TN DEX i : , LlP .
OGILYIF , ~ !
ARLA .
M.--lllicroscopic and Systematic icltudy of llfadreporarian T_l'pes of Corals .
Phil. Trans . .
ll , 1896 , vol. 187 , pp. 83-345 .
Corals .
~Iicroscopic and Sy ; tematic Study of Marlreporarian Types .
Oii !
Tie , Maria ~1 .
Phil. Trans. , ll , l8lJ6 , vol. 187 , pp. 83-345 .
[ 83 ] III .
Mic1\#183 ; oscopic and Systematic Study of JJ[adreporcu\#183 ; ian Types of Corals .
By Miss MARIA M. OGILVIE , D.Sc .
Commnnicated by Sir ARCHIBALD GEIKIE , F.B.S. Received July 24 , _:.._Read November 21 , 1895 .
CONTENTS .
Part I. Microscopic Structure of the Skeletal Parts .
Introductory : Historical Account of the Literature Chapter I. Microscopic Investigation of the recent genus Galaxea , , II .
Recent and Fossil Astraeid Types . .
, , III .
Classificatory Characters of Astraeidae , , IV .
Fungia and Siderastraea .
, , V. The Family Fungidae . . .
PAGE ; .
12a .185 , , VI .
Eupsammia : Systematic Position of the Eupsammidae .
193 , , VII .
Some " Perforate " types-Tu1\#183 ; binaria , JJ !
fodrepora , Porites .
203 , , VIII .
General Review of Macroscopic and Microscopic Skeletal F~atul'e_s .
223 Part II .
Phylogeny of Madreporaria based upon Skeletal Striwture .
Chapter IX .
Evolutionary Changes .
253 " " " X. Transitional Nature of Triassic and Jurassic Genera .
XI .
Suborders of ]Jadreporaria XII .
General Scheme of Classification List of Abbreviations in figures .
PART I.-MICROSCOl'IC STRUCTURE OF THE SKELETAL PARTS .
INTRODUCTORY .
HISTORICAL ACCOUNT OF THE Ll'l.'ERATURE .
THE following account brings forward the most important contributions , zoological and palreontological , to the study of the origin , the microscopic structure , and morphology of the Madreporarian skeleton .
The earlier writings in the first half of this century are neglected , since exhaustive accounts of those have been given by BRONN , and by EDWARDS and HAIME .
DAN A's"\lt ; brilliant researches on recent corals and their habits , introduced many * D.A.NA , " Exploring Expedition of Captain WILKES .
" " Zoophytes , " 1846 , Atlas , 1849 , Philadelphia .
M 2 11.1.97 fundamental changes in the systematic arrangement of corals .
The large folio atlas published by him in connection with the WILKll : S ' Exploring Expedition in 1849 , is in constant use in every large School of Zoology and Palreontology , and has made known the beauty and variety of the existing Madreporarian world to a wide circle of naturalists .
Very soon this work was followed by another , equally great , which embraced fossil as well as recent corals .
The publication of the " Natural History of Corals , " by EDWARDS and HAIME* in 1857 , marks an era in the scientific literature of the Coelenterates .
Their work treate1l Alcyonarians and Zoantharians , the class Zoantharia E.H. including Actiuian , Antipatharian , and Madreporarjan corals .
The classification of the Madreporaria , which EDWARDS and HAIME carried out , still holds its place in all text books .
Slight departures from it have been occasionally suggested in special works , but only such as affected a small group of genera .
The broad systematic features are still accepted in the sense laid down by Enw ARDS and HAIME .
During the latter half of the century , our zoological knowledge of the Madreporaria has made enormous strides of progress , largely as a result of the material brought by the " Challenger " Expedition .
A question which has common interest both for zoologist and palreontologist is the relation of the soft parts of the polyp to the hard calcareous or horny skeleton produced in most corals .
It ltas been found that the mode of development of the skeleton may be very different in the difierent sub-classes of Coelenterates , and that within each sub-class the greatest variety of form may obtain , both with regard to the elemental parts of the skeleton and their particular manner of com bina.tion to build up a connected framework .
The " Sclerite " cind the " Spicule " of EDWARDS and HAIME .
At the time when EDWARDS and HAIME wrote , comparatively little work had been done in this direction .
Those authors were familiar with the spicules of Alcyonarians , and assumed that the compacter Madreporarian skeleton was made up of similar structural elements .
They recognized tw:o layers of tissue in the body-wall of the polyp- a " dermic " and an " epidermic " -and said that the " derniic .
" tissue gave rise to " sclerodermites " or " sclerenchymatous nodules " ; that those remained free in Alcyonarians , but united together so as t.o form cribriform or compact lamellre and trabeculre , making up various skeletal parts of Madreporaria ( septa , wall , costre , synapticulre , colurnella ) .
The epidermic tissue laid down thin , impeiforate larninre , * MILNE EDWARDS and J. HAIME , " Histoire Naturelle des Coralliaires , " Paris , 1857 .
A series of separate publications preceded this work , 0\#163 ; which the most generally known are " Recherches sir la Structure et la Classification des Polypiers , " Paris , 1848-9 , and " A Monograph of the British Fossil Corals , " London , 1850-55 .
SYSTEMATIC STUDY OF 'l'HE MADREPORARIAN TYPES OF CORALS .
85 forming the " sclerobasic axis " in Alcyonarians , the basal tabulrn and dissepiments , and the epitheca in Madreporaria .
In a few cases individual sclerodermites could also be observed in these epidermic laminre ex. Melithrna ( ' Coralliaires , ' vol. 1 , p. 33 ) .
The sclerenchymatous nodules laid down by the dermic tissue were said to be of two kinds:(1 ) Fusiform-shaped , partly horny , partly calcareous in compo~ition .
( 2 ) Irregular , six-sided bodies , in which the calcareous element dominated .
Both varieties might occur in the skeleton of the same polyp ; indeed , the figures illustrative of both are nodules belonging to Alcyonium digitatum ( loc. cit. , vol. 1 , p. 31 , Atlas , Plate Al , figs. ld , 3c , 3d ) .
In the descriptive text accompanying the figures the term " spicule " is used for the fusiform-shaped nodules , and " sclerite " for the irregular , more calcareous nodules .
We find , however , that the distinction thus made between " spicule " and " sclerite " soon became invalid .
KoLLIKER used both terms synonymously in his ' !
cones , ' published 1864 , applying them exclusively to the skeletal bodies , lying often closely packed , but always individually distinct , in the dermal tissue of Alcyonarians .
EDWARDS ' a. , nd HAIME 's view , as enunciated in 1857 , that the Madrepontrian skeleton was built itp of Sclerites , in part or wholly coalesced .
The " Complete Sclerenchyma " was the term used by EDWARDS and HAIME for the Madreporarian skeleton , and was defined by them as the " result of the successive and continuous development of the calcareous nodules ( " sclerites " ) , which , instead of being produced in a miliary manner throughout the whole surface of the dermal tissue at once , invade it at one part and another , and unite together as solid Iamellre , the whole of which form the fundamental and essential part of the corallum , properly so-called . . . .
In the least perfect form of this sclerndermic tissue each nodule produces small branches , at whose extremity other nodules are developed ; this takes place in such a way as to form linear series , which have the appearance of small knotty stems constricted from time to time , and to those we have given the name of " Poutrelles."* In a slightly more perfect form of the scleroderm these nodules produce lateral branches as well , which completely coalesce with those of neighbouring nodular series ( poutrelles ) , and transform a row of trabecula into a kind of cribriform lamella .
Usually the sclerodermites approach each other more closely ; their branches multiply considerably ; soon there remain no more spaces between them , and a continuous and compact lame1la results from their union " ( Zoe .
cit. , vol. 1 , p. 32 ) .
The Madreporarian septum , was said by the same authors ( loc. cit. , vol. 1 , p. 57 ) to " show all the modifications described above in speaking generally of sclerenchymatous lamellre .
" Farther , they said the septum was composed , in the Poritidre and all trabeculate coral types , of a single lamella or leaf ( " feuillet " ) , but when the septum was compact one could often distinguish " two parallel lamence or leaves , united * " Pout.re1les " = " Trabeke1 " of PRATZ .
The nomenclature of EDWARDS and HArnE was retained by PRA'l'II ( see below , The Family Fungidre , ' pp. 185-7 ) .
86 DR. MARIA lVL OGILVrn ON THE ' MICROSnOPIC AND either directly or by an intermediate tissue of a deeper shade , " e.g. , Desmophyllum .
In other cases an empty " space could be seen between the two leaves , e.g. , Mussa fistulosa .
' ' The " intermediate tissne " in the middle of a compact septum corresponds with what \\7as called the " dark line " of the septum by subsequent authors .
The conception of EDw ARDS and HAIME 'was altogether more a general scheme on which they might base their clam , ification .
In histological details it has had to undergo much correction .
But it showed a thorough grasp of the leading structural features , and had the great merit of giving prominence to a unit of structure , viz. , the sclerenchymatous nodule .
To this " unit " they sought to trace back all the varieties and complexities shown by the skeleton in different Madreporarian families .
Nothing could be more concise than the description quoted above of the building-up of the Madreporarian sclerenchyma .
It is like the bold sketch of a master artist thrown upon the canvas , into which the apprentice hand might later work the details .
At the same time , criticized scientifically , the whole scheme was ventured upon an assumption , since the individual " sclerenchymatous nodule , " to which reference was ma , de , had only been proved to exist in Alcyonarians Its presence in Madreporaria had been argued from analogy .
The surface of the free Alcyonarian spicule may be smooth , granulate , spiny ; the same varieties may be oLserved in the parts of the " complete sclerenchyma .
" in Madreporaria .
Again , the regularity of the rows of granulations on the surfaces of the septa in many Madreporarian genera , as well as the regularity of the " branchlike " union between trabecuhe in other genera , was sufficient to suggest that " nodules " were placed in vertical succession upon one another .
The irregular " sclerite " might afford many points of contact with neighbouring sclerites , hence the wholly calcareous " scleritic " form of nodule , instead of the " spicular , " was regarded by EDWARDS and HAIME as the unit of structure in Madreporaria .
Its origin was also referred to the dermal tissue , in which it wa.s known that the free " spicules " and " sclerites " were developed in Alcyonaria .
Correction of Special Points in EowARDS and HAIME 's Statement of Co ?
'Ctl Structure was made by LACAZE DuTHIERS ~nd KoLLIK~R in 1864 .
The first correction of the general laws laid down by EDWARDS and ; HAIME was made by LACAZE DuTHIERs , * when he proved that the hard axis of the red coral originated in the mesoderm ( = dermic tissue , E.H ) , and was not an epidermal structure ; farther , the axis originated as free spicules , and not in laminate form .
The spicules became subsequently coalescent by means of farther calcareous deposit .
It was then thought that the sclerobasic axes of other Alcyonarians were formed in the same way , until , some twenty years later , VON KocHt proved that EDWARDS and * LACAZE DU'l'HIERS , " Histoire Naturelle du Corail , " \amp ; c. , Paris , 1864 .
t G. v. KOCH , " Die morphologische Bedeutung des Korallenskelets , " 'Biolog .
Ceutralblatt , ' Bd .
2 , Nr .
19 , January , 1883 .
SYSTEMATIO STUDY OF THE MADREPORARIAN TYPES OF CORALS .
87 HAIME had been partly right .
In the Gorgonia polyp , and possibly in certain other Alcyonarians , the sclerobase was of epidermic origin .
In 1864-65 appeared KoLLIKER's* great work or } the fine tissues and skeletal_ structures of Protozoa and Crelenterates .
The second part was devoted to Hydroz9a and Actinozoa , but , among the latter , chiefly to Alcyonaria .
Only one short reference is made to the Madreporarian corals ( loc. cit. , Part 2 , pp. 168 , 169 , :6.g .
A ) .
The figure is a transverse section through a part of the skeleton of an " Astrrea , " showing groups of fibres radiating outward from points not exactly central in position .
KoLLIKER writes in the context:-\#183 ; " These calcareous fibres are grouped in such a way that they produce sometimes star-shaped figures ( fig. A ) , sometimes feather-like forms . . .
Farther , one observes in these calcareous fibres lines of layering ( Schichtungslinien ) , which , in indi , \#183 ; idual cases , occur so regularly at intervals of 0\#183 ; 002 to 0\#183 ; 004 millim. that they remind one of the transversely-striped fibres of the enamel of vertebrate teeth .
In many species , however , these lines are only very indistinctly marked , or cannot be recognized .
" \#183 ; KoLLIKER was of opinion that the skeleton was formed in the main by a calcification of the " inner crenenchyma " of the polyp , but that no free calcareoits bodies were formed in the process of ca .
Zcijication ; " for , in the first place , the soft ccenenchyma of the l\iadreporaria , so far as has been already investigated , possesses no calcareous bodies ; and , secondly , even the thinnest lamellre of a corallum show .
now here a trace of such bodies .
' ' Ea1\#183 ; lie Views on Septal Symmetry in Corals .
Meantime several of the first palreontologists of the day were occupied with Palreozoic corals belonging to the subdivisions Madreporaria rugosa , E.H. , and M. tabulata , E.H. The names of LINDSTROM , NICHOLSON , KuNTH , suffice to recall the nature of the work which had begun in the decade 1860- 70 .
As I have limited my attention in the present paper to the M. rugosa among Pal:::eozoic corals , I shall indicate here only the main contributions which concern that group .
The characteristic feature of Palreozoic , in comparison with Mesozoic and Cainozoic \#183 ; Madreporaria is not the possession of the tabulate base in the central area of most genera , neither is it the thick , rugose epitheca or theca commonly present .
It is the pinnate mode of development of the septa in relation to THREE particular septa and the resulting bilateral symmetry of the calyx .
KuNTHl " first proved the laws whici1 governed the insertion of new septa in most Palreozoic corals , the appearance of main and counter septa in one diameter of the calyx , and of two alar septa in the cross diameter ; the appearance of additional " pairs " of septa on either side of the main septum and on the remote side of each of the alar septa .
The new septa are " - ' A. KoLJ .
JKER , 'Icones Histiologicre , ' Leipzig , 1864-65 .
t KuN'rH , " Beitrage sir Kenntniss fossiler Kora1len , " 'Zeitsch , d. D. Geol .
Ges .
, ' 186~-70 . . .
' - ' coalescent at their point of origin always with the main septum or with one or other of the a1ar septa , and this gives externally a characteristic branching appearance of the septa from the epitheca inwards and upwards towards the centre of the calyx .
Although this feature is generally shown throughout all stages of growth in Palreozoic Madreporaria , there are frequent enough examples of Palreozoic species in which the septa appear , : pinnately " only in the earlier stages of development , and afterwards are inserted irregularly or with more or less radial symmetry .
KuNTH 's observations had a great influence on the thought of the time , and his results seemed conclusive that tetrameral or bilateral symmetry of the septa characterized the majority of Palreozoic Madreporarian corals , while radial symmetry was accepted as the almost invariable rule in Mesozoic and Cainozoic corals .
According to EDw A.RDS and HA .
IME , the radial symmetry was founded on a cyclical number of six primary septa , simultaneously developed , and new septa were inserted in definite " orders " of subsequent cycles , the full number of septa being always a multiple of six .
HA .
ECKEL proposed the recognition of two main subdivisions of Madreporaria , ( i. ) M. tetracora .
Zla , corresponding to " .L11 .
rugosa , " E.H. , Palreozoic corals with tetrameral symmetry ; ( ii .
) M. hexacoralla , the Mesozoic and Cainozoic corals witn cyclical symmetry of septa , including M. aporosa and M. perforata , E. H. According to those views , there seemed an immeasurable gap in the evolutionary history of Madreporaria between Palreozoic and Mesozoic corals .
The Publication of LA .
CA.ZE DuTHIER 's " Development of Corals " in 1872 .
LACAZE DuTHIERS~~ brought forward in 1872 more exact observations on the developmental stages of corals , and his results deviated in very important features from the laws laid down by EDw A.RDS and HA .
IME .
His researches included the development both of Actinians and of Madreporaria , and the main results are familiar in every text-book .
In the case of Astroides , an Eupsammid common in the Mediterranean , he proved that the septa did not begin to be formed until twelve primary mesenteries had been developed .
These mesenteries appeared in pairs , two pairs always following in rapid succession ; in other words , developmental pauses took place after the appearance of four mesenteries , then after eight , and after twelve .
At the last pause , the twelve mesenteries became equal in growth , and only then the septal invaginations of the basal body-wall took place as a cycle of twelve radial pouches alternating in position with the mesenteries .
The insertion of ' further septa and mesenteries is comparatively of less importance than those of the primary cycles .
From LACAZE DuTHIER 's account of development it was made clear that , at any * H. de LACAZE Du'l'HIERS , " Developpement du Coralliaire , " Prem .
Mem. , sir Jes Actiniaires sans polypier .
( ' A rcb .
de Zool. Exper .
, ' 1872 , vol. 1 , pp. 289-396 ) .
Sec. Mem. , sir les .A.ctiniaires a polypier .
( ' .Arch .
de Zool. Exper .
, ' vol. 2 , pp. 269-348 ) .
SY3TE : M:A.TIC STUDY OF THE :M:ADREPORA .
RIA .
N TYPES OF CORA .
LS .
89 rate in Astroides , the radial symmetry was not base\lt ; l on the presence of a primary cycle of six radial invaginations , as EDWARDS and HAIME had supposed .
On the contrary , the paired insertion of mesenteries followed precisely the same relative order of insertion as that which governed the tetrameral symmetry of the Palreozoic corals .
Nearly ten years later , KocH~ ; proved , in Caryophyllia and other aporose corals , that the increase in the cycles of septa did not take place according to EDw ARDS and HAIME 's laws for Hexacoralla .
The " Primary " Median Lamella , a:i , d " Stereopla-smi'.c " Lateral Laye1\#183 ; s distinguished in the Septa of Madreporaria .
A nearer relationship of Tetracoralla and Hexacoralla being little thought of at the time of KuNTH 's work , the interest of palreontologists centred in the working-up of special faunas characteristic of certain geological ages and provinces .
About the same time valuable contributions were being made to Palreozoic Madreporarian literature by NICHOLSON , LINDSTROM , DYBOWSKI , SCHLUTER , and attention came to be more and more directed to the preparation of thin microscopic sections .
The broad septum of Zaphrentoids and Cyathophylloids , as well as the free septa .
I spines of Cystiphylloids , were repeatedly figured .
( I have here adopted the recent nomenclature of Palreozoic groups given by NICHOLSON , ' Manual of Palreontology , ' I. ) .
The " two-leaved " septum described by EDw ARDS and HAIME was the general appearance presented by sections of Palreozoic Madreporarian septa , and in nearly all cases a median broader or narrower area represented the " intermediate tissue of a deeper shade " described by the same authors .
The origin of the intermediate tissue was not understood ; the great variety in its appearance , and its small thickness relatively to the rest of the septum , was perhaps its most striking feature .
LINDSTROM regarded it as a primary median septal leaf of independent structure , while the lateral layers on both sides of it were called by him " ste\#183 ; 1\#183 ; eoplasm , " or secondary calcareous deposit of lamellar nature.t DYBOWSKI , on the other hand , distinguished only two leaves or laminre in the septum .
A suggestive paper , written by VON KocHt in 1882 , showed a certain return to the fundamental doctrine of EDWARDS and HAIME , that the skeletal structures were compos~d of a number of individual units .
But the nomenclature had altered ; * G. v. Koc H , " :M:ittheilungen iiber das Kalkskelet der Madreporaria .
I. Das Vermehrungsgesetz der Septen .
" 'Morph .
J ahrb .
, ' Bd .
VIII .
, 1881 .
t LINDSTROM , " Contributions to the A.ctinology of the Atlantic Ocean .
" 'Kon .
Svenska .
Vet .
Akad .
Handl .
, ' Bd .
XlV .
, No. 6 , p. 17 , 1877 .
t v. KocH , " Mitt .
iiber die Structur von Pholidophyllitm Loven ?
' .
, E. H. , u. Oyathophyllurn , sp. ?
aus Konieprus .
" 'Palreontogr .
, ' vol , 28 , p. 216 , l 882 .
90 DR. MARIA M. OGILVIE ON 'l'HE MICROSCOPIC AND instead of " sclerites " and " sclerenchymatous nodules , " we hear only of \#183 ; septal " spines , " such as exist"d in numerous Palreozoic corals .
And this was only natural , since the septal " spine " had been made familiar in a number of figures , whereas the " nodule " remained unproved as an individual structure .
KocH instituted a direct comparison between the microscopic structure of the septal spines in a Palreozoic genus and that of the " two-leaved " compact septa , both in Cyathophyllum , a Palreozoic genus , and in Caryophyllia , a recent Turbinolid .
He recognized a distinct area of indefinite structure in the middle both of " spine " and of " septum , " and outside it a lamellar " stereoplasm " of fibrous structure .
From that time forward the view of LINDSTROM was adopted with more security , and was generalized for the structure of the septa in Madreporaria of all ages .
The brilliancy of the decade 1872-82 in Zoological Research .
A strong stimulus had been given to the zoological research of Actinozoa by the embryological works of LACAZE DuTHIERS .
We may , in fact , regard the decade 1872-82 as a high tide in the 7.oological literature of corals .
The late Professor MosELEY* was then producing his brilliant series of papers touching various classes of Crelenterata .
His general reports of " Challenger " material appeared in No. II .
of the " Challenger " Expedition volumes , and several papers on special genera appeared in different magazines .
Amongst the most interesting of his researches were those on dirnorphisrn in the zooids of the Alcyonarian Heliopora crerulrea ; t and again the dimorphism in the families Milleporidre and Stylasteridre.t He proved that both those groups , which had been formerly classed with Actinozoa , were in reality Hydrozoa .
The researches of the HERTWIGst gave insight into the fine histology of the muscula , ture , the nerve-system , and the generative organs of the Actinians .
A couple of years later , in 1882 , Professor R. HERTWIGS published his 'Monograph of the Actinians , ' belonging to the " Challenger " Expedition material .
The " Report of the Reef Corals " in connection with the same expedition was executed by Mr. QUELCH .
II KLUNZINGER 's , r three volumes of the 'Corals of the Red Sea , ' appeared in the years 18 77-80 , accompanied by a series of photographic plates .
This author suggested * MOSELEY , " The Structure and Relations of the Alcyonarian Heliopora Coorulrea , " \amp ; c. , ' Phil. Trans. Roy .
Soc. , ' 1876 .
t " Structure of a Species of Millepora occurring at Tahiti , " 'Ann .
Not .
Hist. , ' 1876 .
" Preliminary Note ou the Structure of the Stylasteridoo , " 'Ann .
Not .
Hist. , ' 1877 .
t H1m'l'wIG , 0 .
\amp ; R. , " Die Actinien an::i , t.-histol .
untersucht , " 'Jen .
Zeits .
, ' vol. 14 .
S HER'l'WIG , R. , " Die Actinien der ' Challenger ' Expedition .
" Jena , 1882 .
II QUELCH , " Challengei Reports , " vol. xvi .
, 1886 .
This report is notable for the description of the curious Recent , coral , Moseleya , as well as for the reasons given by Mr. QUELCH in favour of the abandonment of the Rugosa ( Tetracoralla ) as a separate sub-order of Madreporaria .
- , f C. B. Kr.uNZINGEH , " Die Korallthiere des Rothen Mere \lt ; Js .
" Berlin , 1877-80 .
" Wachsthum der Korallen , " 'Wtirtemb .
Jahreshefte , ' 1880 ; Jahrg , 36 , p. G2 .
SYS'TEMATIC STUDY OF THF , MADREPORARIAN TYPES OF CORALS .
9i many changes in the systmnatic arrangement of Madreporarian families and subfamilies made by EDWARDS and HAIME ; in many cases agreeing with VERRILL in preference .
A separate paper , written by KLUNZINGER , in 1880 , gathered together the special data concerning modes of budding and the increase of Madreporarian polyps .
1881-82.-The Discovery of the Calicoblastic , or Lime-forming Cells , and of the Ectodermal origin of the Madreporarian Skeleton .
Two authors , v. HEIDER~ ; and v. KocHt may be said to have almost simultaneously discovered the existence of a skeletal-forming layer of cells in the Madreporarian polyp .
The same year saw also a purely paheontological work devoted entirely to the discussion of microscopic details of the skeleton and applying those with success to the systematic grouping of genera .
The series of microscopic researches published by P.aATz+ are unique in their kind ( vide p. 20 ) .
Careful study of this paJreontological work indicates a certain connection between the observations made by P.aATZ on Jurassic and Cretaceous genera and those made on recent corals by v. KucH and v. HEIDER .
But , from the fact that each author had already concluded his own work before seeing the other 's , written contemporaneously , the zoological and the pala:iontological interests have remained , so to speak , estranged and have mutually missed all the significance which they might have had .
Moreover , v. KocH and .
v. HEIDER did not agree in the details they gave of the skeletal-forming cell , and the matter has been somewhat in abeyance .
The zoologists who have since published papers support v. KocH ; hitherto no palreontologist has taken up the question .
As I find it to be one of considerable importance , I shall enter fully into the points of agreement and difference between KocH and HEIDER .
HEIDER proved in the paper quoted , the general identity in the anatomy and histology of Cladocora with facts previously determined for Actinians .
He writes , " The brothers HRRTWIG came to a conclusion ( based on their investigations of the mesenteries in Actiniarn : ) with which , so far as my research has gone , I can only agree .
It is , in the main , that Actinians and Cora.ls ( Madreporarian ) are still more closely allied with one anoth8r than they have as yet been regarded " ( loc. cit. , p. 28 ) .
He then describes for the first time the lime-forming cells , to which he gave the name " calicoblast .
" He found that the cell-nucleus was active during the earlier change8 of the cell-contents from organic into inorganic conditions , but gradually shrank as the cell became filled with the granular or fibrous deposit .
From that he concludAd that the cell ceased its organic life when the change had completed and became * A. v. HEIDER , " Die Gattung Cladocora , Ehrenb .
, " 'Sitzungsber .
der K. Akad .
der Wiss .
Wien , ' December , 1881 .
t G. v. KocH , " Uber die Entwicklung des Kalkskeletes von Asteroides calycularis u. dessen Morphologiscbe Bedeutung , " 'Mittb .
Zool. St. Neapel .
, ' 1882 , p. 284 . !
E. PRATZ , 'U eber die verwandtschaftlichen Beziehungen einiger Korallengattungen mit haupti ; achlicher Beriicksichtigung ihrer Septalstructur , ' " Palreontographica , " vol. 29 .
N 2 bit J\1ARiA J\f .
OGILVIE ON THE :MICROSCOP1d AND 7Jractically a small body of calcareous fibre , S. These calicoblast cells , were iu REIDER'S opinion , of mesodermal origin ; he thought that the ectoderm apparent on the basal embryonic disc disappenred before maturity .
HEIDER's- paper was published in December , 1881 , and in the same month KocH had finished his manuscript , " On the Development of Astroides calycularis , " quoted above .
Therein he describes how the ectodermal layer of cells on the basal embryonic disc of Astroides continues in the mature polyp and separates out the calcareous deposit of the various skeletal parts .
Unlike HEIDER , KocH had observed no calcareous particles or fibres within the ce1ls of this ectodermal layer , and he therefore looked upon them as secreting cells , actively sPparating out calcareous matter , and at the same time continuing their own organic existence .
HEIDER had not pursued his research to the skeleton , in order to see what became of the calcified cells which he h'ad seen .
KocH , on the other hand , had in previous papers already given special attention to the mature skeleton .
He now showed , by means of transverse sections , that the embryonic skeletal deposit was laid down in the form of individual calcareous " spheroids " or " ellipsoids , built up of concentricallylayered , always rhombic crystals .
" The1=e grouped together here and there , and gave rise to an embryonic skeletal disc .
Then the polypal wa11 pouched itself upwards in a ring of radial folds , and the ectoderm on both sides of the pouches deposited similar " spheroids " or " ellipsoids .
" In this way a cafoareous , radiallyplaced structure took origin within each of the invaginated pouches , and these radial structures were the septa of the young Astroides polyp .
KocH , however , did not follow out the mode of combination of these " spheroids " or " ellipsoids " to form the fully-developed septum .
Yet the transverse sections of the septum of Garyophyllia which he published in the same year ( vide ante , p. 89 ) , give no sign whatever of being composed of individuaLspheroids or ellipsoids , but show fibres arranged parallel with one another and radiating outwards on both sides of a median darker plane in the septum .
These embryonic " spheroids " of Astroides gave therefore no basis for general statements with regard to the Madreporarian skeleton .
An important feature of the embryonic spheroid may be at once noted .
KocH \#183 ; writes that near the peripher_y and towards the centre of the basal disc , the spheroids were more or less free from one another , and seemed then to attain what might be regarded as their normal development .
" At such places , " he says , " the \#183 ; most of them showed very plainly a dark streak dividing them into two halves .
" HEIDER took the opportunity , in a later paper , * of comparing KocH 's description of the lime-secreting cell with his own .
After an examination of Astroides and Denclrophyllia , HEIDER so far modified his earlier opinion as to say that the calicoblast cells might be ectodermal in origin , and wander towards the edge of the mesoderrn , forming there one or more cell-layers .
He still maintained the complete * A. R. v. HEIDER , 'Arbeiten aus dem Zoolog .
Ins. zu Grai , ' vol. l , No. 3 , " Kon ; illenstudien .
" Leipzig , 1886 .
SYS1'EMA'i'IC STUDY OF 'l1HE MADRElPORARIAN 'l1YPES OF CORALS .
93 conversion of the calicoblast cell into inorganic deposit , and gave figures with fibre and granule-containing cells .
The crystalline " spheroids " and " ellipsoids " of KocH might possibly , he suggested , be formed by the combination of a number of the changed calicoblasts .
Although HEIDER had made thin transverse microscopic sections of ClCldoeora , he did not advance our knowleclge of the fine structure of the mature skeleton beyond the stage which KbLLIKER had reached .
In fact , HEIDER frankly admits in his paper , in 1886 , that he is at a loss to account for the difference between the simple deposit observed in the embryonic skeletal disc , and the various complex structures of the mature skeleton .
He surmises " that the individual skeletal parts in the Madreporaria are formed in a manner analogous with the skeleton of Alcyonarians , " but that in the former the calicoblasts produce a connected calcareous layer at the outer limit of the mesoderm , " whereas the analogous cells in the Alcyonarian polyp form , as is well known , calcareous spicules , which remain isolated throughout the whole life of the polyp " ( loc. cit. , p. 180 ) .
This surmise of HEIDER 's shows that he stood much in the same position with regard to the skeleton of Alcyonarians and Madreporarians , as EDW .ARDS and H. AIM had done thirty years before .
Instiiead of the " sclerenchymatous nodule " ( equiv. sclerodermite , sclerite , spicule ) which the authors of " Coralliaires " had pronounced to be common to both groups , the term " spicule " was now limited in its application to the Alcyonaria , while the " spheroid " or " ellipsoid " of KoLLIKER and KocH depicted the unit-form in certain Madreporaria , and the " septal spine " in others .
Too little was known of the " spheroid " to venture upon either its closer identification with , or differentiation from , the " spicule , " yet there was always an inclination towards identifying .
Short Review of Kocn 's Works on the Maclrepo1\#183 ; Clrian Skeleton .
VON Koen has done more than any other author in investigating the details of microscopic structure in different groups of Maclreporaria .
In 1877 he published a short account of Stylophora cligitcita , PALL .
As.* The habit of the coral is branching ; the calices are very small , about 1 millim. diameter , canaliculate coenenchyma extending between the calices and supporting the coenosarcal continuations of the polypal body wall on slight echinulations .
I may state here that the echinulations on the coenenchyme are , according to my results on microscopic structure , the exact homologue of " granulations " on the surfaces of septa , \amp ; c. Koen described the skeleton in its minute structure as " built up of more or less regular ellipsoids , often extended to a considerable length , " and agreeing with that .
* G. v. KocH , " Mittheiln .
iiber Ccelenteraten , Anatomy von Stylophora digitatci , PALLAR , " 'Jenaischc Zeitschr .
, ' vol. 11 , pp. 375-81 , Plate 2~ , 1877 .
94 bR .
MARIA M. OGlLVIE ON THE MICROSCOPIC AND previously figured by KoLLIKER .
Of the echinulations , he writes : " The small irregularly outlined prominences on the surface appear to be the ends of the incompletely developed crystals which make up the individual ellipsoids .
" The illustrative sections are both transverse and longitudinal .
The transverse sections show the zig-zag outlines round the radiating groups of fibres called spheroids and ellipsoids ; * but the longitudinal section ( Plate 22 , fig. 6 ) does not carry out the idea of either a spheroid or ellipsoid of fibres .
Instead of individual symmetrical groups of fibres , there is a continuous long axis , which corresponds to the centre of radiation in the transverse section , and from which the fibres diverge in a uniformly upward and outward direction .
Such a longitudinal section gives indisputable proof that the skeleton of Stylophora is NOT built itp of individual ellipsoids .
The terms " spheroid " and " ellipsoid , " as applied at any rate to Stylophora , are merely descriptive of au appearance seen in transverse section .
Nevertheless KocH constantly stated in subsequent works that the Madreporarian skeleton was in most cases made up of calcareous spheroids and ellipsoids .
In 1879 a short paper was published by KocH , t in which we learn that he had tried in vain to isolate the " crystallographische einheitliche Element " in the skeletons of different sub-cla.sses of Anthozoa .
He found that in all the types examined , small rhombic crystals were the ultimate structures , but he could not discover how these crystals arranged themselves into the larger crystal-groups , nor could he demonstrate that any such groups ( ellipsoidal or spheroidal ) were separable .
So far as Alcyonarians came in question , he could state that the crystals were arranged concentrically and .
regularly in the spicule , yet in such a way that their axes more or less crossed each other .
" In all other forms of calcification , " KocH concludes , " I have not got much beyond what is already known " ( loc. cit. , p. 322 ) .
His next important paper is that in 1882 , which I have already referred to on p. 89 .
He makes there a minute examination of two Palreozoic and one recent species , and has nowhere occasion to mention " ellipsoidal " or " spheroidal " groups of crystals .
He compares the solid septum of Cyathophyllum sp. ?
with the septal spines of Pholidophyllum Loveni , and finds their nearest structural analogue in the Alcyonarian spicule .
As this paper of VON Koca 's has largely influenced palreontological literature , I shall give a sketch of its contents . .
Pholidophyllum Loveni+ possesses no leaf-like septa ; but longitudinal series of * Stylophora : a transverse section may be seen in NICHOLSON'S 'Manual of Palrnontology , ' vol. 1 , P\#183 ; 245 , fig. 125A .
t G. v. KocH , " Bemerkungen iiber das Skelet der Korallen , " 'Morph .
Jahrb .
, ' Bd .
5 , 1879 , pp. 316-22 , Plate 22 .
t Pholidophyllwrn Loveni , a Silurian species , had been described under various generic names .
EDWARDS and HAJME desc1\#183 ; ibed it as Oyathophyllum ( ?
) Loveni ( ' Brit. ~'oss .
Corals , ' p. 280 , Plate G6 , iigs .
2 , 2A ) .
LINDSTROM removed the species from the genus Oyathnphyllitm , and erected for it a new genus Pholidophylliim ( ' On some Operculated Corals , E-ilurian and Recent , ' by G. LINDSTROM .
Wishy , 1870 ) .
He called attrntion to a " covering of small , ( \#189 ; millim. ) , very thin scales , tightly clustered SYSTBMATIO STUDY OF THFJ MADREPORARIAN TYPES OF CORALS .
95 septal spines project from the wall into the lumen of the calyx .
The spines may vary in length and thickness , their microscopic structure is , however , always the same .
An internal mass , according to KocH , of small crystals , indifferently arranged , occupies the central part of the spine , which appears clearer in the centre and darker at its peripheral area , by transmitted light .
The peripheral area is then surrounded by a zone of radiating calcareous fibres , which always is transparent and clear .
Lamellar lines of thickening a.re shown passing from spine to spine , and giving ri : : ; e to a thecal thickening composed of their compacted basal portions .
This lamellar thickening is the equivalent of LINDSTROM'S stereoplasm .
Transverse sections are then given by KocH of the broad leaf-like septa and theca in Cyathophyllum .
sp. and Caryophyllia cyathus ( recent ) .
He writes : " A dark streak runs along the middle of the septum more or less regularly , it appears to be sometimes made up of several individual parts , and possesses a radiating structure .
At the sides of this dark streak , which I shall name the 'primary streak ' ( ' Primarstreif ' ) , a clearer , likewise crystalline , and usually finely striped mass lies , which is limited by a darker edge , but is not very sharply defined from the perfectly transparent mass of thickening ( Ausfti.llungsmasse ) .
" Sometimes LINDSTROM'S term " stereoplasm " has been applied by other authors only to the outer " Ausfollungsmasse , " or " secondary thickening " of KocH in leaflike septa ; sometimes it has been used to include both the outermost thickening and the finely-striped crystalline fibres as described by KocH .
When used in this latter sense , it has then been directly brought into contradistinction with the dark " Primarstreif " of KocH .
This want of uniformity in the use of the term " stereoplasm " arises mostly from the vagueness of ideas with regard to the essential structure , as well as an uncertainty how much of the septum is really " primary " and how much " secondary .
" And it is hopeless to attempt carrying out a strict terminology , when authors are in doubt about the principles which are meant to be conveyed by the terms .
KocH does not enter into the question of the difference between the " Primarstreif " in its microscopic character and the radiating fibres .
He simply makes a general comparison of the darker central mass in the septal spine of Pholidophyllum with the primary streak of the septa in many Madreporaria , or with the darker , inner part of the large spicules of many Alcyonarians , while the clear , radially-built together in longitudinal rows along the costre ; " these he thought to be " secretions of an exothecal nature " homologous with the opercula of certain Rugose corals , with the epithecal valves of Oysti phyllum prismaticum , and possibly also with the scales and opercular valves of recent Prirnnoa , Alcyonarians belonging to the Gorgonidre .
This homology , indicated by LINDSTROM in 1870 , derives new significance , since we now know that the whole skeleton of Madreporaria is of ectodermal origin ( cf. LINDSTROM , 'Geol .
Mag. , ' 1866 , Plato 14 , fig. 40 , and ' Ofvers .
Vet .
Ak .
Forh .
, ' 1868 , Plate 6 , fi g. 4 ) .
Similar scales are present on the epitheca of Rhizophyllum Gotlandicmn Roem .
, from the Silurian beds of Gothland .
outer part would answer to the corresponding parts in the corals named .
The lamellar deposit , which represents the wall in Pholidophyllum , he regards as identical with the " secondary layers of thickening " in Madrepora and other recent corals .
In the same year , 1882 , KocH 's embryological paper on A:stroides calycularis was published ( ante , p. 91 ) , and the primary skeletal deposit was described by him as entirely composed of calcareous spheroids and ellipsoids .
In the following year a paper* appeared which may be regarded as a resume of this author 's results on the structure and morphology of the Madreporarian skeleton .
He distinguishes three main types of structure among corals:(1 .
) " The simplest forms of calcareous skeletal structures are the so-called spicules observed in Alcyonarians " ... .
" further , in a number of fossil corals , one is led to think of the possibility of the skeleton being composed of coalesced spicules , as , e.g. , with some probability in Pholidophyllum , then , perhaps , also in Syringophyllum and similar genera .
" ( 2 .
) " The connected calcareous skeleton of the Madreporaria .
" " The fine structure of the Madreporarian skeleton is fairly simple .
It is always composed of crystalline spheroids .
" ( 3 .
) The skeletal structures of the fossil Rugosa appear to be closely associated with those of recent Madreporaria , since considerable identity may be recognized in the fine structure of both .
t These results are far from presenting any general truth with regard to the skeletal structure , and its mode of deposition .
Neither do they yield any sound basis for morphological \#183 ; comparison .
For example , the type of structure which KocH NICHOLSON , + and others had observed in transverse sections of the recent genus Caryophyllia is that indicated in Group ( 3 ) above as typical of Rugose corals of Palreozoic age .
On the ot.her hand , close analogy had been demonstrated by KocH in 1882 between this type of structure in Group ( 3 ) and that demonstrated for Pholidophyllum .
A year later , the latter was brought into affinity with the Alcyonarians in Group ( l ) above , and separated from any connection with its Rugose and recent neighbours of the previous year .
The septal spine of Pholidophyllum is , in short , looked upon as a kind of spicule , and if we were to follow KocH 's papers * KocH , " Morphologische Bedeutung des Korallenskelets , " 'Biolog .
Centralblatt , ' vol. 2 ; No. 19 , pp. 586 , 587 , January , 1883 .
t This refers to the septa of Oaryophyllia and Oyathophylforn , as well as to a number of examples given by KocH , and other authors .
In 'Palreontographica , ' 1882-3 , Bd .
29 , KocH 0ontributed an article on budding , " Die Ungesrhlechtliche Vermehrung ( Theilung und Knospung ) einiger Palceozoischen Korallen , " in which occasional reference is also made to septa !
structure ( Zoe .
cit. p. 347 ) , " Stauriaand Fascicularia show , in their general form , as well as in the fine structurfl , great identity with the recent genera , .Mussa , Oaryophyllia , Gulaxea ( the coenenchyme excepted in the last ) .
Everywhere distinction can bfl plainly made between the primary layer of the septum and the stereoplasm .
" t NICHOLSON gives an instructive figure on septa !
structure in 'Manual of Palreontology , ' vol. 1 , p. 247 , figs. 127 , A , B , 0 , where Oarvophyllia is compared with " Rugose " types .
SYSTEMATIC STUDY OF THE MADREPORAR !
AN TYPES OF CORALS .
9o/ of 1882 and 1883 , the inevitable conclusion to be drawn would be , either that the compact septum of Cyathophyltum , Caryophyllia , \amp ; c. , was composed of individual spicules , or that the whole septum had a structure entirely analogous with a single spicule .
The type of structure in Group ( 2 ) of KocH is based upon the spheroid and ellipsoid seen in trans1 , erse sections of certain Madreporarians , and in embryonic stages of Astroides calycularis .
This type is characteristic , it is assumed , of recent corals ( Madreporarian ) ; at the same time the Rugose corals seem closely allied .
After all , therefore , a review of KocH 's work leaves us with a number of facts and several important figures , but in as great confusion as ever regarding the ultimate structural unit in Madreporaria .
It is almost a relief t_o return to the sclerenchymatous nodule of EDWARDS and HAIME .
For what is this spheroid ?
Is it a spicule 1 We can see and separate the spicule , then why not the spheroid ?
It is noteworthy that the spheroid has been presented from first to last by one or two transverse sections , no longitudinal section has demonstrated it .
The " septal spine " of Pholidophyllum , might be compared with a spicule , it certainly never could be called a spheroid .
Neither had the primary streak and stereoplasm of Caryophyllia , \amp ; c. , ever betrayed , in figures given , any individual structure akin either to a septal spine , a spicule , a spheroid , or sclerenchymatous nodule .
Progress of work on the Septa of Palceozoic Corals .
As I have already said , several palreontologists engaged in the study of Palreozoic corals prepared their sections for microscopic work , but the state of preservation was seldom sufficiently good to permit of all the details being seen .
NICHOLSON* accomplished much in the field of Madreporaria tabulata E. and H. , and for the Monticuliporoids .
He has also published a number of sections illustrating i , ; pecial work on the M. rugosa .
In 1879 , ScHLUTERt published thin sections accompanying his work on Devonian corals , and the same author continued his special study of the corals of the Eifel and Rhine districts in 1880-1-2 in a series of papers.+ The thin sections show the general relations of the different skeletal parts , but lead us no further with regard to fine structure .
FRECH's~ papers , also on Devonian corals , give several sections in which more detail may be seen .
A longitudinal section made through a septum of * NICHOLSON , " On the Structure and Affinities of the Pa1ooozoic Corals , " 1879 ; " Monograph of the Girvan Silurian Fossils , " NrcHOLS0N and ETHERIDGE , 1878-80 .
t SCHLUTER , " Devonische Korallen und Diinnschliffe v. Spongophyllum Kunthi u. Calophyllum paucitabulatum , " ' Sitzungsber .
d. Vert. de Naturh .
ver .
d. Pr .
Rh .
in Westphalen , ' 1879 , p. 402 .
t Later papers appeared in 1880-81 of the same journal ; " Zoantharia rugosa aus d. Rbeinisch .
Devon .
Mittel und Ober , " ' Sitzungsbericht , No. 3 , Ges .
Naturf .
Fr. -zu Berlin , ' 1880 ; " Ueber Never Korallen \lt ; l. Mitteldevon d. Eifel , " December , 1882 , 'Sitzungsber .
d. Niederrh .
Ges .
f. Natur .
and Heilkunde , p. 203 .
S FRECH , " Die Korallenfauna des Oberdevons in Deutschland , " 'Zeitschr .
D. Geol .
Ges .
, ' 1885 , p. 21 .
08 1JR .
MARlA M. OG1LVIE ON 'l'HE MH : JROSCOP10 ANO Cyathophyllum minus Roem .
sp. ( loc. cit. , 'Zeitschrift , ' 1885 , p. 31 , Plate 1 , fig. 2c ) , shows several rows of dark points , each row forming a central zone surrounded by light-coloured " stereoplasm .
" In the latter , no structure of any kind is presented .
The author describes this septum as consisting of coalesced spines directed upwards and inwards .
The rows of dark points in longitudinal section correspond to the primary streak ( Primarstreif , KocH ) in the middle plane of the septum .
FRECH 's sections , therefore , corroborated KoCI-r 's comparison of the structures of the Pholidophyllum spine and Cyath')phyllum septum ( cf pp. 253 , 264 ) .
In the same year FRECH published a sequel to his paper on Devonian corals , entitled " Ueber das Kalkgerti.st der Tetrakorallen " ( ' Zeitschrift , ' 1885 ) .
As I shall have occasion to refer to this in Part _II .
of the present paper , I shall merely remark hem that it gives a short account of the main structural types recognized in the septa of Tetracoralla up to that date .
A good longitudinal section is given of Enclophyllurn , showing dark axes of the septal spines surrounded by " structureless stereoplasm " ( Taf .
41 , fig. 6 ) .
No further work with regard to septal structure is given by FRECH in his next paper/ published 1886 .
W AAGEN'st description of the permo-carboniferous fossils of the Salt Range includes a number of specially interesting Coelenterate types .
The Rugosa are represented by .Lonsdaleia and Arnplexu , s. A figure of a transverse section of Lonsdaleia Indica shows septa " distinctly composed of two parts , a central one , which is of dark colour , and a peripheral or coating one , which is light coloured and finely fibrous .
" The darkcoloured part , which can , however , when differently preserved , appear as a distinct white line , represents what has been call :d by KocH the " primordial stripe " or " primordial septum " ( Zoe .
cit. , p. 899 , Plate 101 , fig. 3b ) .
The fibrous structure of the " peripheral " part of the septum is shown very c1early in transverse and longitudinal sections given by SCHAFER+ in the course of a paper on Phillipsastrrea .
" The septa , " he writes , " are formed of bundles of calcite fibres which lie on either sitle of the median plane of the septum .
From this starting point on this p ] an each fibre passes upwards and outwards in the direction of the proximal end of the septum .
" He compares the " feather-like structure thus produced " with that described by PRATZ for Tharnnastrrea ( see below , pp. 187 , 188 ) .
In these " fibrous bundles , ' ' proved by PRATZ for the first time in certain Mesozoic and younger genera , and by SCHAFER for the first time in a Palreozoic genus , we have a correct rendering of microscopic longit , udinal sections which showed fine details of structure .
No use is made by either of those authors of the terms primordial septum , stereoplasm , * FRECH , " Die Cyathophylliden und Zaphrentiden des Deutschen Mitteldevons , " 'Pal .
Abb .
, ' vol. 3 , Part III .
, Berlin , 1886 .
t WAAGEN , " Salt Range Fossils , " 'Palreontologia Indica , ' ser. 13 , vol. 1 , 1887 , Calcutta .
: ) : R SCHAFER , " On Phillipsastrcea n'ORB , " ' Geol .
Magazine , ' London , September , 1889 , vol. 6 , p. 398 , Plate 12 .
SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF \#183 ; coRALS .
99 secondary thickening , outfilling mass , ellipsoids , spheroids , spicules , \amp ; c. !
SCHAFER demonstrated farther that each calcite fibre was a crysta1lographic individual , as KocH had shown .
in certain recent Madreporaria .
And he pointed out that the vesicles and tabulre agreed with each other in their histological structure , but differed from the septum in so far as the position of the deposit was concerned .
" In the tabulre and vesicles the calcification could only take place from below upwards ; the septa , on the contrary , were arched over by the ectoderm , and the calcification could continue both upwards and on either side .
" Progress of Work on the Septa of Mesozoic and Cainozoic Corals .
One of an important series of systematic papers was published in 1887 by ORTMANN ( see foot note , p. 185 ) , in which this author , like QUELCH , abandoned the suborder Madreporaria Rugosa .
He then laid down a new system of classification ( see pp. 312-3 I 4 ) .
In 1888 , a paper appeared by HIND * on the typical species of the genus Septastrma , S. Fo1\#183 ; besi , E.H. , from the Miocene Formation of Virginia and Maryland .
He said that the " primary streak " of the septum in this genus was really double .
The septa could readily be split in the middle plane , and the inner face on both halves differed from the outer .
Each of the inner layers had a lattice-like appearance , being composed of well-marked linear ridges , parallel with each other , and either vertically or obliquely placed , while transverse growth-lines crossed the ridges at intervals .
HIND pointed out that the powdery calcareous deposit present in those inner ]ayers might possibly be the result of subsequent change and disintegration in the course of fossilization , and need not necessarily be looked upon as the original form of the skeletal deposit in the " primary streak .
" The peripheral parts of the septum were composed of radiating fibres .
In 1890 , a " Monograph of the Coral Fauna of the Triassic Strata " in different parts of the Eastern Alps was published by FRECH.t The systematic aspect of this Monograph is treated below ( Part II .
, Chapter X. ) ; little special work was done on the microscopic structure of skeletal parts .
The only thin section is one made transversely through a calyx of Isastrma profunda , REUSS , var .
nov .
Mqjor ( loc. dt .
, p. 22 , Plate 5 , fig. 4c ) .
I shall quote the author 's description , as it best shows the standpoint at present taken by FRECH:"The microscopic structure of the septa agrees with that observed in the case of Devonian Cyathophyllids .
The wall ( thec'1 ) and the septa originally consist of light-coloured spines ( trabeeulre ) and are later thickened by a darker-coloured skeletal mass ( stereoplasm ) .
The primary spines are , in cases of good preservation , plainly visible as isolated pillars in the middle of the septa , and th eir diameter is * G. J. HIND , " On the History and Characters of the genus Septastrcea D'ORB .
, " 'Q .
J. Geol .
Soc. , ' J 888 , pp. 200-226 , Plate 9 .
-t F. FRECH , " Die Korallenfaunfl , der Trias , " 'Palreontographica , ' vol. 37 , 1890---91 .
0 2 no longn added to , even in the most strongly-developed septa .
The strengthening of the septa and wall takes place wholly by deposited stereoplasm , which shows clearly the structure frequently described by PRATZ , VON Koen , and myself .
The " primary streak , " composed of primary spines , is placed , as a rule , somewhat excentrically within the septum .
Numerous spinate continuations ( " Seitendornen " nov .
norn .
) stand out from the sides of the septa ; in their arrangement they follow the course of the original primary spines , notwithstanding that they consist , like the sides of the septa , of stereoplasm ... .
" We may conclude , from the analogy dr:awn with Cyathophyllum that FRECH believes a single granulate ridge in an Astrreid septum extending from the wall inwards is the equivalent of the septa !
spine in Pholidophyllum , Cystiphyllurn , and other Palreozoic genera .
But he gives to the term " spine " a quite special application .
By it he means only the central part , that part which KocH had described in Pholidophyllum as clearer centrally , and darker peripherally , when seen by transmitted light ; that part which FRECH himself had shown to be composed of a series of dark points connected or disconnected in certain species of Cyathophyllum , and which , in other cases , he describes as " isolated rods " or " pillars " in the midst of a structureless stereoplasm ; that part which W AAGEN and FRECH himself found very variable according to fossilization , and which HIND surmised had been probably entirely disintegrated in Septastrrea , and whose ultimate appearance could not be looked upon as trne indication of its original condition .
Yet the " primary streak , ' ' or " primordial septum , " is said to Le composed of those questionable ' ' primary spines .
" And all the rest of the septum , including the " radiating fibres , " as well as the " A.usfiillungRmasse " of KocH , is looked upon as " stereoplasm , ' ' or secondary thickening .
I shall have occasion to show , in the course of my paper , how entirely untenable those ideas of FRECH are with regard to primary spines and septum .
Since FRECH refers to PRATZ , in the matter of this stereoplasm , I may remark that PRATZ never for a moment confounds with stereoplasm the radiating fibres which wholly composed the individual trabeculce of the septum in the genera he examined .
He used the term " trabecula " in quite a different sense from FRECH-in the sense of " poutrelle , " E.H. and never had occasion to differentiate any central " spine " or " trabecula " from an outer stereoplasm or secondary thickening .
I do not quite see why FRECH introduces the new name of " Seitendornen , ' ' since the septal eminences , to which the name is applied , have always been known satisfactorily enough as " granulations .
" I may so far anticipate the results of my own work as to say that these granulations have , in the Triassic Astrreids examined by FRECH , a very distinct structural bearing .
Not the whole of a septal ridge , in Montliva1 tia , Thecosmilia , \amp ; c. , is the equivBclent of the primitive " septal spine , ' ' seen in Cystiphyllum and other Palreozoic genera , but only so much of it as corresponds on the : : ; urface of the septum to one of these granulations .
An important contribution to our knowledge of the septal types among Jurassic corals was made by KoBY ( see foot note , p. 27 5 ) in a s~ries of figures with descriptive notes .
SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
101 Progress of Zoological Work on Recent Madreporaria .
Several papers by BouRNE and by FOWLER , published in the 'Quart .
J ourn .
of Microscopic Science , ' may be said to have continued the anatomical work which had been done in England by MOSELEY , in Germany by KocH and HEIDER .
In referring to the papers of BOURNE and FowLER , I direct my attention here to points which concern the Madreporarian skeleton .
Morphological and anatomical features will be referred to in the course of this paper .
The excellent microscopic transverse sections of wall and septa deserve to be noted , they are among the best which have been given .
Also the various terms for the median area of the septum--primary streak , primary septum , primary spines -are scarcely used by those authors .
The particular areas in question are fitly called in general " centres of calcification , " whether the area be seen as a " line ' ' or a " point " in transverse section .
The " centre of calcification " was , with those authors , a neutral term , taking nothing for granted about the dark line ( ex. Euphyllia Flabellum ) or the dark points ( ex. Mussa ) themselves , but indicating merely the fact observed that the calcareous fibres~ in all cases , diverged outwards from those areas .
Both BouRNE and FOWLER had abundant opportunity of confirming KocH 's statement that the skeleton was entirely of ectodermal origin .
At the same time , they accepted HEIDER 's term of calicoblast for the lime-forming cells of the ectodermal layer .
They farther agreed with KocH in regarding the calicoblast as a secreting cell , separating out the calcareous material which gave rise to the skeleton , while itself remaining intact as an organic cell of the polypal ectoderm .
The obserYations made and figured by HEIDER of calicoblasts containing fully-developed fibrous deposit , and the nucleus shrunken or vanished , were entirely discredited both by BOURNE and FOWLER .
Another explanation was given by FOWLER of the curious , elongated\#183 ; cells with fibre-bundles , observed by HEIDER ; while the roundish cells with calcareous , granular , and fibrous contents were little regarded .
BouRNE'-l- sa.id HEIDER might have inferred the rod-like ( fibrous ) appearance within the cells could not be calcareous in structure , otherwise the rods would have been dissolved by the action of the acids used in decalcifying the specimens .
In defence of HEIDER I would draw attention to the fact that the latter saw those fibres in cells , many of which were still highly protoplasmic , and that my experience has been a very slow action of decalcifying acids on the contents of such cells .
The skeleton proper dissolves rapidly , but the superficial layers next the body-wall of the polyp dissolve slowly and incompletely , leaving protoplasmic remnants .
On the other hand , both FOWLER and BOURNE recognized two main varieties of ca1icoblasts in the ectoderm , " rounded or polygonal , soft-looking , granular cells , " and " long , narrow , columnar cells .
" The same varieties had been described by HEIDER , * BOURNE , " On the Anatomy of Mussa and Euphyllia , and on the Morphology of the Madreporarian .
Skeleton , " ' Q. J.M. S. , ' No. CIX .
, p. 25 , Aug. , 1887 .
in the soft condition , and also in a more calcified state containing the " Stabchen , " or small rods .
To FowLER , ,'i : first , and then to BOURNE , we owe an important observation with regard to the position occupied by the long columnar cells .
Those clothe , in Lophohelia and in Mussa , the uppermost , most actively-growing parts of the septa at and near the periphery of the calyx , whereas the ectodermal cells elsewhere are smn.ller in these two genera .
In the genus Fimgia no long columnar cells were observed by BouRNE , the whole ectoderm being composed of smallish celJs .
FOWLER and BouRNE have accomplished much for palreontology ~y confirming , without doubt , the presence of a continuous ectoderm as the external layer of the body-wall in Madreporaria .
This layer produces no skeleton on the oral surfaces of the polyp , but it is the active skeleton-former throughout its whol~ aboral surface , whether smooth or crumpled , single or invaginated .
Those authors thus refuted , conclusively , the idea from which HEIDER has never freed bis mind , that the mesoderm lay next the skeleton , for nothing else can be supposed to be meant when HEIDER writes tha , t the " calicoblasts wander in from the ectoderm to the edge of the mesoderm .
" At the same time there was an end made to any direct comparison between the origin of the Alcyonarian spicule and that of the various iSkeletal parts of Madreporaria ( antea p. 93 ) .
I find , however , that KocH , FowLER , and BouRNE are wrong in their conception of the calicob]ast .
My investigation of superficial layers of the skeleton have given fullest confirmation of HEIDER 's idea , that the calicoblasts were actualJy converted into calcareous groups of fibres .
I hope to show , in the course of this paper , that the fibre-containing calicoblasts which lie next the skeleton are shed off , so to speak , from the polyp , new cells constantly taking their place in the ectoderm by celldivision .
The shed calicoblasts build up successii1e layers of calcified cells , which hang together at first by their cell-walls , and ultimately , as crystalline changes continue , form the individual lamina : of the skeletal structures .
The whole Madreporarian skeleton is composed of such larnince , any apparent variation in microscopic structure is accounted fo1\#183 ; by some difference in the particular shape and position , locally , of the ectoderrn .
The Microscopic Researches of l'RATZ .
I briefly referred to the work of PRATZ in connection with the publications which marked the year 1882 ( antea p. 91 ) .
His rernarch deals chiefly with a number of fossil genera which had been referred to Fungidre and to Poritidre by EDWARDS and HAIME .
Some of those are Cyclolites ( Cret .
) , Leptophyllia ( Cret .
) , Thamnastrcea ( Tri.-Tert .
) , Trocharcea ( Up .
Jur .
) , Microsolena ( Up .
Jur .
) , Astrceomorpha ( Tr. ) .
* FOWLER , " The Anatomy of the Madreporaria , " III .
, 'Q .
J. M , S. , ' Aug. , 1887 , " Lophohtdil ) prolifera , " p. 9 , SYSTEMATIC STUDY OF T.H:E MADREPORAR !
AN '11YPES OF CORALS .
103 Of recent genera , Coscinanea and Siderastrceci are included , but those were not examined in the same detail as the fossil genera .
PRATZ was enabled , by means of careful selection of well-preserved material , and the preparation of very thin sections cut in all directions , to attain a clearness of conception with regard to the essential structure of the skeleton in those genera .
He held himself wholly within the limits of the types examined , building up for them an analytic table of their classificatory features , as complete in its way as it also was .original .
I shall defer the consideration of this table to later parts of the present paper ( Part I. , chaps .
iv .
and Y. , Part II .
, chap .
x. ) .
It is not too much to say that no other author shows in the same high degree as PRATZ a thorough acquaintance with and mastery of the past and current literature of Madreporaria , and has held himself at the same time so free from mental bias , and so much aloof from the changing opinions and the controversial interests of contemporaries .
The result is , that his work is one of the few which is from beginning to end alive with scientific instinct .
It is a work whose high value can only be slowly recognized , since few are sufficiently versed in the study to realize what PRATZ accomplished at a time when nothing but confused terminology and hazy ideas of skeletal structure were finding expression in the rapidly written papers of some of his fellow-workers .
I have never made the personal acquaintance of this author , yet his original microscopic sections , now in the possession of the Munich Museum , have largely influenced my work .
The terminology which PRATZ uses is EDWARDS ' and HAIME 's .
He shows that in the fenestrated septa of Microsolena , Thamnast1\#183 ; cea , \amp ; c. , the trabeculre ( poutrelles , E. H. ) are composed of a series of nodules , and the nodules are themselves made up of smaller symmetrically-arranged prominences .
The structure of the trabeculre is wholly tibrous , radiatiug :fibres passing upwards and outwards in bushlike or pinnate fashion from an ideal axis in the centre of the trabecula .
PRATZ writes in one place that he " thought he could distinguish smaller fibre-bundles passing to each of the smaller prominences of a nodule ; " of this he was , however , uncertain .
What he proved was the complete bush-like group of fibres corresponding to each nodule of the series , in Gyclolites for example ( cf :fig .
38c , below ) .
My work will show that these smaller :fibrous bundles to which he referred are universally present in Madreporaria wherever the skeletal surface is finely granulate .
Although microscopic sections of Madreporaria had been frequently made and described before PRATZ published his work in 1882 , he was the :first to apply such work systematically .
The genera examined were mostly grouped by PRATZ in a sub-family PReudoastrreinre , which he referred to the family Fungidre , at the same time pointing out a certain similarity in their habits of growth with Astrreinre .
A few genera were formed by PRATZ into a sub-family Pseudoagaricinre ; this subfamily will be shown below to be unnecessary .
I have purposely removed the work of PRATZ out of its due succession in the 104 DB- .
MA .
RI.A M. OGILVIE ON '1.'H.1\#163 ; MICB .
OSCOPIC A.ND historical review of the literature , because my own paper may be looked upon as the direct continuation of the methods which PRATZ used .
Purpose of the Pr\#183 ; esent Paper .
My interest in Madreporaria began while I was still a student m Professor RAY LANKESTER 's Class of Zoology and Comparative Anatomy , at University College , 1889--90 .
I then determined to make the group a subject of special research , and to study its palreontological aspect .
With this in view I came to Munich in spring , 1891 .
There , by the permission of Professor VON ZITTEL , I had all the advantages offered by the large coral collection in the Pa1reontological ' Museum of Muni ch University .
Dr. RUDOLF SCHAEFER , Custos of the Museum , who had already published special work on Palreozoic corals , kindly helped me in my early studies .
To him I owe much .
Amongst other things , he pointed out to me the interest attached to microscopic investigation of fossil corals , taught me various methods of making sections , and has throughout my farther work guided me in my selection and use of literature .
In this way he directed me towards the line of study which I have mainly fo1lowed , and whose results are embodied in this paper .
At the same time the necessity for such work was being borne in upon me by my own experience in preparing a monograph of the Upper Jurassic coral fauna from Stramberg .
The longer I worked with the identification of genera and species , the more dissatisfied I became with the conventional methods of work .
The synonymy of Jurassic corals is wont to be fairly extensive , and it soon becomes clear that the individuality of the author has mw::h to answer for in the erection of genus , species , or variety .
No fixed standard of research can be traced in the various monographs of the special faunas , and the general results obtained by different authors may be highly satisfactory from a particular paheontological or geological point of view , but cannot be said to have an absolute value for the science of Madreporarian evolution .
When my descriptive work for the Stramberg monograph was completed , and I tabulated the list of species in order to compare them with the fossil faunas next in age , I felt that the apparently exquisite Upper Jurassic age of my fauna had been very largely a foregone conclusion !
W'hile the list gave no indication of the close relationship which existed , on the one hand , between the fauna I had examined and any typical Palreozoic fauna , and on the other , between it and any typical recent fauna .
The arbitrary line cutting off the Tetracoralla or Madreporaria Rugosa ( Palreozoic ) from all later Madreporaria had been disproved by a goodly number of the Stram berg corals .
Neither was the radial arrangement of septa in cycles by any means very definite or very general in the Jurassic types .
There was , therefore , a certain protest against the first basis of sub-division in Madreporaria , viz. , the tetrameral or hexameral symmetry of the septa .
SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
105 Again , a certain sameness runs through the external macroscopic features of most of the Madreporarian genera , which happen to be well represented in Jnrassic strata .
The genits can , as a 1\#183 ; ule , only be identified with certo.i'.nty ofter a careful microscopic investigation .
Many genera erected by FROMENTEL canuot be criticised or made use of because of the want of such data .
Again , one finds in some authors the habit of the corallum outweighing any considerations of fine structure .
Under the genus " Latirnceand\#183 ; m , " as many varieties of Madreporarian types are hidden at the present day as could ever have been al tributed to the name of " ~Jadrepora " in the earliest works on corals .
In short I found myself face to face with th_e fact that the genera and families of EDWARDS and HAIME were no longer quite in touch with more recent work , and yet no recent work gave any general standard of research .
The literature seemed split up into ever so many factions .
The only hopeful prospect seemed to be in microscopic work , and that I determined to do before publishing my monograph of the Stram berg fauna .
The collection of St. Cassian corals which I had myself made offered a number of specimens in admirable state of preservation .
Besides , these corals being of Triassic age had a particular phy]ogenetic interest attached to them .
Dr. HoFER , Custos of the Zoological Collection in Munich , kindly handed over to me several recent corals which he had collected from the Red Sea .
These were then specially prepared for the examination of the relation between the polypal and skeletal parts .
Dr. SCHAEFER placed at my disposal his own collection of recent Madreporaria , which has greatly facilitated my studies .
I selected carefully the recent corals , with which I went to work , as far as possible making an examination of the skeleton of corals whose internal anatomy had already formed the subject of special zoological study .
I chose also leading types from the Astrreidm , Fungidre and Perforata ; good microscopic sections had ah\#183 ; eady been published of Caryophyllia and Flabellnm , two well-known types of Turbinolidre .
Thus I hoped to arrive at some fairly embracive series of different structural types -if such existed in il1ad1'eporaria .
The result has proved that different structural types do exist , and form a surer method of determining Madreporarian genera and groups than any other method known to me .
While engaged in the work of this paper , I learnt that Dr. VoLz , * Assistant to Professor FRECH in the Breslau Museum , had commenced a monograph of the St. Cassian corals , which should form the second part of " Triassic Corals .
" The first part of this publication by Professor FRECH had appeared in ' Palreontograpbica , ' 18\#163 ; 10 .
* Since this paper was entered at the Royal Society , I have received the Introduction to the paper on St. Cassian Corals by Dr. VoLz , and it deals with microscopic features observed in these Triassic corals .
" Die Korallenfauna der Schichten von St. Cassian in Siid-Tirol .
" ( Inaugural Dissertation by WILHELM .VOLZ ; Stuttgart , Schweizerbart'sche V ~rh.g , 1895 .
) I therefore abandoned my original intention of including St. Cassian types in the present paper , and have limited my treatment of Triassic corals to thA already published Zlambach fauna , ( vide chapter x. belvw ) .
As there is a representative collection of this fauna in the Munich Museum , I have been in a position to examine original spECimens and make microscopic investigations when necessary .
I am glad to have this opportunity of thanking GEHElMRATH VON ZITTEL for his kind permission to work in the museum and library of the Palreontological Institute -a permission to which has been added much personal encouragement throughout these years .
GEHEIMRATH VON ZITTEL has , moreover , kindly revised for me the system of Madreporarian families which I propose in the concluding chapter of this paper .
CHAPTER !
.-MICROSCOPIC INVESTIGATION OF ' THE RECENT GENUS GALAXEA .
Tr-rn SKELETAL PARTS .
The genus Gc1laxea is a colony-building form in which the polypal tnbes are surrounded by a delicately-formed ccenenchyme , only the actually inhabited calicinal Fig. lA .
car .
car .
Fig. 111 .
s. C. th .
c11 .
en .
Branching corallite of Galaxea with surrounding camenchyme ( twice natm-al size ) .
tali .
ps.c .
Enlarged drawing ( diagrammatic ) of half of the calyx , showing the skeletal parts : s = septum , c = costa , th = tbecn , , tab = .tabula , ps.c .
= pseudoco1umelln , en = coonenchymc , .
portion of the tubes rising freely above the ccenenchyme .
Two corallites are shown in fig. lA , twice their natural size .
The upper parts of the calyces have been cut across .
The upper surface of the ccenenchyme is about half an inch below that of the corallites .
Part of the ccenenchyme has been removed in order to show the branching of the corallites .
Fig. lB shows the lanceolate , exsert septa , and their relations to wall and basal structures .
Dnriug the life of the colony , the ccenenchyme is covered by living tissue , the SYS'rruSIA .
TIC STUDY OB ' THJ~ MAOREPORA.lUAN TYPgs OF CORALS .
107 , : ccenosarc .
" The ccenosarcal tissues are the continuation of the body-wall of the corallites , and enclose canalicnlar extensions of the visceral cavities of the polyps .
It is\#183 ; convenient to distinguish an " oral " and an " ctborcll " body-wall in the sense in which FOWLER applied those terms , viz . .
, the " oral " body-wall for the upper and outer surface of polyp and of ccenosarc , the " aboral " body-wall for the lower and inner surface .
The latter lays down the various skeletal parts of the corallum , external to its ectodermal layer of calicoblast cells .
The oral body-wall takes no part in forming the skeleton , although the ectodermal layer is continuous with the lime-secreting ectoderm of the aboral wall .
These relations are diagrammatically represented in fig. 2 .
Fig. 2 .
001\#183 ; , A. Car , B. GALAXEA-camenchymatous type .
o. = oral body-wall ; a. = aboral body-wall ; cs .
= ca:mosarc ; cn.=camenchyme ; e. =edge-zone ; s.=septum ; c.= costa ; th.= theca ; tb.= tabula ; is.l.= interseptal loculus .
( Compare the similarly treated figures , 22 , 35 , 36 , 56 .
) The figure is a longitudinal section through two corallites of Galctxea and the ccenenchyme surrounding them .
In Corallite A the aboral body-wall is shown at that part where it clothes one of the principal septa .
A dotted line indicates the position of the same body-wall at the part adjacent to the septum , where it clothes the theca and the basal tabulrn within the theca .
In Corallite B the aboral bodywall is supposed to be cut through at the part resting on theca and tabulre , while the \#183 ; dotted line indicates i~s position on a principal septum .
It w~ll .
be seen fro_m the figure that the septum projects some distance above the theca , i.e. , is " exsert , " and also projects outwards beyond it , giving rise to a costate portion " c. " The aboral wall clothing the margin of the septum is therefore carried out much nearer the oral wall , and the visceral cavity or its continuation over the lip of the corallite is there necessarily narrower than in the adjacent thecal portions .
p 2 ios DR. MARfA M. OGILVIE ON THE MICROSCOPIC AND The '\#183 ; intrathecal " visceral spaces in the interseptal loculi of the calyx are represented outside the calyx by " extrathecal " spaces between the aboral body-wall on the theca and the oral body-wa11 .
The term " Randplatte " has been given by HEIDER and KocH , and has been generally accepted for these extrathecal portions of the visceral cavity .
I propose to translate this useful term into English as the edgezone .
It is understood in the term that the mesenteries of the interseptal loculi are continued into the intercostal loculi .
" Ccenosarc , " on the other hand , signifies an extrathecal part into which the mesenteries do not continue ; a canalicular system may be developed in it , but has not any immediate connection with the mesenterial loculi of the calyx .
In Galaxea a well-developed edge-zone extends over the lip of each corallite , and its body-wall , oral and aboral , is continuous with that of the ccenosarc .
The meseuteries are , however , not continued into the ccenosarcal parts of the colony ( 1iide FOWLER , " Madreporaria , " 'Q .
J. M. S. , ' February , 1890 , " Galaxea esperi " ) .
The under surfa.0e of the whole colony is covered by an epitheca with concentric furrows .
It takes it , - origin from the same part of the aboral body-wall which deposits the ccenenchynie throughout the colony .
The edge-zones of the peripheral polyps are not continued into ccenosarc , but are closed at their lowest portions , and the epithecal deposit is laid dowu by the angle between aboral and oral body-wall.# In the same paper some suggestions are made with regard to the absence of the " Randplatte " in various colonies , a subject frequently referrerl .
to in the papers of FOWLER .
EDw ARDS and HAIME gave to the ccenenchyme of Galaxea the special term of peritheca , as its appearance differed on the one hand from the costal parts between the corallites of many Astrreids , e.g. , Heliastrrea , and on the other hand from the ribbed or _ echinulate ccenenchyme in the Oculinids and the " Perforate " types .
Without entering at present into the question of the morphology of parts , it will be seen from fig. 2 , above , that the " peritheca " in Galaxea is the deposit made by a common ccenosarc , and as such may be included with the ccenosarcal deposit in Pocilloporidre , Madreporidre , and ~thers , under the general term " Ccenenchyme .
" One can distinguish various structural types of ccenenchyme , porous or compact , granulate , costate , smooth , \amp ; c. , but these distinctions are not such as to necessitate special names .
We have then the following structural parts composing the skeleton of Galaxea , ( cf figs. 1 and 4 ) .
( 1 ) Radial structures , the septa and their extrathecal costal continuations .
These have in Galaxea striated surfaces and serrated edges .
( 2 ) Tangential structures , the parts of the wall present between neighbouring * The best general discussion of the relations obtaining between the " Randplatte " of the polyps and the ccenencbyme in ccenenchymatous colonies has been given by BOURNE , in his_ " Anatomy of Mussa \#183 ; n , nd Euphyllia " ( ' Q. J. M , -S .
, ' No. 109 , pp. 41-43 ) .
SYSTEMATIC S1'UDY OF THE MADREPORARIA~ TYPES OF CORALS .
l 09 sept.a .
These are in \#183 ; Galaxea smooth on their inner surface , striated on their outer .
( 3 ) Basal structures , the columella in Galaxea at the central part of the calyx represents irregular prolongations of the septa ; the calicinal tabulm are smooth basal deposits from the aboral wall ; the ccrnenchymal tissue is in Galaxea fine\#183 ; and Yesicular , composed of small " dissepiments , " and is the basal deposit of the ccenosarcal part of the aboral wall .
( 4 ) Epitheca , a concentrically-furrowed basal plate formed as the edge-zones of the peripheral polyps are withdrawn .
Lateral Surf ace of a Radial Structure .
On examining a " radial structure " closely with hand lens , the serrre at the edges am seen to be the pointed tips of a pair of strire on the opposite surfaces .
The strire Fig , 3 , a , d , sr .
str .
g.c. a.cl .
Lateral surface of a radial structure : str .
= str-ire ; sr .
= serrre ; a.d. = area of divergence of strire ; g.r ; .
= growth-curve .
are dotted along their course with minute granulations .
Fig. 3 is an enlarged drawing made from a slightly-polished septum , the actual width of the septum from the columellar to the costal portion is not more than 2 millims. The \#183 ; strire are 110 DR. MARTA lVL OGILVrn ON Tl-IE MICROSCOPIC AND disposed fan-shaped upon the surface , diverging from a line about half-way between wall and colmrn~lla .
The lighter curves running transversely across the strire mark successive septal edges , corresponding to distinct periods of growth in the life of the polyp .
The part between two successive curves represents the addition made to the septum ( incl .
costa ) during a single period of polypal growth , and I shall , in future , call such a part the growth-segment of the septum , or a septal seg11ient .
MrcROSCOPIC SECTIONS THROUGH SEPTA OF Galcixea .
Location of the " Dark Line " in Transverse Sections of the Calyx .
A polished microscopic section cut transversely through a septum of Galaxea shows the appearance , already described in the introductory chapter , of a " dark line " in the median plane , and on both sides of it lighter-looking " stereoplasm " composed of radiating calcareous :fibres .
Galaxea , is one of the forms in which the " dark line " is plainly seen to be broken up into a number of " dark points , " and sometimes short , dark lines .
These component parts of the " dark lines " have been in recent coral literature variously named spines , septal spines , primary spines , trabeculce , longitudinal ridges , while the dark line itself has been called ' ' primary streak , " " prirna1\#183 ; y septum , , " " Urseptum .
" The " dark line " in the wall of Galctxea is similarly constituted of separate " points " and " short lines .
" Its position in the wall is , however , much nearer the inner than the outer surface , a fact indicating that the deposition of calcareous substance takes place chiefly on the extra-thecal part of the thecal fold .
A diagrammatic representation of the position of the " dark line " in the septa and wall of Galaxea is given in fig. 4 .
Fig. 4 .
d.l. d.l. Diagrammatic section , showing the position of the " dark line " in the radial and tangential skeletal elements of Ga.laxea .
Throughout Madreporaria we find that the " dark line " is extremely variable in its position in the wall , but in the septa of all types , the " dark line " is always in the median plane .
\#183 ; SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
111 Each of the radial and tangential structures of the skeleton has two lateral surfaces culminating in a serrated edge , and in the living state the polypal flesh folds over the edge and clothes both surfaces .
In the tangential structures , i.e. , wall , of Galaxea the fold of the aboral body-wan descends much farther over the outer surface than on the inner or calicinal surface .
In the radial structures , i.e. , septa and costal , both the lateral surfaces are clothed by the polypal flesh to equal distances below the serrated edges .
Comparison of the actnal structures in tho skeleton with the transverae sections made of them , displays at once that the " dark line " in the section corresponds with the sharp edges of wan and septa in the skeleton .
Further , that the symmetry of the calcareous deposit jn the septa , and its asymmetry in the walI is correlated with equality or inequality respectively in the length of the flaps of the polypa1 folds .
The consid\lt ; ~ration of these fundamental relations between the polyp and the skeleton is necessary before we can understand the microscopic sections of the dry skeleton .
For the preseut the " dark line " is to be thought of as a structural appearance in the sections correspon\lt ; ling to the sharp angle made by the polypal folds where they bend over septa an\lt ; l wall .
And I shall make no use of the terms " pri.m ary septum " = da rk 1m\#183 ; e , and " stereop1 a sm " or " second ary t h.1 c k enr .
ng " = t l1 e fibrous deposit \#183 ; on both sides of it .
The distinction which those terms virtually express is not valid according to the results of my work .
The Growth-lines of the Septum .
Fig. 5 is a drawing of a typical transverse section of Galctxea , showing in its median plane dark points , and short , zig-zag , cla1\#183 ; k lines , around which fibres are placed , and radiate outwards to the surfaces .
In a fairly thick section that is all that can be said of the fine structure .
But in thinner sections ( fig. G ) an examination with high-power lens shows exceedingly fine wavy lines crossing the calcareous fibres at right angles .
As the fibres are arrauged in groups around the " points " and " zig-zag lines " of the median plane , the wavy cross-lines foll into corresponding curves around those " points " and " lines .
" The same curves are followed by all the cross-lines in one half of the septum , from median plane to surface .
The margin of the transverse section is , in consequence , notched in appearance ( cf fig. 6 ) .
The curves between the notches in the section correspond to granulations of the surface stri~ .
The width from any one of the wavy cross-lines to the next is about '0036 millim. , or , roughly calculated , about twenty-seven of them are pl'esent in one-tenth of a millimetre .
KoLLIKER described , in 1864 , the striped appearance of the fibres in the transverse section of an Astrma , which he figured ( vicle antea , Introd .
, p. 87 ) , and since that time most authors have referred to a " larnellar " appearance of the so-called " Ausfhllungsmasse " or stereoplasm , without making a closer investigation of the sn bject .
]12 DR. MARIA M. OGILVIE ON THE MICROSCOPIC AND The " cross-lines " seen in transverse section represent actual lamellre .
This is proved by the fact that the fibres are crossed by similar wavy lines in sections cut longitudinally .
Tangential longitudinal sections show them best , in radial longitudinal l !
- , ig .
5 .
Transverse section of a few septa and .
part of the wall ( low power ) , .Fig .
6 .
d , p , gr. learn .
gran .
Transverse section 0\#163 ; part of septum and wall showing the radiating lines of structure passing out from the " dark points , ' ' and also the wrtvy cross-lines corresponding to suocessive " growth-lamellre " ( high power ) ; d.p. = da1\#183 ; k points ; g1 .
, latn .
" " 'g.rowth , lamella ) ; gl'an .
= granulation .
sections they are naturally very obliquely cut , and appear with zig-zagged edges irregularly outlined against one another , The lines may be compared with the growth-lines ( ' ' Anwachsstreifen " ) of other animal groups , and each individual SYSTEM .
A.TIC STUDY OF THE M.A.DREPOR .
A.RI .
A.N TYPES OF CORALS .
113 lamella is simply a calcareous-growth-lamella .
In Galaxea we have seen that these lamellre are not smooth plane surfaces , but are thrown , as it were , into a series of gentle pleats around definite axes in the median plane of the septum .
The axes are represented in transverse section by the darker " points " and ' ' short lines " in the centre of each series of curves .
Structural Relation of Septal Granulations to the " Centres of Calcification " in the Median Plane .
Finding that no matter what family of Madreporaria I selected from , the skeleton always showed these wavy growth-lines , I felt it absolutely necessary to examine them in detail .
Within each tiny wave a minute group of crystalline fibres lies .
Sometimes the fibres are almost parallel with one another , sometimes more decidedly divergent .
The inner part of the wave towards the median plane looks darker by transmitted light than the outer curve of the wave .
And since each wave of the wavy line shows the same structure and appearance , the whole succession of lamellre presents in sections a striped or banded appearance .
If a section 'happens to cut the lamellre very obliquely , the ends of the fibres are exposed in successive lamellre .
This gives a bright beaded appearance to the wave edges of the lamellre , and considerably heightens the effect of alternating dark and light bands .
The fibres of the individual lamellre are laid down always in the same direction , and with the same crystallographical orientation , so that , viewed with weaker magnifying power , the short fibres of successive lamellre seem to merge with one another .
The original disposition of the fibreR , end to end , in successive lamellre , is therefore one cause of the apparent radiation of fibres around the dark points in the median plane .
By reason of continued crystalline change , the individual lamellre may become ill-defined , and the fibrous structure arranges itself in definite bundles according to the primary curves of the growth-lamellre .
In other words , a transverse section made through a granulation on one of the side surfaces of the Galaxea septum proves that a radiating bunch of fibres extends into it from a dark point in or very near to the median plane of the septum .
And a transverse section of a septum passing through two opposite or nearly opposite granulations on the two surfaces of the septum shows two radiating bunches streaming outwards , apparently from a common centre , of opaque , structureless aspect , iu the median plane of the septum .
Probably it is owing to this very frequent appearance in transverse section that the " spheroid " and " ellipsoid " came to be used in speaking of the Madreporarian skeleton ( Introd .
, pp. 93-97 ) .
For sometimes the " centre " is really circular in section , sometimes it is elliptical .
I may state at once that , among the nnmerous corals I examined , the individual growth-lamellre were best seen in such specimens as had been not very long removed from the sea-water .
In them the fibres , even viewed with low magnifying power , were plainly seen to be interrupted by lines corresponding to growth-lamellre .
In oldish recent specimens , the radiating groups of long fibres struck the eye at once , while the original lamellar structure was less marked .
In poorly-preserved fossil material the lamellre could usually be traced , but only as mottled markings across a pronounced fibro\#183 ; us structure .
At the same time one cannot generalize altogether , as I have seen almost aU stages in a single specimen .
THE SUPERFICIAL GROWTH-LAMELLA EXAMINED .
The Skeletal Elements Composi11g it .
The _natural qnestion suggests itself-what are the microscopic characters of the outermost growth-lamella upon the surfaces of the septum ?
On examining , I found that not only the surface of the septum , but also the surface of all the skeletal parts of Galaxea , showed a continuous stratum of miuute calcareous elements , laid down like roofing-slates , or like epidermal scales upon one another .
Similar minute Fig. 7A .
sr .
sr .
sr .
= senre ; sc .
= scales .
The irregular lines running roughly parallel with the serrated septa ] edge indicate growthlamellre which have been cut during the polishing of the section .
SC .
Fig. 7B .
s.e. s.e. = septal edge ; gran .
= granulation covered with the calcareous scales .
gram .
elements were present in all other genera which I examined .
To give an idea of their appearance and size , I have shown them in fig. 7 A as they are arranged on two neighbouring teeth or " serrre " at the edge of a septum .
The relative size of serra to septum will appear by comp:1rison with fig. 3 , where a whole septum is uniformly enlarged to about sixteen times its real size .
Fig. 7B is another drawing of the scalelike elements showing them on part of a thicker , more striated and granulate septum .
SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
115 In each of the skeletal elements a number of :fibres is present , as shown in :fig .
8A , drawn under high-power lens .
The fibres are not always in the form of divergent groups , but appear often to lie more loosely side by si.de .
The thick fringe of :fibres at the edge is produced by the similar skeletal elements of subjacent layers .
The Fig. SA .
sc .
Cal careou s scales on the s urface of a dissepiment ( from the cc:enenehyme of Gala :reahighly magnified ) .
breadth of these :fibre-containing elements varies on ai1 average from \#183 ; 01 millim. t o \#183 ; o 15 millim. , the height is much less , about \#183 ; 003 millim. Their natural position on the surface is with the broad edge directed obliquely upwards and outwards relatively to the septal plane , and the fibres arrange themselves in the same direction .
The Isolated Skeletal Element .
I made many attempts before I succeeded in isolating the individual pieces .
They came apart with least trouble from the almost transparent dissepiments of the Gala xea ccenenchyme .
A few lamellre as seen on the surface are represented in :fig .
8B ; these broke down fairly readily into their constituent cells .
When thus separated , most were seen to be entirely filled with a bush of minute fibres , others had granular contents , or partly :fibrous , partly granular ( fig. 80 ) .
Their resemblance to the calicoblast or lime-forming cell , which HEIDER* first discovered and described , was so striking that I at once sought for farther proof of the identity of the two .
It has been frequently mentioned by writers on corals that organic remnants , after removal of the polyp , may be found on the skeleton .
Wherever on the surface I found such remnants , they consisted of calicoblasts which showed in shape , size , and contents , the varieties already drawn by VON HEIDER .
The cells were round , or obovate .
The contents varied from yellowish organic cell-material to the inorganic fibrous condition .
Comparison of my own observations on several corals with the * A. R. VON HEIDER , '1 : id Introd .
, p. 91 , ' ' Die Gattung Oladocora " and " Korallenstudien .
" Q 2 figures given by VON HEIDER left no doubt that the isolated skeletal element was a calcified calicoblast cell .
The scale-like arrangement on the surface presents irregularities-sometimes like thicker zones , sometimes thicker patcheH ( cf fig. 7 A ) .
Those are doubtless due to irregular disposition of the calicoblasts as they originally separate from the ectodermal polypal layer .
We may look upon the superficial layers of the skeletal elements and of incompletely calcified calicoblasts as the outer layers of a many-layered ectoderm .
The ectoderm of the Madreporarian polyp is figured by KocH , FowLER , BouRNE as a simple layer of cells .
I have observed , on the contrary , that a section through soft and hard parts shows an ectoderm , sometimes composed of a simple cell-layer , sometimes several cells deep .
REIDER'S figures indicate a similar observation .
Fig. 8B .
f c. Galaxea dissepiment , showing several growthlamelloo ; those at or near the surface show organic remnants ( org .
) while those below have passed almost wholly into the fibro-crystalline ( f.c. ) condition .
01''.J .
Fig. 8c .
Individual scales separated from dissepiment = calicoblasts in various stages of calcification .
The calicobbsts remain adherent to one another in dense groups or may be more uniformly distributed .
And in this manner they are gradually left behind on the skeleton , and completely calcify , while active cell-division develops constantly new ectodermal cells .
The calicoblasts adherent to the skeleton represent such as were already in course of losing living continuity with the polyp at the time when the polyp was removed from the skeleton .
The above observations made on the skeletal surfaces will be seen to carry out fully the opinion of HEIDER , that the skeletal deposit forms , in the first place , within the calicobbst ( cf. Introd .
, pp. 91-93 and 101 , 102 ) .
I pointed out in the preceding chapter that KocH 's view of the extra-cellular deposition of the skeleton had been accepted by FowLER and BOURNE .
The figures given hy the two last-named authorities are confined entirely to preparations made from completely decalcified SYSTEM .
A.TIC STUDY OF THE M.ADREPOR .
ARIAN 'l'YPES OF CORALS .
117 specimens .
My observations have been made partly on dry skeletons , partly on skeletons from which the soft parts have been freshly removed , organic tissues remaining here and there adherent , and pa , rtly on preparations showing the body-wall in position .
Coalescence of Calicoblasts to Form a Growth-lamella .
Having thus succeeded in isolating the structural unit of the Madreporarian skeleton , it was easy to trace the connection between it and the growth-lamellre .
Actual measurement proved that the distance between lamellar bands was the same as the height of a single calicob]ast-in Galaxea , from \#183 ; 0025 millim. to \#183 ; 0035 millim. And the minute groups of fibres in the individual waves of the " wavy n outline of the lamellre agreed in their appearance and dimensions with those of the calicoblasts , and of the isolated skeletal elements of the superficial layers .
Moreover , in studying the surface of the skeletal parts , I had frequent opportunity of observing that the outlines of the cells became vaguer in proportion as the calcification was more complete .
The growth-lamellre , immediately below the surface , still showed , in sections ) indications of the individuality of the fibrous groups , but no actual limit .
I concluded therefore that the growth-lamence were originally layers of calicoblasts , and that their wavy outlines corrP-sponded to the originally independent cells .
THE FASCICLE OF FIBRES .
Longitudinal Tangential Septal Section .
( Fig. 9 .
) Comparison with fig. 3 will show that a tangential section of a septum will meet the surface strire obliquely to their axis of length .
I tried , with little success , to make slant sections which would meet a single stria parallel with its axis .
Fig. 9 is a small part of a slant section , showing in the axial part of the strire ' ' short dark lines " and knob-like ends , which are , in a general way , directed alternately upwards and outwards towards surface granulations .
The study of a series of sections showed that it was a definite plan of growth which gave rise to this alternation in direction .
At the beginning of any particular period of growth , the calicoblasts are grouped at first around a common centre of calcification at the upper edge of a stria , but , ultimately , two centres are developed on either side of the median plane .
In this way , a pair of granulations are marked out on opposite surfaces of the stria , usually slightly alternate with one another in position .
It will now be clear why in a transverse section of the Galaxea septum there should be sometimes a single centre in the middle of a group of fibres , sometimes two centres , and why a centre may be seen either in the median plane or out of it .
At the same time this is by no means to be looked upon as a general feature of septal structure .
The " double centre " is , indeed , frequently met with , but it seldom happens that the two centres diverge so far from one another out of the median plane as is the case in Galaxea .
It is this which produces , in thick transverse sections of Galaxea , the effect of a zig-zagged dark line in the " middle " of the septum , instead of the appearance common in Astrreids of a straight " dark line .
" Fig. 9 .
f. c.c : .
f. = fascicle of fibres emerging in a surface grauulatiou .
c.c. = centre of calcification .
The tangential longitudinal section demonstrates farther that the calcareous fibres are directed upwards and outwards from the centres , or short axes , of calcification to the lateral surfaces of the septum .
Thus the bundle of fibres is formed emergent at a granulation ( cf antea , p. 113 ) .
To this bundle I shall apply the name of a " fascicle " of fibres .
It will be found that when the granulations are small , as in Galaxea , they represent usually the emergent ends of individual fascicles .
On the other hand , large granulations usually represent a number of fascicle-ends which have become coalescent .
Every fascicle of fibres may bA said to have an axis or ideal line of direction .
The dark lines in the septum are nothing but the ideal axes of calcification around which the active deposition at the edge of the septum took place .
Ea.ch uprightly-standing " axis " in longitudinal sections represents the height added at that part of the septal segment during a single growth-period , the actual deposit being the fibres of opposite fascicles .
The latter are then lengthened during the next growth-period , while a new axis forms directly above the preceding , and a new pair of fascicles takes ongm .
Radial Longitudinal Sections .
Full confirmation of what I have said with regard to successive axes of calcification and successive fascicles is given by radial longitudinal sections of the septum .
SYSTEMATIC STUDY OP THE MADREPORARIAN TYPES OF CORALS .
119 Fig. 10 is a longitudinal section passing through the septum and columella , the thecal part of the septum being left out .
The correspondence of the vertical series of fascicles seen in this section , with the striw on the septal surface ( cf fig. 3 ) , may be very easily demonstrated practically , by watching the various stages during polishing and preparation of a thin section .
It may also be deduced from measurements .
The complete width of this septum is about 2 millims. , and by using the hand lens one can count 15 or 16 strire on the surface .
On the other hand , the width of a series of fibres measured with the micrometer is on an average \#183 ; 12 millim. Pig .
10 .
a.cl .
a.d. The stria , branch out to right and left from an area of divergence ( a.d. ) in the plane of the septum .
The individual fascicles are met in the section , fig. 10 , sometimes transversely to their axes ( e.g. , near the colnmella ) , sometimes more obliquely .
In the latter case the distance from one fascicle in a series to the next above and below it is so slight that the individuality of separate fascicular groups is lost , and the fibres appear to he arranged continuously .
I made also a few microscopic preparations of the thinnest septa with mere indications of granulate eminences upon their surfaces .
These showed conclusively that each granulate eminence , however small , surmounted a " dark point " in or near the median plane .
The usual figure presented was that of a stellate group of fibres ; where the surface had been obliquely cut in polishing , the figure became distorted and elliptical .
In all cases the limiting lines between adjacent fascicles were irregular , and corresponded to the superficial furrows around granulations .
The general bearing of these observations will be better understood in the course of the paper .
Already , however , I have verified the statement which I made in Introd .
, p. 93 , with regard to the homology of the ccenenchymal eminences in Stylophora , and the granulations present on the septa and costre of many Madreporaria .
Both the " ccenenchymal eminence " of Stylophora , and the " septal granulation " of Galaxea are the ~xternal expression of form produced by the growth of the fibres of the fascicle in a definite direction .
Rctdial Longitudinal Section th1\#183 ; ough Wall and Septum .
I have given in fig. 11 a drawing of a longitudinal section which passes very obliquely out of the plane of the septum into that of the wall between two septa .
Only a small pRrt of the septum is introduced , that part in which the series of fascicles are all directed towards the costal continuation of the septum .
Fig.11 .
cos fa .
sep htr , i .
c.th .
c.tr .
Radial longitudinal section through wall and septum of Galaxea .
c.th .
= centre of calcification in the theca .
c.tr .
= centre of calcification in the septal stria or trabecula .
Little fibrous deposit is seen on the inner side of the dark line in the wall , whereas on the outer side , long fascicular groups extend to the serrate costal ridge , which is present on the outer surface of the wall between two septa-costal ridges .
The shadowy line in the drawing giving apparent continuity to the dark centres and axes in the median plane is an attempt to reproduce the effects got by changing the focns of the microscope .
SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
121 The transverse sections , fig. 5 and fig. 6 , show apparently contradictory facts regarding the relations of the " dark lines " in wall and septum .
In fig. 5 they are drawn sepamted from one another , in fig. 6 they appear continuous .
The same corallite varies in this respect in Galaxea at different parts , according to the relative rapidity of growth between septum and wall .
When the " dark line " of the septum passes continuously through the wall , more rapid growth of the septum is indicated .
Structure of Tabul03 , Dissepiments , and Coluniella .
The complications in the skeletal structure of the septa and wall arise as the direct result of the wrinkled and pitted character of the ectoderm which envelops those parts of the calyx ( cf " Mussa , " p. 136 ) .
If we examine the structure of a dissepiment we find it much simpler .
The surface of either tabula or dissepiment in Galaxect is smooth , no serrre , no strire , no granules , and consequently its finer structure betrays no " dark points " and no radiating ' ' fascicles " of fibres .
We have in the dissepiment Fig.12 .
g.l Galaxea dissepiment : g.l. = growth-lamellre .
a smooth structure composed of successive growth-lamellre , laid down on one surface only , and either horizontalJy or slightly arched .
The fibres of the lamella are seen in sections to be grouped in tiny " waves " just like those of the septal lamellre .
Fig. 12 shows the regularity of the banded lamellre and the fibres placed perpendicular to their surfaces .
The columella is not laid down as a flat basal structure in Galaxea , but as occasional continuations of the septal edges .
In any single calcareous piece of the irregular columellar framework , the same structure may be observed as in a marginal serra of thg septum .
Reference to fig. 10 will illustrate the structure .
Brief description of the illustration of a typical Galaxea septum , .
Fig. 13 gives a diagrammatic representation of a Galaxea septum .
It will be seen that the strire diverge towards both sides from a line AB 011 the surface of the \#165 ; D~CCXOVI.-l3 , septum .
At the same time these strice are set with granulations .
In order , now , to show the arrangement of the " axes " of the fascicles and that of the growth-lines , the septum is represented as if cut in different directions .
Fig. 14 shows a septum of Mussa cut similarly , and given here for the sake of comparison with Galctxeci .
Fig. 13 .
C E F A B Diagrammatic view of the structure of a Galaxea septum .
AB direction of the " area of divergence " of septal strioo . .
CD tangential section through a pair of strioo on the opposite surfaces of the septum .
1 !
)F tangential section cut obliquely to the direction of strioo .
Tn the upper part of the :figure one must imagine the one half of the septum removed and the median plane exposed ; in the lower part of the figure the septum is complete and the striated surface seen .
Corresponding to the strice in the lower part , vertical series of fascicles are present in the upper part , and from the axis of each fascicle the calcareous fibres diverge in an upwardly radiating direction .
At the upper edge of the septum the fascicles give rise to :fine serrce .
SYSTEMATIC STUDY OF THE IYiADi:tEPORA"RiAN TYPES O~ .
, CORALS .
123 A horizontally-cut surface is shown by the septum at the middle of the figure .
In such a section the calcareous fibres radiate outwards from the axes of the fascicles towards the surface of the septum .
At the left side of the figure a section , CD , has been cut exactly through two oppositely placed strire .
It will be seen here how the axes of the fascicles at first rise upwards , and then bend alternately towards the one surface or the other .
At the same time the arrangement of the fibres is shown round the axee , and their relations to the surface ; these are such ihat the axjs of each fascicle ends in a granulation at the surface .
Fig. 14 .
Similarly cut septa of Mussa , showing its more complex structure ( ref .
, pp. 186-139 ) .
Four successive growth-segments are shown in longitudinal section .
Distinct " fans " of di.erging trabeculm are present in each septal ridge .
In Galaxea ( fig. 13 ) the whole septum is composed of one such " fan " of diverging trabeculm .
The growth-lamellre are also indicated in the figure .
Viewed in a horizontal section , they follow the direction of the surface in a general way , enclosing the axes of the fascicles like a hemispherical series of concentric laminre on either side .
But in tangential section CD , the lamellre are no longer parallel with the surface .
They still surround the fascicles concentrically ; at \#183 ; the same time , since new lamellre are constantly laid down around the upwardly-growing apex , each individual famina makes an angle with the surface ultimately presented .
This is the necessary resnlt of the growth in height of the septum .
A tangential longitudinal section , OF , cut anyhow through the septum , passes obliquely through strim , and gives a more complicated figure .
Each lamina ( aYerage width \#183 ; 003 to \#183 ; 005 millirn .
) is a deposit of calicoblasts , the R 2 124 DR. MARI .
A lVI .
OGILVIE ON THE MICROSCOPIC AND wai , y outline corresponding to originally separctte cells .
I have indicated to the left the scale-like arrangement of such a series of calcifying calicoblasts at the surface .
Naturally , any individual series is not completely caJcified before the next is laid down , but calcification is incomplete to some depth below the surface , and the regular :fibrous laminre are the result of slow crystallization ( cf fig. 8B ) .
One of the most important features in the figure is the relation of individual fascicles to the successive septal segments .
It will be remembered ( p. 110 ) that I defined the growth-segment of a septum , as the pwrt between two growth-curves , representing the additions made to the height of a septum during a single growth-period .
The figure shows that each growth-segment is composed of a horizontal row of individual centres ( axes ) of calcification .
The fibrous deposit around each axis is coalescent with that of neighbouring axes , and the growth-segment is therefore compact .
But each of the individual fibrous groups of the septal segment is only one in a vertical series .
It is necessary to have a name for the vertical series , and no better name can be chosen than trabecub ( = " poutrelle , " E.H. ; cf Introd .
, p. 85 ) , so long as it is clearly understood to refer to the fibrous deposit itself , and not to the " dark " centres of calcification .
The trabecula of Galaxea is composed of a vertical SP-ries of fibrous groups , each of which is deposited around a distinct axis of calcification within a single growth-period .
The axis of calcification bends out of the median plane , usually forking to opposite sides .
The fibrous deposit around the divergent forks of the original axis bulges out on the lateral surface of the septum , and gives rise to a granulation .
SuJ , e1'.fidally , the trabecula of Galaxea is made apparent in the form of a pair of strim nn the opposite lateral surfaces of the septum .
The " serra " in which an opposite pair of strire culminate at the edge of the septum , is the tooth or growing apex of the trabec:ula .
The average distance from apex to apex is \#183 ; 1 to \#183 ; 12 millim. in Galaxea .
The axes of the fibrous groups are lens-shaped , and uniformly disposed in the median plane ; in other words , calcification in Galaxea goes on rather bilaterally than radially .
I shall give the name of trabecular part to the individual fibrous group formed by the fascicles at one centre of calcification .
The fibres of a trabecular part may be differently arranged in different genera , and even in one septum there are sometimes differences .
Galaxea affords an example of perfect unif01\#183 ; mity in the structure of its travecular parts , and , therefore , also in the structiwe of its septalsegment .
Thus one might any day erect an exact model of the delicate calcareous framework of Galaxea , using , as the fundamental unit of structure , a minute bush of :fibres representing the last most highly calcified stage of the calicoblast .
the calicoblasts would have to be cemented together to form a lamella , but while building up the model in purely lamellar stages , in which the polyp built up the original , care would have to be taken to mould each new lamella into the form of ridge and furrow , granulation and SYS'l'E.MA'l.110 STUDY 01 !
' THE MA : DREPORARIA : N 'l'YPBS OP CORALS .
125 groove , which give to the septa and wall of the corallite its beauty of design , and which express outwardly the limits of the fibres belonging to the individual fascicles .
And by the time we had built such a model of Galaxea , and another of Mussa , of Heliastrcea , and of the largest possible number of Madreponiria , we would know\#183 ; something of the classificatory value of their skeletal features .
That the skeleton of the coral , when thoroughly examined , is by no means a small aid towards the establishment of a natural system of claRsification , has been testified by M. EDWARDS and HAIME 's treatment of the Madreporaria .
Certain external appearances have been found to be characteristic of the various Madreporarian families and to be correlated with features of difference in the polyp itself .
The main families recognized by EDWARDS and HAIME half a century ago , are recognized to-rlay after much searching study of the soft parts of the typical genera .
The close and critical examination of external features to which those authors submitted the \lt ; ;:orals , practically exhausted the subject from a macroscopic aspect .
Still , the microscope could put many points in a clearer light , as PRATZ has already shown .
And now the identification , proved above , of REIDER'S calicoblast or lime-forming cell with the ultimate skeletal unit on the surface of all skeletal structures , the demonstration of the part which the fibrous cell-unit plays in building up a definite skeletal lamella , makes that lamella the cell-for-cell equivalent of the ectoderm itself .
Is the ectoderm smooth--so is the skeletal lamella .
Is the ectoderm wrinkled , pitted .
grooved the skeletal lamella is then ridged , granulate , fluted , and the fibres of the whole series of skeletal lamellre , instead of being arranged parallel to one another , are grouped into fascicles .
The fascicles may vary in shape , and may be gathered together in different ways , their arrangement being necessarily illustrative of the original irregularities of the component lameHre , and at the same time therefore illustrative of the form of the body-wall of the living polyp where it is applied to the supporting skeleton .
Bearing this in mind as the fundamental law of Madreporarian skeletal structure , it is clear that the starting-point of any examination of the coral skeleton with a view to its classificatory value , is , -firstly , the variation of the calcified calicoblastic growth-lamella from a smooth plane surface to the most complicated furrowed patterns ; secondly , its shape at all growing edges where the primary double layer of deposit forms either in laminate shape or around separate centres in the hollows of the ectodermal fold .
'11HE SO-CALLED " PRIMARY STREAK .
" Appea1\#183 ; ance of the " dark points " and " dark line , ' ' In the course of the historical accou11t of the literature I slwwed tl1at recent speculation was inclined to regard the " dark line " of septum a11d wall as structurally 126 DR. MARI .
A M. OGILVIE ON THE MrciWSCOPIC .AND independent of the lighter-looking " stereoplasm " or fibrous tissue .
The dark line in the sept.urn has attained in fact the \#183 ; name of primary septum , and where it has a disconnected character , the individual points have been regarded as sections of " primary spines .
' ' This view has lately found expression more especially in the works of FRECH ( Introd .
, pp. 99 , 100 ) , FowLER and BounNE have spoken of the " dark line , " , v hither seen as a continuous line ( ex. Eitphyllia , Flabelfom ) or as disconnected points , in Rection8 ( ex. Mussa ) , as a " centre of calcification , " using " centre " in a general sense .
My results fully justify the term " centre of calcification .
" The appearance of the " dark line " with transmitted light , although generally opaque , is not always so .
It varies very considerably even in perfectly fresh skeletons .
The " points " in a transverse section of Galaxea , for example , appear at , one place homogeneous , and yellowish or dingy-brown in colour , in another place the " point " seems a fairly large , circular area filled with granular , powdery material , and then it is usually dark .
Regarding the " line , " it may be said that very often .
the mere fissure between the two halves of a septum , which forms readily in all sections , often gives the appearance of a median " line " ; but such a fissure usually ends irregularly in smalJer cracks , or bends suddenly out of the median plane across one of the halves .
It \#183 ; is , therefore , easy to distinguish it from a " dark line , " which breaks up into a series of " dark points .
" Several authors have observed that the " dark line " in certain genera is so constituted , but it was long looked upon as a fairly accepted fact that the median " dark line " in other genera , e.g. , Euphyllia , Flabellum , was a really continuous line representing a special median lamina .
HIND ( Introd .
, p. 99 ) , in examining Septastrcea , a genus structurally resembling FlabPlliim , found , instead of a distinct lamina in the median plane of the septum , a separable fenestrated tissue on the inner or median surfaces of both halves of the septum .
Those are the already observed facts on which the theory rests of a " primary septum " composed of " primary spines .
" No one has sought to explain the origin of the said " spines , " or how it came about that they are so remarkably varied in size , colour , composition , whereas the rest of the septum , throughout the whole group of Madreporaria , has a perfectly definite fibrous calcareous structure .
Sirnilct1\#183 ; Dark Bands of the Growth-lamence .
In studying the " dark line " of the median plane of the septnm , an important consideration appeared to me to be the exact similarjty in appearance presented by the dark bands at the inner surface of each of the growth-lamellre composing the fibrous portions of the septum ( cf below , figs. 30A and B ) .
It was not difficult to recognize the cause of the latter appearance .
For each calicoblast which I had observed at the superficial parts of the skeleton showed an unchanged cell-remnant at its proximal skeletal end .
This fact can Le confirmed by the use of stains .
SYSTEMATIC STUDY OF THE M.ADREPOR .
ARTAN TYPES OF COR .
ALS .
127 When , now , the small size of the individual calicoblasts is remembered , it will be readily seen that adjacent proximal ends would give the appearance of a continuous band .
These band8 , like the " points " in the median plane , sometimes appear brownish .
The highest magnifying power never displays any structure in the case of points or bands that appear dark , but only shows a _general black amoi-phous substance , which I can only regard as the carbonized residue of the originally unchanged organic parts of the calicoblasts .
The ultimate appearance of these portions would naturally vary according to the rapidity or slowness of subsequent changes .
And infiltration of water might entirely remove the residual products and introduce foreign material ; very often small crystals of salt are present .
Not infrequently the hypha:i of fungi and alga:i penetrating the skeleton spread more especially in these lamellar bands , the mere fact that the growth-lamella:i are there apposed to their neighbours affording an easier ingress for the hypha:i .
Decalcification proved , however , that the hypha:i ramified anywhere and everywhere , as well as in the lamellar planes .
Further , decalcification brought out very clearly the identity of the medicm dark bands or points with those of the fibrous lam ence .
Usually , as decalcification proceeded , small tears appeared where the dark points of the median plane had been , and also in the lamellar bands ; at the end of the process , the same places Rhowed occasionally undeterminable fragments , organic and inorganic , or were entirely empty .
Thinking the " axes " of the fascicles appeared sometimes virtually solid , I thought of possible chitinous intermixture , and tested them alongside of spicules of Alcyonarians and sclerobases , partly calcareous , partly horny .
The results for the " fascicle axes " of Galaxea were , however , always negative .
I shall frequently have to refer to the so-called " dark line " and " primary septum " in the course of the succeeding chapters .
Meantime , from examination of Galaxea , it may be taken that at the growing edge of the septum , the invaginated septal folds are pushed up into a series of small conical pouches , to which in the skeleton the Aerrated tips of the surface strire correspond .
The calicoblasts apposed to the free serrated tip arrange themselves with their proximal ends practically coalescent , and thus give rise to a ring or\#183 ; circular mass of organic points ; by continued growth , the successive rings of points become the apparently homogeneous , yellowish or yellowish-brown axis in the median plane .
It may also be noted that the fibres immediately surrounding the axis are diffused with finely intermixed yellowish substance , ' ' indicating probably that the calicoblastic cells at the growing margin of the septum are at the time of deposition less completely converted into calcareous fibres than those at some distance below the margm .
* the colour was sometimes so intense as to suggest that it was inherent in the skeletal tissue in these parts .
Brown-coloured barklike " epitheca " has been des0ribed for cert.ain Turbinolids .
'l'h{l question deserves farther investigation , FOWLER and BOURNE have observed the specialized character and large size of the calicoblastic cells at the growing points of the skeleton in several corals , among others Galaxea .
Both these authors regard the calicoblasts as ectodermal cells , which merely secrete calcareous substance without being themselves calcified .
They correlate the large size of these cells at the growing-points with more active deposition of the calcareous substance on the skeleton .
This view is thoroughly confirmed by my observations ; but since I find that the cells themselves become incorporated in the skeleton , it becomes readily understandable that those large , actively secreting cells should be incompletely calcified and retain an organic residue of some significance .
This organic residue is at the actual growing apex .
Whether it becomes afterwards carbonized and dark-looking , or may be otherwise chemically acted upon , even be entirely hollowed out in some cases , is a secondary matter .
The term " spine " is , according to niy 1'esults , entirely inappropriate , and is given in ignorance of the cause of the appearance in question ; on the other hand , the term of " centre " ( resp. " axis " ) of calcification is thoroughly fitting and may be retained .
It is of interest to compare the figures given by VON KocH of the emb1\#183 ; yonic ikeletal deposit ( KocH , Zoe .
cit. , " Development of Astroides , " 'Mitth .
Neapel Stat. , ' 1882 , Plate 20 , fig. 7 ) .
The transverse sections of " spheroids " and " ellipsoids " show calcareous fibres radiating around a central point where no fibres are present .
The centre of the " spha::roid " or " ellipsoid " is the same as the " dark point " in the so-called " primary streak " ( Primarstreif ) of the septum .
This is rendered clear by comparing the figures given by VON KocH , in the same plate , of the growth of the septum vertically upward from the basal disc .
At the same time no one thinks of calling the centre of the embryonic sphreroid a primary sphceroicl l Why then the centres of the mature fibrous groups " primary spines , " and the median area in which they happen to be m03t marked ( because the deposit from opposite flap8 there coalesces ) a " p1\#183 ; i11iary septitm " ?
In all cases we have simply to do with centres and axes ( ideal ) of calcification , around which the calicoblasts are grouped in the living polyp , and from whi'ch therefore similarly oriented fibres ultimately radi:ate when complete calcffication has taken place .
CHAPTER IL-MICROSCOPIC INVESTIGATION OF RECENT AND FOSSIL ASTRJEJD TYPES.-.iliUSSA .
THE SKELETAL p ARTS .
The genus Mussa is a -well-known living representative of the sub-family Astrreinre , fam .
Astrreidai , E.H. , in which the septa have toothed or spined upper and inner edgeR , and have usually strongly-ridged lateral surfaces .
Galaxea has previously ranked as an Eusmilinid genus , i.e. , with smooth septate edges , although the closer study followed out above has shown it to have serrated septa .
It \#183 ; was desirable to compare the fine structure of th~ Galaxea calycinal parts1 more especir-1.ll ; y- the septa1 SYSTEMATIC STUDY 011 ' THE MADRl~PORARIAN 'l'YPES OF CORALS .
129 with those of Astrreinre genera .
And first I selected Mussa , since the anatomy of the polyp had been thoroughly investigated by BOURNE ( Introd .
, p. 101 ) .
s.e. sp. gran .
Fig. 15 .
a.d. c.e. a.,1 .
\#183 ; Upper or growing part of a Miissa septum ; s.e. septal edge ; c.e. costal edge ; sp. spiniform teeth .
gran .
granulations coalescing on a septal ridge ; a.d. area of divergence of septa ] and costal ridges .
gr. g.c. str .
Fig. 16 .
A single spiniform tooth ( 40 times magnified ) of one of the small , fine septa in Mussci , showing upwardly diverging strioo ( str .
) growth-curves ( g.c. ) and variously sized granulations ( gr. ) .
The septum of Mussa runs out into spines of unequal length and size ; these are the 130 DR. MARIA M. OGILVIE ON THE MICROSCOPIC A~D " dents spiniformes " of EDWARDS and HAIME .
In transverse section a single spiniform tooth is elliptical in outline , the middle part being distinctly thicker , and forming on the septal surface the most prominent part of a ridge .
The surface ridges are profusely covered with granulations , of irregular size and distribution ( fig. 15 ) .
Frequently the granulations on the middle of a ri\lt ; lge are larger , and coalesce with one another in any direction .
The shorter thinner septa of the higher cycles in Mussa have less prominent marginal spines and surface ridges .
On each spine granulations are profusely distributed , but are of smaller size , and fine strioo may be distinguished on the lateral surfaces of a spine , resembling the strire on the Galaxea septum ( fig. 16 ) .
That these cannot be observed superficially on the thicker septa is due to the richer deposit of lime which obscures the original lines of structure .
The wall in Miissa is entirely formed by lateral thickening of the septa at the peripheral part of the calyx .
A certain part is also taken by the dissepiments which are placed here at a high angle .
Sections transversely cut through the wall show , however , no independent tangentially-placed ring of calcification in the intervening dissepimental deposit .
This type of wall is characterized as a pseiidotheca , in contra~ distinction to a theca , with distinct thecal centres of calcification , such as has been described for Galaxea .
MICROSCOPIC STRUCTURE .
Identity in the Fine Structitre of the Mussa " Spine " and the Galaxea Sept1.,1 , m. A drawing is given in fig. 17 A of two adjacent spiniform teeth of a thick septum of .ll1ussa , after slight polishing of the surface .
Then the essential lines of structure corresponding to the strire of Galaxea are easily demonstrated .
And again the same evidence is affo1~ded of the periodic growth upwards of the colony .
Successive growth-segments , defined by distinct curves of growth between them , may be seen in each spine .
Each segment is composed of a curving row of segment-parts parallel with the septal edge .
The latter diverge from one another at a greater angle than in the septum of Galaxea , and vary more in their respective sizes and height , but the structure of each individual segment-part is the same in allf namely , a diverging group of fibrous fascicles .
Fig. 17B represents the lower of the two spines in fig. 17 A , sufficiently magnified to show the fibrous fascicles .
Both figures may be compared with similar preparations of the Galaxea septum ( figs. 3 and 10 ) .
The individual parts of the successive segments are attached directly end to end , so as to form continuous trabeculre , diverging from one another in fan-shaped form .
As , moreover , the individual parts of the same growth-segment are rigidly attached side by side , the whole structure is necessarily compact .
The compactness is here associated with very close setting of the original centres SYSTEMATIC STUDY OF THE :MADREPORARIAN TYPES OF CORALS .
131 from which during any given period calcification proceeds .
At the beginning of any particular growth-period the calcification goes on only locally at certain points of the calicoblastic layer of the colony .
It then spreads outwards from those original centres as growth continues , and forms , before the close of the said growth-period , one continuous compact septa ] segment .
In Galaxea the septa .
I segment is in the form of one simple curve ; in Mussa it embraces several curves according to the number of septal teeth , each of which repeats the str\#183 ; ucture of the single curve of Galaxea .
Fig. 17A .
s.r. g.c. Two spiniform teeth in a thick lrfossa septum ( low power ) .
s.r. septal ridge ; g.c. growthcurves marking three successive growthsegments in that septal tooth .
Fig. 17B .
g.l. sr .
tr .
One tooth ( high power ) .
tr .
trabeculm ; sr .
serra ; g.l. growth-lines .
The " area of divergence " of the tmbeculre runs along the middle .
The edge of a spiniform tooth in Mussa gives clear indication of this structure ; it is finely serrated , each serra being the growing apex of a trabecula , just as in Galaxea .
The microscopic ~tructlire gives , therefore , conclusive evidence that each individued spiniform tooth and corresponding ridge in the Mussa septum , is homologous with the complete Galaxea septitm .
The terms " Stria , " " Ridge , " " Spine .
" Hitherto the term ' ' stria " might mean , according to EDWARDS ' and H Arnm 's precedent , either the faint ridges on a Galaxea septal surface , or the broad , stronglydeveloped ridges on the septum of types like Mussa .
Yet a very perceptible difference may be seen by looking merely at the lateral surfaces of the septum in s 2 DR. MARI .
A M. OGILVIE ON THE MICROSCOPIC AND both cases .
In Galaxea , each " stria " is really linear , being almost knife-edged , and having a single granulation now and then upon it .
In Mussa , the ridge has no sharp edge , but is a broad swelling , on which numerous granulations are scattered , apparently irregularly .
The " stria " in Galctxea is the surface appearauca presented by a single vertical series of trabecular parts-is , in fact , uni-trabeculate ; while the ''ridge " of Miissa is an appearance produced by a diverging group of trabeculro-is , in contradistinction , multi-trabeculate .
I shall in future give more than a descriptive value to the terms " stria " and " ridge .
" The term , " : , tria " may be limited to the superficied ctspect on one sidP- of a bilctterally symmetrical trabecula ; the term " ridge " to the superficial aspect on one side of a group of trabeculw .
In Mussct , the component trabeculre of a ridge are all bilaterally symmetrical .
The :fibres aggregate either on both sides of one centre , or of two centres not di verging far from the median plane of the septum .
It is necessary to define those terms clearly , as I wish to avoid the use of the word " spine , " which has been applied by FRECH ( rf .
Introd .
, p. 100 ) to ridges on the surface of Astrreinre septa .
This same word " spine " has been used in several different senses ; - ( 1 .
) For the su1.face ridge , The appearance in question is not a spine at all .
( 2 .
) For the dark points seen in transverse section in the median plane of the Astrroid septum .
Latterly the term " spine " has been qualified by '~ primary " when used in this sense , but has been invariably applied to the axis of a complete vertieal series .
Again , this " primary spine " was regarded as the equivalent of " trabecula , " i.e. , both were entirely limited to the dark central area around which fibrous deposit aggregated .
EDWARDS ' and HAIME 's use of " poutrelle " and PRATz ' use of " 'frabekel " referred to the actual calcareous structures , and that is the sense in which I use trabecula .
The " primary spine " is a term which must inevitably be given up , since its existence as an " isolated pillar " ( FRECH ) has never been proved .
( 3 .
) For the spiniform tooth at the edge of septa in Mussa , lsophyllia , \amp ; c. ( 4 .
) Fo1\#183 ; the short , spinate , calcareous growths ( septal spines ) projecting horizontally into the calyx in the place of septa in Madrepora , as well as for similar growths in Palreozoic genera , whatever their position in the calyx .
These are the structures for which originally the term " spine " was used , and for which it must be retained .
I shall throughout the present paper limit the application of the term " septal spine " to this original use of it .
It -w1ll be found to be the morphological equivalent of a single trabecular part in Mussa , Galaxea , \amp ; c. ( 5 .
) For , \#183 ; spinate outgrowths from the theca of Turbinoliclw .
The ornamentations on the Turbinoloid theca take sometimes one form , sometimes another ; SYSTEMATIC S'11UDY OF THE MADREPORARIAN TYPES OF CORALS .
133 when they are spinate , the term " spine " is justly given .
At the same time , the context makes it usually clear that these are thecal outgrowths and have no connection with septal spines .
ProJection Figure , representing the Distribut\#163 ; on of Trabeculce in various Septet .
In order to bring out clearly the homology between the septum of Galaxea and the ridges of the M1issa septum , I give a comparative diagram of their component Fig. 18 .
a Gala:l'ea ( whole sPptum ) .
b Miissa ( two ridges of a septum ) .
Heliastrcea ( septum ) .
cl~G Goniastrcea ( costa , septum , and palus ) .
Scale diagram , showing the distribution of " centres of calcification " seen in transverse sections of various septa } types .
trabeculre ( fig. 18 ) , as represented by the distribution of the centres of calcification in a transverse section of the septum .
The dimensions of the figures are drawn to true scale .
Owing to their large size , only two of the ridges of Mussa are introduced ; some six or seven ridges are present in the complete septum .
Each point in the figures represents the ideal axis of a trabecula .
The projection figure of the .
Galctxea septum ( fig. 18A ) may be described as 0 !
1e symmetrical series of points , representing the symmetrically diverging group of trnbeculre .
Tbat of the Mussa septum ( fig. 18B ) , if completely given , would be a symmetrical series of points , as many times repeated as there are ridges on the septal surfaces .
And as the trabeculre in the middle of a group are more closely packed together than those on either side , the whole projection figure reminds one ?
f a graphic representation of sound waves .
After realizing the projection figure of Mussa , we may transfer it with but little variation to many other genera belonging to the group of Astrreinre .
In .l-Ieliastrceo , ( fig. 18c ) , instead of uniform bilaterality in the fine structure of the trabeculre , those in the thickest part are especially marked out from the others , inasmuch as several cent.res of calcification are arranged round an ideal axis , radially and at the same time 134 DR. MARIA M. OGILVIE ON THE 1vI1CROSCOPIC AND Fig.19 .
c.c. c.c. c.c. diss .
gr. gr. diss .
g.l. 'l'ransverse section through a thick spiniform tooth of the Miissa septum ( low and high powel ' ) ; c. c. centre of calcification ; gr. granulation ; g.l. growth-lines , signifying successive growth-lamelhe ( p. 125 ) ; diss .
dissepiment .
Fig. 20A .
gr. c.c. Transverse section through the thin intermediate part of a M1issa septum between two spiniform teeth .
Several tiers of granulations are met on either side of the median plane .
Fig. 20B .
gr. Diagrammatic drawing of a thick spiniform-toothed ridge and of the thin 1:nterrnediate part ( int .
) with closely placed , E1mall granulations .
RYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
135 to some extent spirally .
The same is true , in a less degree , of the closely packed centres of calcification in the Goniastrcea septum .
T1Ymsve1\#183 ; se Sections through the Septal Spiniform Teeth .
As a glance at the septum of Mussa ( fig. 15 ) shows , it is jmpossible to make a transverse section through a complete septum , cutting all the spiniform teeth perpendicularly to their axes of length , since such a section would represent a curve nearly parallel to the margin of the septum .
This fact must be remembered in the consideration of the accompanying transverse sections .
The thick part of one spiniform tooth of a. septum has been selected for fig. 19 , as the axes of the fascicle pairs in the median plane are there met almost perpendicularly , and thus show the regularity of their distance from one another the cross-diameter of a fascicle is about \#183 ; OS millim. The continuation of this transverse section to the interdental part on both sides is not included in the figure ; it naturally shows the fascicles more and more obliquely cut .
The lateral parts in fig. 19 , marked " gr " represent the emergent ends of fibres belonging to surface granulations .
The granulations are particularly large on the thickest part of the ridge ; the section has cut obliquely across the septum .
On the other hand the granulations are much smaller at the thin ends of the individual teeth or ridges , and a section which is fairly thick may cut through several upgrowing rows of granulations towards the septal edge .
Such a section is given in fig. 20A , and to elucidate it , a drawing is made ( slightly diagrammatic ) of the surface of the ridge at both its thickest and thinnest parts ( fig. 20B ) .
Fascicles of fibres must be imagined within each granulation shown in the surface drawing , their axes necessarily placed obliquely upwards in the direction of growth .
The thin part with crowded small granulations is that which is represented in section in fig. 20 .
The surface relations are very easy of demonstration by studying the fine edge of a septum with strong lens and then gradually polishing it and watching the stages .
My chief reason for insisting upon the relation of the granulations is because the appearances in fig. 19 might seem at first sight to suggest a lateral deposit of " stereoplasmic " nature differing from a median " primary " deposit .
Such sections are highly characteristic of the granulate septa in Astrreinre , and more especially for thick septa of calyces belonging to a full-grown corallum .
The greater lengthening and thickening of the granulations in the latter case is undoubtedly the result of the continued growth in thickness of the skeletal parts after their full growth in h~ight has been attained .
Microscopically examined , the extra thickness is produced by continued lengthening of fibres round fascicular axes .
If this increase goes on long enough , the granulations next one another may often coalesce .
When the orientation of the fibres is the same throughout growth , from the _laying down of the first laminre at the edge of the septal invagination to the laying down of the last layers some distance below the edge of the septum , it will be seen that ( in the Mussa septum at least ) the terms of " primary septum " and " stereoplasm " cannot well be applied in the sense of two distinct structures .
" Secondary\#183 ; thickening " is , on the other hand , suitable as a general term implying the continued growth in thickness characteristic of the first and second cycles of septa , * and in respectively less degree of other cycles .
But it may not be taken as meaning any difference in the microscopic calcareous laminre of- the septum .
Fig. 21 is a transverse section through the thinnest septum present in a complete Fig. 21 .
c.c. Transverse section of a very fine septum \#183 ; 05 millim. thick ( cf. fig. 19 ) .
calycinal section .
Its interest is in the fact that , although its thickness is only \#183 ; ~5 \#183 ; millim. , the distance from one centre of calcification ( c.c. ) to the next is \#183 ; OS millim. , exactly the same as we have already observed in the thickest septa .
We may deduce , therefore , that in Mussa there is a minimum distance of \#183 ; OS millim. between one trabeoular axis and the next , irrespective of the pa1\#183 ; ticular " order " to which a septum belongs .
Comparative Results derived from the Minute Structure of Skeletal Parts in Galax ea or Mussa .
I have not thought it necessary to repeat the details of microscopic structure in Mussa , in so far as they coincide with those of Galaxea ( cf. fig. 14 ) .
In both , the growth-lamella is formed by the coalescence of calcified calicoblasts , about \#183 ; 003 millim. in height .
The lamella is composed , therefore , of fine calcareous fibres set at right angles to its waved surface .
Successive Iamellre are smoothly extended at the base of the polyp and form there the smooth dissepiments .
In considering the septum , which differs in the two genera , it will be well to bring it into relation with the ectodermal layer of the polyp ( ref : figs. 2 and 22 ) .
The basal body-wall of the Astrreid coral polyp , which rests on the dissepiments , is pocketed or invaginated upwards into the visceral cavity ; an\lt ; l the invaginations take place all round the polyp , so as to form a complete ring of radial folds .
Each * The septa of these cycles are usually called primary and secondary septa .
This use of the term " primary septum " refers to its better development and its early appearance in the embryo .
SYSTEMATIC STUDY OF 'l'HE MADREPORARTAN TYPES OJ !
' CORALS .
137 flap of the fold is drawn at a number of places inwards to the visceral cavity of the polyp .
Corresponding to the irregularities on the endodermal or visceral surface of the body-wall , small pits are present on the ectodermal , skeletal-producing surface .
Subsequently , the skeletal layer of the septum is an exact cast of tlw form of the ectodermal flap ; that i~ , the original form of every pit ( of various shnpes and sizes at different parts of the flap ) is represented by that of the corresponding eminence on the skeletal la.yer .
The component calcified calicoblasts of the layer have their fibres set at right angles to the sides of the pit , and the eminence of the growthlamella assumes a hemispherical , conical , or any other form , according to the shape of the ectodermal pit ; in Mussa , the pits are mostly cylindro-conical .
In all cases the fibres radiate around what was formerly the axis of the pit .
By continued deposition of lamellre a fascicle of fibres is determined , whose axis coincides with this axis .
Fig. 22 .
MUSSA.-Non-camenchymatous type ; o. = oral body-wall ; a.= aboral body-wall .
e. = edge-zone ; s. = septum ; c. = costa .
is.l .
= interseptal loculus ; ic.l .
= intercostal loculus .
ps.th .
= pseudotheca ; ep .
= epithcca ; d. = dissepiments .
( cf. also p. 156 , and figs. 2 , 35 , 36 , 56 ) , The ectodermal pits at any one growth-period are arranged in ro , vs in Galaxea and in .Mussa , and as the polyp grows upward , the new pits are exactly above the foregoing .
Hence , in the skeletal surfaces of the septum , we find a continuous corrugation or " stria , " on which the granulations are placed .
The structure on any one stria is the same in Mussa and Galaxea .
But there is an important difference in the distribution of strire on the septa of the two genera , a difference which results from the particular form of invagination of the body-wall into radial folds .
This difference will be made clear by a comparison of fig. 13 with fig. 14 .
The latter snows the :MDCCCXCVI.-B .
\#183 ; rr l :J 3 D'J , .
:'11.\lUA .
M. OGILVrn O::-l ' THF ; MICROSCOPIC AND septum of l , ; fussct cut in various directions , tra , nsversely and longitudinally , and shows the relation of the surface sculpturing to the internal microscopic structure .
In Galaxea one long radial fold is formed , each flap of which is pleated , and the edge delicately crimped out ; in each crimp of the edge a serra of a septal stria develops .
In Miissct , the radial fold is notched out into a series of pockets ; within euch pocket a " spiniform tooth " develops , n.nd the flaps on either side of the pocket are goffered anJ crimped in a system exactly corresponding to that marking the single long pocket of the Galaxea septum , .
The notching out of the septal fold of llfussli allows it to have a greater areal extension , the pockets of the body-wall springing awn.y from one another all along the low curve of the broad oval disc .
This gives rise to a certain divergent arrangement of the pockets on one side and the other of the peripheral area of the calyx , the apices of the pockets being directed at right angles to the curve of the oral disc .
At the same time , calcification goes on independently within each pocket , and produces there a fan-shaped group of fascicles in each growth-period .
The " pockets " of the septal fold of l , ; Jussa are sufficiently deep to allo-N at least some three or four successive growth-segments to be within them .
During each growth-period a varying number of growth-lamellre are deposited , more in thick septa than in the higher septal cycles ; and , since further deposition goes on during several successive periods in Mussa , the septa of the main cycles attain in this genus unusual strength and thickness .
A very important point in .ilf ussa , is the effect produced by the stretching away of neighbouring pockets from one another .
The ridges corresponding to the pockets appear afterwards definitely arranged on the radial strncture .
They pass invmrds and upwards from the periphery towards the centre of the calyx , and outwards and upwards from the periphery towards the costal edge .
The periphery of the calyx is indeed marked out as an " area of cliver\#183 ; gence " for the surface ridges of serJturn and cos ta .
It is at this area of divergence that new pockets are invaginated , and therei : ore new ridges intercepted , during the course of upward growth ; and the ridged radial strncture is always highest at that part , i.e. , most exsert ( antea , fig. 15 ) .
So fa.r as I have yet gone , I may distinguish two types of the Madreporari , m : : , eptum :( 1 .
) Galaxea .
The radial invaginations of the aboral body-wall , in which radial skeletal structures are formed , have fluted flaps , and lay down a septum with striated , granulate surfaces and a serrated m. .
a rgin .
( 2 , ) Mussa .
The radial invaginations of the basal body-wall are drawn qut into long free pockets , broad at the base and narrowing towards the apex .
Each pocket has fluted flaps , and lays down a corrugated , granulate ridge , with finely-serrated margin .
The whole radial structure is composed of a greater or less number of snch ridges , which form adjacent ridges directed upwards and inwards in the " septal " portion of a radial structure , and SYSTJ::MATIC STUDY OF TH:1\#163 ; :MADREPORARTAN TYPES OF CORALS .
139 upwards and outwards in the " costal " portion .
Between septal and costal portions there is , therefore , in JJ1 ussa an " area of divergence " of 1\#183 ; iclge~ not represented in Galaxea .
This coniplexity of structitre is characteristic : of " Astr(Rince " genera .
, as distinguished from a large number of the genera bithei\#183 ; to classified under the " Eitsmilinw .
" E.H. The structure of both these septal types is shown in various sect.ions , diagram~ rnatically treated , in figs. 13 and 14 , above .
In both .Llfussa and Galaxea , I have been able to prove that there is no reality in the terms 1wimary spines , _1n\#183 ; i111ary trabeculCP , p1'imcl1'y septum , .
The anatomical researches of FOWLER and BouRNE have proved that the ectodermal calicoblasts are larger and often differently shaped above the apical growing-parts of the skeleton .
FowLER has proved this in Lophohelia and Gala , xea , BouRNE in Mussci and R1tphyZlia .
At the same time , no such observation , vas made by BouRNE on Fungia , and it , cannot by any means be generalized .
I have indicated above the influence which the size of the cells and activity of the deposit may have in producing the appearance of the " dark line"= " primary septum " or " dark points"= " primary trabecuhe " at the median area of the septum ( antea , p. 126 ) .
According to the amount of unchanged organic cell material , , ve may expect a more marked difference in the appeRrance of the individual growth-lamellm .
The growth-lamellm deposited at the growing-edge of the septum nre those which are afterwards observed in the median plane .
They are at the same time those which have usually more unchanged organic material an\lt ; l a less clear crystalline deposit ; those which in fossil material are found in the most highly disintegrated condition and have the most varied appearances , called forth by secondary change .
No definite limit can be drawn between growth-lamellre nearer the centre and nearer the periphery .
Secondary change begins apparently at the centre and e.xten\lt ; ls outwards , sometimes affecting the whole series of fibrous growth-lamellre , which originally composed the septum .
HELTASTR .
, EA .
The Corallitm .
The corallum of Mussa is usually described as brancling , that of Heliastrma as massive .
And the difference in the form of corallurn is regarded as correlated with a difference in the mode of polypal increase-fissiparity in the case of 1Vfossa , gemmation in Heliastrwa .
The opinions_ with regard to the budding of coral polyps are still divided ; up to the present it has not been possible to distinguish in all cases satisfactorily between so-called budding , bud-fission and fission .
The diagrammatic figure3 ( 2 , 22 , 35 , 3G , 56 ) are given to illustrate various methods of calycinal union in ca.menchymatous and in non-crenenchymatous colonies .
In Gcdaxea , the calyces are sa , id to be " surrounded " by camenchyme ( ref.antea , p.107 ) , in-Ieliastrcea they are said to be " united by costre , " in Mussa they are usually described as " free .
" \#183 ; r 2 Yet , in all three genera , the soft parts of adjacent polyps n.re continuous .
In .ZUussa , it is true , groups of adjacent polyps become independent of one another in the corallum correlated with the branching growth .
The facts show three different forms of the extracalycinal skeletal parts .
In Galaxea , the costre are narrow , and the intermediate space between calyces is entirely occupied by dissepiments .
In J-Ieliastrma , the costre are so wide as to come usually into contact , dissepiments being developed between them .
In Mussa , there are no dissepiments between the costre .
The mode of incre1se is essentially the same in all three genera , so far as the skeletal structures afford evidence .
Side branches are in all cases in open communication at their origin with the oral surface of a main branch .
General Similarity of the E'J.ternal Slceletal Features in Mussa and Heliastrroa .
A hand-lens examination of the radial and basal skeletal structures of Heliastrma ( cf fig. 23 ) is sufficient to demonstrate that they are built up on the same general plan as those of Mussa .
:Fig .
2'.3 .
e.T .cul .
cal .
s. a.cl .
c. diss .
ps .
th .
--1 !
, ~0 .
ps.col .
d. ps.th .
d. Skeletal parts of Heliastrroa .
cal .
, calycinal parts ; ea.\#183 ; .cal .
, extra calycinal parts ; s. , sept-nm ; a.d. area of divergence of septal and costal ridges ; c. , costa ; r. , ridge ; d. , dissepiment ; ps.th .
, pseudotheca ; ps.col .
, pseudocolumella .
The radial structures , compnsmg septum and costa , show in Heliastrma as in Mussa , prominent teeth at their edges , and ridges on their lateral surfaces .
The ridges diverge from a peripberal " area of divergence , " passing in the sam:e radial plane inwards and outwards .
This " area of divergence " is the thickest part of the radial structure .
At some distance below the uppermost edge of septum and costa , the tangential thickening at this arec:1 , is almost sufficient to bring the neighbouring radial structures into contact .
The dissepiments at the same area are strongly developed and placed more or less vertically , afforcling a means of coc:1 , lescence of the expand.ed " arens of divergence .
" Sometimes independent " centres of calcification " are present like those in the wall of Galaxca , but less regular in position .
Inasmuch as the wall is not composed of clearly-de6ne\lt ; l tangential structures , bnt may be SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
141 variously developed , it approaches the " pseudothecal " type of Milssa .
There are , however , all transitional stages in different genera bet ween a " pseudotheca " arising from lateral septal thickening and a " eutheca , " such as that observed in Galaxea .
In Heliastrcea , us in Mussa , growth-curves may be distinguished on the septal surfaces , marking the limits between successive septal growth-segments .
A " diagonal " pattern is traceable on the lateral surfaces of the septum in Heliastr'wa , owing to uniformity in the breadth of the septal ridges and the height of the septal segments .
In other words , the additional height attained by the ridge during a single growthperiod is from \#183 ; 35 millim. to \#183 ; 45 millim. , and that is also its average breadth .
Only one large granulation is marked out on the ridge during a growth-period ; hence each vertical series of large granulations on the surface corresponds to successive stages of growth of the ridge .
These stages are again indicated by the successive platforms of dissepiments in the interseptal loculi , many of the dissepiments resting on the granulations of adja.cent septal surfaces .
Although the external appearance of the individual ridges in Heliasv~ea and Mussa differs considerably , their essential microscopic structure is very similar .
In both a number of trabeculre are present .
But the fibres are arranged according to two different systems in }Ieliastrcea ( cf fig. 18c ) .
First , there are small-sized traheculre , " tr .
, " in which the fibres practically arrange themselves in two oppositely- placed fascicles during a single growth-period .
This is the system already observed in ~Mussa .
Second , there are larger-sized trabeculre , " Tr. , " in the middle of each ridge , corresponding in position to .
the larger gran~lations .
In these trabeculre the " centre of deposit " is in reality a , group of centres , very closely phced .
Fibrous fascicles proceed from two , three , or more centres , and pass outward to the opposite lateral surfaces .
Each large granulation represents , as a rule , co3.lescent fascicles belonging to this group .
The appearance in transverse section of one of these larger tra1eculre , is that of radially-arranged fibres instead of the almost wholly two-sided arrangement of the fibres in the smaller trabecuhe .
I propose to call the large-sized trabeculre , with a group of centres , compound trabeculce , and the smaller-sized trabeculre simple t1\#183 ; abecillce .
Tr\#183 ; ansverse Sections .
The microscopic structure of the trabecnlre composing the septum in Heliastrcea , is represented in the sections .
Figs. 24A and 24B are horizontal transverse sections of the septum .
In fig. 24A it will be seen that in some parts the dark-looking centres of calcification are closely placed in rows ; occasionally a pair of axes are cut corresponding to the portions of fascicles bending outwards to the surface .
At intervals in the section thicker trabeculre are present , in which the growth-lamellre become strongly concentric and bend in out-springing curves around a common centre .
In the centre there is not one dark point nor two , as we have observed 14~ DR. MARIA M. OGILVIE ON THE MICROSCOPIC AND in sections of Galaxea and il1ussa , but three or four distinct centres of calcification arn present .
The concentric growth-lamellre curve in and out around them in the middle part , although the later curves no longer follow the individual centres but enclose all .
The diameter of the smaller , " simple " trabeculre , is from \#183 ; 04 to \#183 ; 05 millim. , Fig. 24A .
c.t , \#183 ; .
c. Tr. ps.th .
c. 'l'r .
d. Transver ; ; ; e section of two adjoining sept !
t of Heliastrr.ea , one of tho septa hwing split along the rnediau plane ( containing the so-called " dark line " ) of the septum .
c.'l'r .
Centres of calcification in compound trabecnloo .
c.tr .
, , , , , , simple trabeculoo .
ps.th .
Pseudothecal , and d. , dissepimental deposit is shown to be without individual centres of calcification .
Fig\#183 ; .
24B .
tr .
Tr. Thick central po1\#183 ; tion of !
t septal ridge formed by a " compound " trabecula ( 'l'r .
) ; and the thi11ner lateral wings of the same ridge passing into corresponding furrows-only " simple " trabeculre ( tr .
) are present in these .
( Magnified 70 times .
) that of the larger trabeculre varies according to their position , those at the peripheral part of the septum being largest .
Their average diameter is from \#183 ; 09 to \#183 ; 12 millim. Fig. 24B shows four small fascicles in the centre of a large trabecu]a , and the fibres radiate upwards and outwards .
It may be said that radial symmetry marks SYSTEMA'l'IC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
143 the system of fascicles in the thick , central portion of any septal ridge , while bilateral symmetry marks the thinner , lateral wings .
Longititdinal Section ( Radial ) .
Fig. 25 shows a longitudinal section parallel \Yith the surface and passing through sp. Fig. 25 .
sp. H2 R1 fr .
tr .
Longitudinal section parallel with the septal plane ; R1 septal ridge met along the median plane ; R2 septal ridge met very obliquely ; tr .
trabeculre ; sp. spiniform teeth .
two septal ridges .
The thick part of one ridge , " R1 , " and the thinner lateral area next it nre met by the section nearly in the median plane ; in the other ridge , " R2 , " the section passes obliquely through the septum towards the outer surface .
The arrangement of the fibres in the different parts comes better out in longitudinal than in radial sections .
The highly oblique position of the fibres in the central " primary " deposit of the compound trabecula , and their less oblique position in the subsequent thickening of the same trabecula is readily perceived .
In the other trabeculre of " Ri " the fibres are also seen to be directed obliquely upwards , but in R2 the figure shows only dark points from which fibres radiate .
As R2 is cut near the surface , and this appearance of the fibres proves that they are cut almost transversely to the ( ideal ) axis of the fascicle , it may be deduced that the fascicles bend sharply from an upward direction perpendicularly outwards to the surface .
The general conclusions for the Heliastr(13a septum which may be drawn from these sections are:( 1 .
) The primary deposit laid down at the edge , at the beginning of an active period of growth , is limited to certain areas at some distance from one another ; this primary deposit is represented by the central part of the main trabeculre .
( 2 .
) As the growth proceeds , deposit takes place more uniformly and fills up the intervening spaces .
( 3 .
) The fibres of the primary deposit are laid down within small conical poucbes of the body-wall , and are obliquely and radially placed with regard to the centre of calcification ; those which are afterwards deposited are laid down from the side-fla , ps of the septal invagination , and have , in consequence , a tend0ncy to be less and less oblique .
The fibrous deposit in the small trabeculre is uniformly oriented throughout ; in the larger trabeculre , however , the first-formed fibres are fairly distinct from the fibrous deposit made towards the close of the same growth-period and during succeeding periods .
In the sections':- of Heliastrcea , yellowish matter was often seen penetrating the fibres , and especially in the laminre near the median plane of the septum .
It often made the fibres seem brown , instead of clear and crystalline as they appeared in other parts .
The organic nature of this yellowish matter was capable of chemical proof by the action of re-agents .
I regarded it as a less altered condition of organic cellmaterial than that which gave rise to the dark " axes " in the Galaxea fascicles .
The Costl .
JJ .
An interesting feature is frequently noticeable in the costate portion of the Heliast1 \#183 ; cea septum ; viz. , the entire absence of small trabeculre .
The ridges often run out Fig. 26 .
Transverse section showing the radial trabeculre ( comp .
) in the intercalycinal portions of the Heliastrcea skeleton ; c.Tr .
costal trabeculre radial in structure ; ps.th .
pseudotheca ; s. septum .
nto longish trabeculre , in no way connected with one another .
The transverse section ( fig. 26 ) shows how these individual trabeculate ends produce a figure of radial sym* Special care was taken in mounting , neither canada balsam nor shellac being used .
Sections wero viewed under water or pure glycerine .
SYS'l~EMA .
TIC STUDY OF THE :MA .
DR}ijPORAR1A.N TYPES OF CORALS .
14-.5 metry corresponding to that which I have described above for the larger trabeculre in the calycinal part of the septo-costa .
But I could not observe any difference in the orientation of the fibres , apparently the costal deposit begins to be laid down somewhat later than that in the exsert septal part , and there is not a primary as distinct from a secondary direction of the fibrous deposit .
Longitudinal Section ( Tangential ) .
Generally speaking , the diameter of a fascicle in a large trabecula of H eliastrcea is about the same size as that of Mussa .
The fascicles emerge rapidly on the thick part of the ridge , and it is their coalescence which gives rise to the large granulation corresponding to an individual growth-segment in the ridge .
Secondary thickening may continue to accumulate on the granulation until the latter comes into contact , in the interseptal loculus , with a similarly thickened granulation projecting from the adjacent face of the next septum .
By this contact a " pseudosynapticula " is formed such as PRATZ has described in the group of Pseudoastrreinre .
It has been thought that " pseudosynapticuhe " were not present in Astrreids , but my observations show them to be fairly common .
Fig. 27 .
]'S .
l:'yn , d. ljl ' .
tl .
s. p ; ; . .
, in .
s. s. Longitudinal section ( low power ) through three adjacent septa , showing " pseudosynapticu ] are " union between them , by means of the continued growth of granulations ; d. , dissepiment ; gr. , granulations ; ps .
syn .
, psendosynapticula ; s. , septum .
A tangential section is given which has passed through three neighbouring septa ( fig. 27 ) , and shows how the dissepiments often weld granulations together .
The same section shows occasional pseudosynapticubr union .
The figure may be directly compared with similarly-cut sections figured by PRATZ in illustr~ition of Cyclolites ( ' Palreontographica , ' loc. cit. , 1882 , Plate 14 , fig. 7 ) , and of l'hamnastrma ( loc. cit. , Plate 14 , fig. 12 ) .
GoNIASTR.:EA .
The Chief Strnctu1\#183 ; al Difference from Mussa and Heliastrrea is the almost complete absence of an Edge-Zone ( Randplatte ) .
I selected Goniastrcea in order to study more especially the relations of the pali to the septa .
The external features of the calyces are Axhibited in fig. 28 .
Goniastrcea has Fig. 28 .
a.cZ .
pa .
.s .
: ; 5--::.~/ .
~~ Skeletal parts of Goniastrcea .
s , ; large septum in one calyx ; s ' .
, small septum in the adjoining calyx ; pa. , pali ; ps.th .
, pseudotheca ; a.d. , area of divergence .
faint ridges on the septa resembling those of Heliastrma , but so small and flat as to be scarcely visible were it not for the presence of distinct pointed granulations upon them .
The walls of adjoining calyces are in contact throughout their entire length , the ri'dges of the radial structures passing from the peripheral " area of divergence " into two septal wings instead of a septal and a costal .
At the same time , it is noteworthy that in any one radial structure there is always a longer and a shorter wing .
In other words , the long septa in any particular calyx corr~spond to the short septa in adjoining calyces , and the short septa in the same calyx correspond to the long septa in adjoining calyces .
The short septa are therefore practically homologous with the costal wings of septa in genera like Mussa and Heliastrma .
This is of interest in the study of the edge-zone ( Raudplatte ) in Astrreids .
The costre are formed by the polypal edge-zone in Mussa and Heliastrma , and it would seem at first sight that genera like Goniastrma and Prionastrma were _devoid of edge-zones .
KLuNzrnGER however notes that very often in these and allied genera an irregular appearance of intermediate tissue occurs in slight furrows between the walls .
In transverse sections of the skeleton also , angular spaces occur between the \#183 ; walls of calyces more or less filled by calcareous deposits .
The latter is laid down SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
147 around several centres of calcification without any definite position relative to calycinal parts .
The edge-zone may at the same time be regarded as much curtailed .
In correlation with this fact we find that increase of the colony is said to take place by means of " central " and " par1'.etal " intracalycinal budding .
As KLUNZlNGER says , central intracalycinal budding is not distinguishable from the fission ( " Theilung " ) in Favia .
A study of various skeletons shows that all these forms of increase are identical with what has been called " septal budding " in fossil Palreozoic forms .
In Goniastrwa , " budding " is indicated in a calyx in two ways :-(1 ) Two pali in front of septa extending from opposite walls grow more rapidly upwards than the rest of the pali and are thic~ned so as to coalesce with the septa and form a diametric line of fission across the calyx-this corresponds to " central " intracalycinal budding ; ( 2 ) Two pali in front of septa extending from adjacent walls of the calyx form with their septa a curving line of fission ( sometimes the septa themselves coalesce without the ai\lt ; l of pali)-this corresponds to " parietal " intracalycinal budding .
Both are merely varieties of " endothecal budding .
" The extreme form of intracalycinal budding is shown by meandroid colonies .
On the other hand polyps with a well-developed edge-zone send out their buds in the edge-zone , the buds being then said to be " extracalycinal .
" An extracalycinal bud ma.y , on developing , remain in communication with the mother-polyp by means of its edge-zone ( ex. Heliastrwa ) ; or the mother-polyp and the bud may gradually become disconnected and the bud grow up as an independent branch ( ex. Caletrnophyllict ) .
Mussa is a genus which produces intracalycinal and extracalycinal buds almost with equal facility , the extracalycinal buds developing as branches , the intracalycinal buds remaining as a " fissiparous " series at the head of each branch .
Both modes of budding may lay the same claim to ancientness , since both occur in Palreozoic colonies , but the intracalycinal buds became in primitive forms always distinct from the mother-polyp .
The earliest cases of fissiparous , meandroid colonies are , so far as I know , of Triassic age , being the " Latimreandroid " variety of the genus Isastrcm .
MICROSCOPIC STRUCTURE .
.Transverse Sections of Septa .
Fig. 29A is a transverse section made through one-half of a calyx .
The septa are seen to pass through the thickened septal wall into the adjacent calyces .
In this figure elliptical centres of calcification are seen in the median plane 01 the septa at regular distances , and corresponding in position to the ridges on the opposite septal surfaces .
Although this is a correct conception of the structural form in the Goniastrwa septum , it has been rendered rather diagrammatic in the drawing , which also omits to introduce the fine growth-bmelhe .
u 2 The transverse sections seldom show such a definite and regular septal structure ; I have drawn in fig. 29B a transverse septal section with fairly typical structural features .
The actual nnmber of growth-lamellre present in one-half of a primary septum is as much as from 40 to 50 ; in the drawing far fewer are given , for the sake of charness .
Fig. 29A .
s. ps.th .
Transverse section ( 20 times magnified ) .
s. , septa ; ps.th .
, pseudotheca .
Fig. 29a .
~~ g~ Transverse section ( 70 times magnified ) of a septum of Goniastrcea , showing the curving lines of the growth-lamellre crossed by the calcareous fibres .
c.c. , centres of calcification ; gr. , granulation .
The centres of calcification are placed very closely , and in the thin part of the septum always singly .
Where the septum thickens in the wall area , two , three , or four centres are often seen massed together , and forming a common group-centre for the growth-lamellre .
The first feature which occurs to one in looking at fig. 29B is the absence of any large trabecuhe in the thin part of the septnm , like those in Heliast1Y1Ja , corresponding t~ ) the prominent parts of the ridges .
The trabecnlre in Gon , iastrma , so far as they can be clearly distingui8hed , have a diameter of \#183 ; 03 millim. to \#183 ; 04 millim. , the :-nnaUest I have yet had occasion to describe .
At the same time the surface SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
149 ridge has a diameter of about \#183 ; 15 millim. Sometimes the centres of calcification appear in transverse section even closer together than \#183 ; 03 millirn .
, and then they may be looked upon as almost coalescent .
Short rows of such closely-placed centres produce the elliptical appearance represented in :fig- .
29A .
The application of higher magnifying power , however , shows that the elliptical centre is in reality a combination of centres .
It follows that the ridges on opposite surfaces of the septum in Goniastrcea are structurally composed of several very small , uniformly-developed trabeculrn ( cf. fig. 18D ) .
Correlated with the small size of the trabeculrn is the fact that the granulations on the ridges seldom correspond to a fascicle of fibres emerging from a single trabecula ; they usually represent the coalesced fascicles passing out from the ridge-series of trabeculrn .
Examples of small and large granulations can be studied in the accompanying sections .
Transverse sections of the septal pali offer nothing different from the septa .
One or two ridges may be included in a single palus , and the structural appearance is the same as in the septal ridges .
Tr:anwerse Section of Wctll ( = Pseudotheca ) .
The thick parts of the septa coalesce laterally to form the wall .
The trabeculrn in the mural part of a septum are much larger than those in the thinner calycinal portion .
Fig. 30A shows three septa in the mural area , the middle septum in the figure being short .
It will be observed that the growth-lamellrn ( cf. fig. 30:B ) arrange themselves in concentric form around the axes of the trabecuhe , and the fibres radiate out freely in all directions .
One might sa.y radial symmetry of the fibres distinguished the large trabeculrn of the mural area , while bilateral symmetry was characteristic of the minute trabeculrn in the rest of the septum .
A \#183 ; well-developed dissepiment rests across the inner point of the short septum in fig. 30A .
It is frequently seen that the close application of the dissepiments to short septa helps to make the wall stronger and thicker .
Tangential Longitudinal Section .
Fig. 31 presents a tangential vertical section through one septum , which has cut three adjacent ridges .
The ridges lettered a and c happen to have been cut where granulations are present on the opposite surfaces ; the ridge b , on the other hand , is cut at a smooth part of its course .
The fibres within the granulations necessarily arrange themselves in the form of a large fascicle , since they are laid down originally at right angles to the planes of successive growth-lamellrn .
The fascicle is all the more marked in Goniastrroa , since the shape assumed by the granulation is extremely conical .
Later1 in stlidying Siclerastrcea , granulations will be observed which have .
a flattish Fig. 80A .
S~ .
g.l. d. Transverse section of \#183 ; the wall area between two adjoining .
calyces of GoniastrC1Ja ; large , compound trabeculoo are present in the septa ( s1s2s3 ) , and thickened difsepiments ( d. ) add farther stability .
Fig. 31 .
h. g.r. Fig. :~On .
c.c. g.l. A small portion of fig. 30A seen under very high microscopic power .
c.c. , centre of calcification ; g .l .
, growth lamcllre , edged by dark bands ( c'f .
p. 125 ) .
a. b. c. fr.ax .
S eptum cut lengthways .
The septal ridge " c " shows an almost transverse section of the ridge trabecula with granulations on opposite surfaces .
tr.ax .
, trabecular centre or axis ; h , hypha of a fungal growth .
SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES 0.1 !
- , CORALS .
151 or gently rounded shape , and in which the fibres are al most parallel with one another .
It is in the case of such blunt , broad granulations that a tendency arises for the granulations of adjacent ridges to unite and form " pseudo-synapticular " bars across the septal surface .
In Goniastr(J3a , however , although the base of the granulations may practically span the whole diameter of a surface ridge , the granulation itself tapers off and remains always isolated from its neighbours .
A few fungal hyph::.e have been introduced into the figure , in their relative proportions .
The growth of fungi and algro disfigured a large number of the thin sections in this and other genera .
Farther drawings are given below ( figs. 33A , 33B ) .
Radial Longitudinal Section .
The radial longitudinal sections through Goniastr(J3a septa were especially interesting .
The specimen mainly examined was a perfectly fresh one , and showed the Fig. 32 .
V1 ' , tr.ax .
tr.ax .
s.e. tr.ax .
tr.a.r .
g.l. Longitudinal section parallel with the septal plane .
Four trabecular axes ( tr.ax .
) run from their apices at the septal edge ( s.e. ) obliquely downwards through the section , and the latter thus meets four different " ridge " series of granulations ( gr. ) ; growth-lamellre ( g.l. ) .
calcareous lamellro in mostly unaltered conditions .
The surface section of a few ridges in fig. 32 shows the general distribution of the lamellar lines , although it was scarcely possible to reproduce them in a freehand drawing in all their fineness of structure .
152 DR. MARIA M. OGilNIE ON 'rHB MICHOSCOPIC AND The ends of the crystalline fibres mark out the edges of growth-1amellre like rows of beads .
In places where the heads of granulations have been rubbed off by polishing , a series of concentric lamellre still gives evidence of the characteristic conical shape of the lamellar deposit .
The polishing has the effect of showing the growth-lamellre of a granulation as it were in ground plan .
Fig. 33A .
\#183 ; d. h. Vertical section of septum penetrated with fungal growth .
Fig. ; ~:fo .
h. , ' , ' , C , Section parallel with septal plane .
d. dissepiment ; h. fungal hyphre ; s.e. septal edge .
Figs. 33A and 33B are given to illustrate fungal growth in coral skeletons .
Fig. 33A is a tangential longitudinal section of one septum of Goniastrcea , which is penetrated by fructifying hyphre .
Fig. 33B is a part of a , radial longitudinal section of a Mussa septum , in which near the edge the hyphre almost give the appearance of a regular canalicular system in the skeletal parts .
I have also had occasion to observe the extreme frequency of algal and fungal growths penetrating fossil specimens .
Summary of the Microscopic Sections .
Summing up briefly the results which may be deduced from these fine sections of Goniastrcea , we find :(I .
) That Goniastrrea , like Muss1 .
and Heliastnea , has septal 1\#183 ; idges , each of which is composed of a number of trabecul(JJ .
At the same time the upper edge of the septum is not deeply toothed in Goniastrcea ; the teeth are flat and have blunt points .
Yet the structure shows conclusively here as in 1.l1ussa and Heliastrma that the septal inva.gination is notched out into a series of main pouches ( corresponding to the teeth ) , and that in each pouch a number of active areas of calcification are marked out .
Calicoblasts are deposited around these areas , and the minute trabeculre are formed .
As thickening proceeds , the original distinction of separate trabeculre becomes rapidly lost in Goniastrcea , .
and the l , arge granulations ~hich form correspond usually in diameter to the .
whole diameter of the pouch , respectively ridge .
( 2 .
) That Goniastrrea , like Mussa and Heliastrrea , and 'unlike Galaxea , has a SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
153 wall mainly formed by thickening of the septa , and still further thickened internally by dissepimental deposit .
( 3 .
) GoniastrrBa , unlike Mussa and HeliastrrBa , has not a deeply-toothed septnJ edge .
This feature is probably correlated with a tendency in Gonictstrma and its allies to have paliform continuations of the septa .
Each such palm , has precisely the same microscopic structure as the septum , and is morphologically a specially " large tooth of great importance in fission .
The identity of the microscopic structure in septum and palus is important , as it at once distinguishes the pali of Goniastrma from those of the Trochocyathus group of Turbinolidre .
In the GoniastrrBa palus one , two , or more trahecular ridges may be present .
It is , in short , a pronounced form of the septal continuations in many Astrreidre which give rise to a so-called " spongiose " columella ( cf Rhabdophyllia ) .
MoNTLIV ALTIA .
Although it is not my intention in this paper to include detailed microscopic sections of fossil genera , I am of opinion that the important fossil genus Montlivaltia has a septum whose structure is nearest to those above described as of compound or radio-symmetric trabecular type .
Hitherto ] ; fontlivaltia , Thecosmilia , Rhabdophyllia , have been classified in the Lithophyllacere , E.H. , along with Mussa ; accord~ ing to their microscopic structure , however , they resemble more closely typical genera belonging to the Astrreacere , E.H. ( = Astrreinre , KLz .
+ Meandrininre , KLz .
) .
They have , like HeliastrrBa , ridges of equal width on the septal surfaces , culminating in .conical teeth at the upper edge .
Moreover , each ridge is marked out by a series of prominent granulations corresponding to successive segments of growth .
Figs. 34A and 34B are sectious made from the Triassic species Montlivaltia norica , * FRECH , and from the Cretaceous genus Montlivaltia rudis , E.H. , respectively .
Both sections I made in order to see finer details of microscopic structure than were already published .
The state of preservation was not sufficient to permit of the identification of all the microscopic details which I had been able to observe in recent genera .
Corresponding to each pair of superficial ridges , one trabecular axis is present in transverse section .
Fibres are radialJy arrange\lt ; l around this trabecular axis ( ideal ) , passing at once from the median plane upwards and outwards towards the surfaces .
Sometimes I could observe that the fibres were grouped in definite fascicles , whose origin in the median plane was marked by au individual point .
The " point " seemed sometimes lighter , sometimes darker than the fibrous structure , according to con\#171 ; - This species is described for the first time in FRECH 's " Monograph of the Corals of the Zlambach Strata " ( ' Palreont .
, ' 1890-91 , pp. 9 and 10 ) .
MCCCCXCVI.-B .
X ditions of fossilization .
It may be that those " points " indicate the axes of smaller trabecu1re acting as compacting tissue between main trabeculre ( cf Heliastrcea ) .
Fig. 34A .
Longitudinal section parallel with the septal .plane .
Each trabecula is seen t , o be a series of fascicles ( cf. figs. 10 , 17u , 25 ) .
Fig. 34B .
Transverse section of alternating thick and thin septa .
Zig-zag appearance of median " dark line " ( antea , pp. 116 , 117 ) .
Very often longitudinal sections through fossil Montlivaltias and Thecosmilias present irregularly dispersed dark points in the axial region of the trabecula .
This appearance has been observed in more pronounced degree also in sections of Cyathophyllum and its allies .
They apparently represent organic cell-matter , originally intermixed with the fibrous deposit at active centres of deposit .
The greater crowding of the " dark points " in the axial area agrees with the explanation given above for similar appearances in recent types ( antea , p. 139 ) .
Importance of the Direction of " Axes of Calcification " in a Septum .
It will now be readily understood that when a transverse section cuts a series of trabeculre almost perpendicularly to their axes , it shows a series of central regions , darker or lighter , as the case may be , around which the fibres radiate .
Eacli central region represents the axis of a trabecula ( in general language an axis of calcificationf and may show one , two , or several " points .
" If , on the other hand , a transverse section cuts trabeculre very obliquely , it will be in reality more a longitudinal section of the individual trabeculre , although it cuts transversely through the septum .
Then each trabecular axis will be represented by a longish series of central points , or by an apparently continuous central " short line .
" A zig-zag position of the " lines " is produced when the rows of granulations on the opposite surfaces of the septum alternate slightly in position .
This is an appearance which is very commonly shown in transverse sections cut through the septa of such Eusmilinids and Turbinolids as are " striated " on the septal surfaces .
It is therefore necessary to compare carefully the position which " strire " or " ridges " take up on the septal surface .
As an example , the " strire " on the surfaces of a septum of 'Purbinaria are almost horizontal with the edge of the septum , while SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
155 those on the flurfaces of a Galaxea septum are variably oblique to the septal edge .
The transverse section of the septum will show in the median plane an interrupted , often zig-zag series of short axes of calcification in the case of Galaxea .
In T.urb1'naria , on the contrary , the transverse Bection will show in the median plane one line of calcification , comparable with that seen in a longitudinal section parallel with a stria in Galaxea .
In other words , the grovi'th-segment of the Turbinaria septum is composed of a single trabecula with a single axis of calcification .
This is the direct homologue of a " septal spine " \#183 ; which is pointed horizontally inwards in the calyx of a Palreozoic coral .
The Galaxea septum is more complicatAd ; in each growth-segment a number of trabeculre are represented by a number of trabecular parts .
These are added at the growing edge , any one of them being the homologue of a " septal spine .
" The segmental parts may be inclined to the vertical at all different angles , from oblique to horizontal .
There are again a number of Eusmilinid and Turbinolid genera , in which the septa are described as " smooth , " having no strire or ridges on their surfaces , although they may have granulations .
Take , for example , Euphyllic1 ; .
The surface of the septum is dotted with minute granules .
By using a strong hand-lens , the granules are seen to occur in the course of delicate strire or ridges which are perpendicular to the septal edge .
About five or six strire may be couuted in 1 millim. The form of the strire is broad , the granulations rounded , whereas in Galaxea the strire are knife-edged and the granulations sharply pointed .
A transverse section of the septum Euphyllia has been published by BOURNE.{ ' It is described by him as composed of parallel fibres arranged on the opposite sides of one continuous median axis ( " centre " ) of calcification .
Actually , however , this " centre of calcification " is made up of a number of centres , very closely placed , and elliptical in outline .
The fibres emerging from those small elliptical axes are practically parallel , each group corresponding to the broad-shaped stria on the surface .
The fibreR are not obliquely placed with reference to the axes of the striiB , but , are almost horizontally oriented .
In other words , the deposit is laid down from the first by the lateral flaps of the septal invagination , and not at the upper edge where the flaps meet .
The absolutely smooth edge of the septum finds in this fact a simple explanation .
The septa of Euphyllia , , therefore , although finely striated on their surfaces , show two important features of distinction from the septa of types hitherto discussed :-( 1 .
) Smooth upper edge .
( 2 .
) The fibres are directed almost horizontally outwards from the median plane of calcification to the opposite surfaces .
These correlated features suggest that the septal invagination is not frilled along * Bomnrn , " On the Anatomy of Mussa and Eitphyllia , and on the Morphology of the Madreporarian Skeleton , " ' Q. J.M. s. , ' August , 1887 , Plate 3 , fig. 7 .
X 2 156 DR. MARIA M. OGILVIE ON 'l'HE MJCROSCOPIC AND its upper edge .
At the same time : the two flaps are not quite smoothly opposed to one another , but are very faintly fluted in the same manner as observed in Galaxea , but with the convexities broader than the intervening furrows .
We may then recognize in the septal structure of Euphyllia a number of trabeculre , parallel with one another and of uniform proportions ; the axis of each being elongated in the plane of the septum .
CHAPTER III.-CLASSIPICATORY CHARACTERS Ol !
' THE AS'l'RAJIDJE , E.H. GENERAL METHOD OF GROWTH .
I. Compactness of the Septum .
The invaginations of the aboral body-wall are formed in long radial lines .
The flaps of any one of the radial folds fall into a corresponding paired system of pleats within which the calcareous fibres are deposited .
The upward growth of the polyp takes place periodically , and at each period of growth a new edge is added to the septum .
The addition is composed of small groups of fibres within each pair of the pleats , each edge-group is attached lengthwise to the similarly-shaped group formed in the preceding period of growth .
And thus the continuous vertical series of fibrous groups arise , for which I have used the name of " trabecL1lre .
" Not only is the edge-group attached lengthwise to its predecessor , it is also attached sideways to the neighbouring edge-groups in course of formation at the same time .
In other words the whole of the new tier of edge-groups which is added to the septum during a single growth-period is firmly \#183 ; compacted before the close of that period .
The compact tier of edge-groups ( = trabecular segments or parts ) makes up one complete growth-segment of a septum .
By this feature of compactness in the method of growth , - the Astrreidre are distinguished from the Pseudoastrreinre , PRATZ , from the Lophoserinre , E.H. , and from the Eupsamminre , E.H. In all these sub-families , and more rarely in the Fungime , the tier of trabecular parts is not a compact septa1 segment .
On the contrary , the edge-groups are not coherent with one another originally , and it is only towards the close of a growth-period that they may be to some extent attached together .
Pores are left primarily between the adjacent groups , although in subsequent growth-periods calcareous tissue may close up some of these pores.* Two fadors therefore determine the compactness of the Astrreid septum-(1 ) the complete coherence of each new tier of trabecular parts as it is laid down ; ( 2 ) the precise serial succession of each new tier upon the foregoing , so that successivelyformed trabecular parts are in line with one another ( cf. Table , p. 252 ) . . .
KLUNZIKGER .
mentions that pores may occur occasionally in the septa of Ast.rreids .
They are , however , of rare occurrence , and never show the definite structural relations to a growth-period described above for those snb-families in which the septa are typically porous .
SYSTEMATIC srl'UDY OF THE MADREPOR .
ARIAN TYPES Ol !
' CORALS .
157 2 .
Compactness of the " Dissepimental " Floo1\#183 ; .
At the close of each growth-period in typical Astrreids , the basal portion of the aboral body-wall situated between the radial invaginations deposits a layer or floor of calcareous dissepiments ( Querblatter ) .
These are usually convex in form , with the convex surface upwards .
The aboral wall draws itself up gradually during the period of active deposit at the septal edge , forming during its updrawal the arched dissepiment .
Finally the period of " pause " sets in , and the body-wall between the septa remains supported on its many-arched floor , which is then completed and thickened .
The formation of dissepiments therefore keeps pace with the addition of new tiers of trabecular parts along the septal edges .
Each new platform of the disFJepiments is as much above the platform previously formed as is the new septal edge above the preceding .
The distance between two successive dissepimental floors is the measurement also of the additional height attained by the septa during a single growthperiod .
The hst-formed floor of dissepiments is always sonie distance below the septal edges .
Usually some five or six septal growth-segments are present between the dissepiments and the growing edges of the primary septa .
This fact allows lateral thickening of the septal growth-segmentR to go on during the interval which elapses between their first laying down at the septal edge and the ultimate withdrawal of the body-wall of the polyp .
Complete withdrawal of the body-wall from any particular tier of trabecular parts in a septum is of course marked by the formation of a dissepimental floor at the horizon just above it .
_ One reason why the primary and secondary ( sometimes also the next cycles in large calyces ) septa are usually much thicker than the septa of intermediate cycles is because they extend to a greater height above the dissepimental floor .
Consequently lateral thickening goes on during a greater number of growth-periods .
Another reason is that the fibrous deposit made during any single growth-period is itself thicker on the main septa than on the intermediate .
The formation of a complete basal layer of dissepiments at each period of growth is a highly distinctive feature for Astrreidre .
In the Funginre , Thamnastrrein re , Lophoserinre , Eupsammidre , the basal support is afforded in the main by cross-bars , " synapticulre , " over which the body-wall still descends in canaliculate fa8hion .
As upward growth ca.uses the aboral wall to be withdrawn from lower to higher horizons of synapticulre , an occasional flat dissepiment may be formed between adjacent synapticulre .
But in none of these sub-families do we find a definite succession of compact dissepimental floors as in the Astrreidre .
On the other hand , the basal deposit is also a means of distinction , within certain limits , between Astrreidre and Turbinolidre .
The Astrreidrn are mostly colonybuilding forms , and grow to various heights .
The Turbinolidre are all simple forms , and the majority are extremely short .
Basal deposit in them is limited to a solid outfilling of the lower part of the calyx .
Without doubt , the solid deposit is added to during the successive growth-periods , just as in Astneids , but there is no free space left by which one may distinguish periodic additions .
There is a group of genera included by EDWARDS and HAIME among the Astrreidre the Trochosmiliacere , E. H. in which the only feature of distinction from Turbinolidre is the occurrence of dissepiments .
The latter , however , are said , in many of the genera , to be " very scanty .
" On the other hand , zoologists have frequently described certain recent simple corals as Turbinolids in which a few dissepiments are present .
As a matter of fact , the distinction is not one which can be sharply made in \#183 ; nature , since all transitions are found between Trochosmiliacere with a rich development of dissepiments , and Trochosmiliacere with solid deposit in the base of the calyx , and but few dissepiments in higher horizons .
I shall have occasion , later , to point out other features in which the Trochosmiliacere show more affinity with the Turbinolidre than with the Astrreidre .
So far it is clear that , with regard to their basal deposit , the simple corals belonging to the Trochosmiliacere afford types transitional between the families Astrreidre , E.H. , and Turbinolidre , E.H. 3 .
Theca and Epitheca in Astrceidce , E. H. The ( ( theca " of the Madreporarian coral is one of the most vexed questions which come under discussion , and I shall have to refer to it again and again in the course of the present paper .
On looking at the drawing of Mussa ( p. 137 , fig. 22 ) , it seems very simple to distinguish a theca upon which the edge-zone of the polyp is applied from an epitheca .
The latter is formed as rings round the theca by the downward lip of the edge-zone .
Again , in 1-leliastrcea , the theca is clear enough , marking out almo_st equal i : ; eptal and costal halves of the radial structure .
An epitheca is apparently absent in Heliastrcea .
It is , indeed , absent around individual polyps of the colony , but there is a common epitheca at the base of the whole colony of polyps .
The same is true of Goniastrcea where a common epitheca surrounds the whole colony , and the individual polyps have a theca only .
A single branch of the Mussa corallum is the homologue of the whole massive corallum of a Heliastrcea or Goniastrcea colony , in so far as each is surrounded by an epitheca similar in origin , similar in structure .
Take again a simple Astrreid coral , e.g. , one of the most familiar of fossil Astrreids , Montlivaltict .
There an epitheca .
surrounds the simple calyx , identical in structure with that of the branch of Mu:ssa or the colony of Goniastrcea .
But there is no theca .
The same observation may be repeated for branching colonies of Thecosmilia , Calarnophyllia , Cladophyllia , \amp ; c. Each branch represents in these cases an individual polyp which laid down epithecal rings outside of and surrounding the septa .
It is the same in many Turbinolidre , in the majority of which an epitheca is SYSTEMATIC STUDY OF THE M.ADREPORARIAN TYPES OF CORALS .
159 present , either as thin rings or as a more solid and often ornamental deposit exterior to the septa .
Authors have sometimes called this deposit theca , sometimes epitheca , and sometimes have with justice got out of the difficulty by calling it both .
The essential difference which may be said to distinguish " theca " from , : epitheca " is that the theca , or wall , mmit be structmally associated with the peripheral ends of septa , whereas the epitheca is in no structural connection with the septa , but is a continuous concentric deposit exterior to all the other skeletal structures of a calyx .
Following this distinction , both simple and colony-forming Astrreid genera may be said to have always an epitheca : on the other hand , a theca may or may not be present in the " simple " genera , while , in colony-forming genera , the branching types may be without a theca ; the massive types always have a theca .
4 .
Entheca and Pseudotheca in Astrmidm , E.H. The terms " eutheca " and " pseudotheca " were introduced by HEIDER and adopted by ORTMANN ( see below , pp. 310-314 ) .
A perfect " pseudotheca " has been figured and described above , under Goniastrma .
It is a \#183 ; theca wholly composed of the thickened peripheral parts of septa .
An " eutheca " has been described under Galaxea .
It is a theca in which distinct wall-lamellre are present between the peripheral parts of the septa .
The two may grow at different levels of the calyx , but unite in course of growth ; or they may grow abreast of one another , and have their centres of calcification continuous with one another .
Amongst the Astrreidre , E.H. , it is only in the sub-family Eusmilinre that an eutheca is developed .
In all the Astrreinre the theca is either a perfect pseudotheca , or a slightly-modified form , in which erectly-placed dissepiments or large septal granulations help to compact and thicken the wall ( ex. , H eliastrma ) .
i ; Occasionally , pores are said to be left jn the pseudotheca ; in the eutheca I have never heard of unfilled spaees .
Typically , however , the theca of Astrreinre may be regarded as compact .
In this feature it contrasts with the porous wall of the Madreporaria perfurata , E.H. , and also with the Pseudoastrreinre and Funginre , in which a porous synapticular wall is present .
THE Sun-FAMILIES AsTRJ .
EINJ .
E , E.H. , AND EusMILJNJ .
E , E.H. The Eusrnilince , E. H. The basis of distinction for the two sub-families of the Astrreidre rests at present on the external features of the septum .
The septum of the Astrmince genera is said to be toothed or spined at the upper edge , and the lateral surfaces of the septum are * The original genus Astrcea of LAMARCK coincides with the Favias and Heliastrceas of later authors .
EDWARDS and HAIME describe the septa in one of the mm ; t common recent species of Heliastrcea ( H. Forskcelana , E.H. ) as cribriform in structure ; yet the same authors give cvrnpact septa in their diagnosis of the sub-family Astrreinre ( ' Coral .
, ' II .
, p. 286 ) .
All such pores are of intertrabecular position , and serve to show that the limiting-line between Astraiidre and FLrngidre cannot be very sharply drawn .
marked by a fan-shaped system of ridges , or by rows of granulations placed piwpendicularly to the septal edge .
The Eusmilinid septum , on the other hand,.is described as smooth-edged , the lateral surfaces frequently bearing granulations placed parallel with the -upper edge .
The types examined in detail above do not carry out these features of distinction .
Goniastrwa , an Astrreinid , has a septum which has faintly-marke\lt ; l ridges on the lateral surface , and very slight indications of teeth at the edge .
Galaxea , ranked as an Eusmilinid , has also a septum with striated lateral surfaces , and slightly serrated edges .
In neither are granulations prominent ; they are in both cases placed along the strim , and are , therefore , perpendicular to the septal edge .
It may be said , and with justice , that Galaxea is altogether out of place among Astrreidre , since it has a well-developed crenenchyme .
Even if Galaxea.be removed , the same difficulty remains .
I have seen plenty of species of Stylina with serrated edges , and all have striated surfaces .
The genera allied with Stylina have similar septa ] features .
Again , the wall of these genera is a pseudotheca of as perfect a type as that of Goniastrcea .
Lately Astroccenia and Stephanoccenia have been transferred by some authors from the Eusmilinre , in the neighbourhood of the Stylina group , to the Astrreinre , on account of their toothed septal edges .
If only the serrated edge were to be considered , several more genera might have been transferred .
But there are several other considerations , e.g. , the thecal and basal deposit , the symmetry of the septa i~ the calyx , the presence of a true columella , or a pseudo-columella , \amp ; c. Other authors , again , have laid no weight whatever on the nature of the septal edge ( see DUNCAN , " Review of the Madreporaria ' ' ) .
The serrated edge is , however , an indication of structure .
There would be no chance of mistaking a good transverse section through the septum of a serrated type , e.g. , Galaxea , Goniastrwa , with a section through a type like Euphyllia , in which all the septa are perfectly smooth-edged .
The transverse section of the serrated type shows a number of distinct trabecular centres of calcification in the median plane of the septum , that of Euphyllia shows one plane centre of calcification , on both sides of which the fibres are arranged parallel with each other .
Moreover , the fibres in the latter case are almost horizontal in position , indicating that they are laid down from the first by the lateral surfaces of the two flaps , whereas the obliquely-placed fibres in the serrated type indicate that the fibres are first laid down along the actual edge of the septal invagination ( p. 15 5 ) .
So far as I have been able to study the structure of the Trochosmiliacere , the septa agree with that described for Euphyllia in the main , e.g. , Plac06milia , Circophyllia .
The disposition of the fibres is the same , and the marked " two-leaved " character of the septum .
Aplophyllia and Rhipidogyra are colonial Eusmilinids , with confluent calyces , whose septa are also built_ on the type of Euphyllia .
All these Eusmilinids stand , therefore , in distinct contrast with the Astrreinre , while a number of genera , such as Galaxea and Stylina , which have hitherto been SYSTEMATIC S1'UDY OF THE MADREPORARlAN TYPES o~\#183 ; CORALS .
161 erroneously classified as Eusmilinids , have septa with a structure , and with external featureEi , resembling the more delicately built septa of certain Astnein re , e.g. , Goniastrcea .
It is clear , therefore , that the sub-family Eusmilinre , E.H. , includes within it two very distinct septal types , the one of which is practically identical with the Tilrbinolid type , while the other is more akin to the simplest Ast1\#183 ; ceicl type .
The two Turbinolid genera , of which transverse sections have most often been made , are Flabellum and Caryophyllia , and BouRNE 's section of Euphyllia , RE1s ' oK .
section of Circophyllia , and other sections already published of Eusmilinids , may readily be compared with those of Flabellum , and Caryophyllia ( see Introd .
, the works of KooH , FOWLER , NICHOLSON ) .
I propose to entirely break up the sub-family Eusmilinre , E.H. A. The simple c0rals , the Trochosmiliacere , may be united with the Turbinolidre.t Many features , in addition to that of septal structure , speak for this alteration the frequent poverty of dissepiments in the Trochosmiliacere ; the presence of a bmellar colurnella , sometimes continuous with a main septum , and giving bilaterality of parts to the calyx ; the presence in other cases of a large styliform columella ; independent of the septa .
Moreover , the habit of the corallum is very similar , the polyp increaRes its size not so much by adding to its height as by adding to its width .
The calyx : becomes , therefore , elongated and compressed in shape .
In the meantime I shall retain EDWARDS and HAIME 's title name of Trochosmilia , only altering the tribe into a sub-family , Trochosmilince .
B. The fissiparous Eusmilinid corals , Rhipidogyr\#183 ; a , \amp ; c. , will be fully discussed in my 1 ' Monograph of the Upper Jurassic Fauna of Stram berg , " shortly to appear .
There the reasons are given for uniting them into a new family , the Amphiastrreidre .
The recent genus Euphyllict is rather anomalous in many ways , but its nearest affinity seems to he with the fissiparous types of the Amphiastrreidre .
The structural relations of the wall and septa are alike in both , also the fine structure and external appearance of the septa .
Type : Euphyllia.-The septum is an enfold from the wall and has a two-leaved appearance , with a distinct dark line , often looking double , in the median plane .
This continuous dark line may be looked upon as a series of very closely set centres of calcification , or as one long axis of calcification .
The calcareous fibres radiate at a low angle outwards from the median plane to the surfaces .
It happens in some genera that secondary surface thickening gives rise to horizontal rows of boss-like granulations , each row marking a single growth-period of the septum .
If we regard the enfolded tissue in the median plane as originally one horizontally * REIS .
See footnote below , p. 286 .
t VERRILL has already suggested that a sub-family , Flabellinre , be formed , including severn , l of the Trochosmiliaceai , with Flabellum and ot.hers of the Turbinolinai , E. H. , mainly on account of the fact that the tentacular disc is not stretched outwards beyond the wall .
MD0U0X.0VI.-B , Y lG2 DR. MARIA M. OGILVIE ON 'l'HE MICROSCOPIC AND placed axis of calcification , the length of the septum , with a single horizontal row of granulations , corresponds to a single growth-segment and is the morphological equivalent of a horizontally placed " septal spine " in Palreozoic forms ; at the same time , moreover , the equivalent of either the septum or the septal spine in recent Madreporas .
The general resemblance of the corallum of Euphyllia to that of Mussa is entirely dependent on the fissiparous budding in both and cannot be regarded as an essential classificatory feature .
BouRNE showed , it is true , that J.11.ussa and Euphyllia were alike in the absence of " directive mesenteries .
" But FOWLER proved this feature in Lophohelia .
, an Oculinid , and HEIDER showed in Maclracis that the position of the muscles in the mesenteries was very variable ( see below , p. 290 ) .
BouRNE 's own suggestion was that their absence in Mussa and Euphyllia was merely an accessory feature correlated with the fissiparous mode of increase .
Euphyllia has peculiarities not yet found in Milssa or any other Madreporarian genus .
The ectodermal cells of the stomodreal tube have digestive functions , and the tube is drawn out into a number of convoluted channels , which lie in the visceral cavity .
The endoderm , on the other hand , forms a highly vacuolated tissue , containing innumerable algal cells ( cf BOURNE , loc. cit. , ' Q. J , M. S. ' ) .
I1'ig .
35 .
Typical Stylinid colony ; letters as in fig. 2 , p. 107 ; a columella is present in the middle of the calyx ; the edge-zone extends horizontally-it is lettered " cs .
" coenosarc , for comparison with Tnrbinaria , fig. 56 , and Galaxea ; psth .
pseudotheca ; ic. l. intercostal loculns .
C. The Stylina Group of Eusmilinicl Genera.- !
propose to treat them as a separate family , the Stylinidre , very closely related to a large number of the Amphia.strmid.re .
The early :Mesozoic Stylinids are , farther , scarcely distinguishable from some \#183 ; colonial so-called " Rugose " genera .
Diploccenfo has two " walls , " and between them lies the broad dissepimental zone so characteristic of Acervularia and allied genera .
Cyathophora and other Stylinids have central tn.bulre .
Again , the varying number of primary septa and the definite cyclical arrangement of the septa had already appeared in Deca ccenia , Pentaccenia , and other young-Palreozoic types .
Galaxea.-This genus has a somewhat isolated position among recent Madre* poraria .
\#183 ; It has most in common with the Stylinidre , but the presmce of a true theca and the limited extent of the costro lead me to place it only provisionally within this family .
SYSTEMATIC STUDY Ol !
, THE MADREPORARIAN 'l'YP .
ES Ob , CORALS .
163 The Astrceiclce , E.H. ( excl .
Eusmilince ) .
The Astrreim:e genera formed , in EDWARDS and HAIME 's classification , a wellmarked sub-family which I would now regard as the sole representative of the family Astn.eidre , E.H. The family would , if thus limited , include almost entirely colonybuilding corals in which neither " ccenenchyme " nor a " true theca " is developed .
In other words , the adjacent polyps of a colony are united directly by their edge-zones , e.g. , H eliastrcea , or the edge-zone may be scarcely developed and the adjacent polyps be united by pseudothecal thickening , e.g. , Goniastrceci .
The leading features of the skeleton are :(I .
) Absence of ccenenchynie .
( 2 .
) The " wall " is pseudothccal in character , the septa always taking the mam part in its construction .
When no lateral thickening of the septa takes place , there is no " wall , " and the septa are supported only by dissepiments .
( 3 .
) A well-developed " epitheca " is always present surrounding individual branches of branching colonies , limited to the base in massive colonies .
It is coalescent with the " pseudotbecal " septal thickenings of calyces at the edge of the colony .
( 4 .
) The septa may have either deeply or slightly-toothed edges , and the septa of the less-developed cycles in some genera may look almost smooth-edged .
The lateral surfaces are always ridged .
and granulate , new ridges being introduced at one definite area in the septum , this is usually the area at which pseudothecal thickening of the lateral surfaces takes place .
This area may be called the area of divergence , since the ridges diverge right and left from it to the septal and costal portions of the radial structures .
Microscopic examination shows that the septa are composed of a number of calcareous trabeculre arrnnged compactly side by side .
The trabeculre usually form small fan-shaped groups corresponding to the individual ridges on the lateral surfaces of the septa .
The calcareous fibres may be arranged with bilateral or radial symmetry relative to the trabecular axes of calcification .
They are always originally deposited by the growing edge of the septal invagination , and are oriented therefore obliquely upwards and outward : : ; from the trabecular axes .
The granulations on the septal surfaces follow the direction of the individual trabeculre .
If the latter form a broad ridge as in Mussa .
, each trabecula has its own distinct row of granulations , and the septal surface appears richly granulate , If the riclgt\gt ; s are narrow , the granulations of the crowded trabeculm coalesce and form one large granulation , e.g. , Heliastrcea .
Large granulations of adjacent septal surfaces may meet across the intersephtl locu]us and form a " pseudosynnpticula .
" ( 5 .
) The floor of the calyx is formed by regularly deposited dissepiments .
A. columella independent of the Repta is rarely developed .
But the inner ends of the longest septa are often drawn out as long teeth or pa]i and inegular]y compacted to form a pseuclocolmnella .
Y2 Systematic Features of Generic Value .
Macroscopic features are :( l. ) The shape of the septal edge , lanccolate , exsert , \amp ; 0 .
, and its minute details .
( 2 .
) The constitution of the columella where such is present .
( 3 .
) The characteristic form of the corallum .
Microscopic features are :( J. ) The special structure of the ridges and granulations on the lateral surfaces of the septa .
( 2 .
) The average dif ; tance from one another of the trabecular centres or axes of calcification in the median plane of the septum .
( 3 .
) The uniformity of the trabecnhe in the septa , i.e. , if all have a single a.xis or paired fascicles , and are , on the whole , bilaterally symmetrical in their fibrous deposit , e.g. , Mussa ; or if all have a " group " axis with a wide circular periphery and have a radially symmetrical fibrous deposit e.g. , 111ontlivaltia .
These are both uniform types .
On the other hand , many types are not uniform in so far as they have some trabecuhe with radially symmetrical fibrous structure , and others with bilaterally symmetrical structure , e.g. , Heliastrcea , Goniast1\#183 ; cea ( cf p. 141 ) .
1Vote on the J ?
o1\#183 ; m and lncrnase of the Cornllum .
Tn examini1ig the Upper Jurassic corals from Stramberg , I found I could arrange a perfectly graduated series of specimens belonging to the genus Thecosrnilia , beginni11g with the highly branched forms whose calyces were at the top of free stems , and ending with low massive forms in which the calyces remained either in contact with one another at their walls or appeared irregularly fissiparous .
No author has given more attention to the features of budding and fission in classifying the Astrreidre than DUNCAN.'-\lt ; In his diagnosis of the genus Thecosmilia ( loc. cit. , p , 81 ) , we read , " Growth by fissiparity and rapid jsolation , or by gemmation and subsequent fissiparity .
" However the fossil skeleton may appear , there can be no doubt that the living p1\#183 ; ocess in the polyp has one and the same result , whether the palreontologist calls it " gemmation " first or " fissiparity " !
And it is hopeless to attempt to strain the same process for purposes of classification into two , hence I doubt the value of DUNCAN'S sub-families .
These are solely based on modes of increase and expressed as AsTRJEIDJE ( 1 ) sirnplices , ( 2 ) reptantes , ( 3 ) gernrnantes , ( 4 ) ccespitosce , ( 5 ) con.fiuentes , ( ti ) a_gglomeratce fissiparantes , and ( 7 ) agglumeratro gemmantes ( DuNCAX , Zoe .
c'it .
, p. 50 ) .
In the same way as for Thecosmilia , one might arrange a graduated series of the * DUNCAN , " A Revision of the Families and Genera of the Madl'eporaria , ' ' Linnean Society , London , 1884 .
SYSTEMA'rIC STUDY 01\lt ; ' THE MADREPORARIAN TYPES 01 ' ' CORALS .
165 genus Isastrcea , beginning this time with the typical massive form of the Isctstrcea cora1lum , in which calyces were clearly walled off from each other , and ending with a typical Lati'.mceandra , in which few calyces have their own walls , most being arranged in confluent series , irregularly walled and winding .
The individuality of the polyp is never so great in Isastrcea as in Tliecosniili'a .
, since the Thecosmilian polyp at the e11d of a free stem carries on a much more independent existence than an Isastrooan polyp on the surface of a compact colony .
Again , although Thecosniilia may also have a massive corallum , the dignity of the polyp never descends so far as in that of a typical Latirmeandroid .
And we shall generally find that there is a certain standard of independence for each genus , which may be faken as a measure of the genus in placing it systematically .
We may , in fact , use the form of the corallum and its habit of increase as a generic feature when we keep in view the highest degree of polypal individuality attained and the lowest .
No sooner do we keep both ends of the scale before us than " gemmation " and " fissiparity " merge in one another .
" Fissiparity " I take to be simply a hastened development of buds under certain conditions and at certain pe1\#183 ; iods of " vegetative g1\#183 ; ow~h " i\#183 ; n the life of the co1'al .
And the buds , which form long mean\lt ; lroid series and never develop fully , probably perform in an Astrmi.d colony the functions of a ccenenchymal area in " Perforates , " namely , the rapid circulation of nutritive material throughout the colony .
GEMMA TION AND FISSIP ARE .I'l'Y .
As I hay already said ( antea , p. 147 ) , all the forms of budding may be reduced to intracalicular and extracalicular , and these are merely varieties in the mode of vegetative increase .
A bud may form in a colony at any part of the oral body-wall , jn Astrreidre it forms usually on the peristorne or above the lip of the calyx .
If the bud is open to the main visceral cavity of the polyp , it is said to be intracalycinal ; but if it forms just on the other side of the rim of the calyx and is open to the continuations of the cmlenteron on the outer side of the theca ( the crelenteral canals of the ' ' Randplatte " ) , the bud is said to be extracalycinal or extracalicular .
In the former case the appearance observed in the skeleton is called " fissiparity , " as the tendency of the corallum is more or less meandroid ; in the latter case , the appearance is c'1lled gemmation , and the tendency of the corallum is towards greater individuality of the calyces and branching growth .
The canalicular prolongations of the ccelenteron on the outer side of the calycinal rim may sometimes be very extensive in length , and may either hold closely to the wall of the calyx ( e.g. , Astneidre ) , or may spread over a wide jnterpolypal surface ( e.g. , Stylinidffi , Oculinidre , Eupsammidre ) .
Buds may form at any part of this extension , but then they are otherwise narned-infracalicular , stolonal , ccenenchymal , \amp ; c. As I have said , in all cases the process is gemmatiun , and the same corallum may develop its buds anywhere .
It is only where canalicular extra , thecal continnations 166 DR. MARTA M. OGILVrn ON 'l'HE MICROSCOPIC AND are practically absent , that the polyp is limited to so-called " :fissiparity , " e.g. , Goniastrceci , lsastrcea .
Take as an example of the redundancy of budding which the coral may show , the following quotation from DUNCAN ( Zoe .
cit. , " Astrangia , " pp. 66 , 67 ) : " There is no doubt that the species . .
\#183 ; ... .
vary in their method of gemmation .
The corallites may entirely arise from the basal expansion , or some may come from the angles between corallites and even from the wall anywhere below the calyx .
Some of these buds are accidentally placed , and have no intimate union with the parent ; but others appear to be tl'Ue infracalicular gemmations .
" Sub-di1 , isions of Astrceidce ( = Astrminm , E.H. ) .
The best basis on which minor sub-divisions of the Astrmidm can be made is probably that of the septal structure .
The principal types of structure have been sho , un in chapter ii .
, they are : ( I ) Montlivaltia or Thecosmilia , ( 2 ) Mussa , ( 3 ) Heliastrcea , ( 4 ) Goniastrcea .
At the same time the relation of the " edge-zone " ( or costate part of the calyx ) to the main visceral cavity ( or septate part of the calyx ) will be found in most cases to be a correlative featnre of some importance .
( 1 .
) Lithophyllicice{I !
( pars .
E.H. ).-Solitary or colonial ; the buds develop as free branches , occupied by one polyp , seldom by a group .
Epitheca is formed round each branch , pseudotheca is little developed .
The septal edges have equally developed , conical teeth , and the lateral su1f aces of septa hai.e uniform ridges , on each of which one main row of granulations i\#183 ; s present .
Genera : Thecosmilia , Cladophyllia , Rhabdophyllia , Cladocora , \amp ; c. ( 2 .
) Mu .
, sctcece.-The buds develop as separate branches .
Intracalycinal budding takes place within the mother-polyp of the branch , so that each branch is the Reat of a small :fissiparous polyp group surrounded by epitheca , Pseudotheca present , close to the epithecal rim of the calycinal group .
The septal edges have teeth iinequally developed , and usually spiniform .
The ridges on the septal surfaces are correspondingly variously developed .
Each r\#183 ; idge has a broad surfiwe on which numerous granulaiions are scattered .
Genera : Mussa , Dasyphyllia , Isophyllia , Symphyllia , Ulophyllia , \amp ; c. ( 8 .
) Astrceacea1.-The corallum is massive .
The buds develop as branches , which r ?
main constantly attached at the rim of their edge-zones .
The epitheca is limited to the base of the whole colony .
Pseudothecm are developed well within each of the individual calyces .
The edge-zones are represented in the corallum by the broad costate portion external to the pseudothecre of the calcyes .
The septal edges have eq ?
ially developed SYSTEMATIC STUDY 01 !
' THI\#163 ; MADREPORARJAN 'l'YPES OF CORALS .
167 teeth ; prominent ridges are present on the late1\#183 ; al su1fctces bearing large1\#183 ; and smalle1\#183 ; g1\#183 ; anulations , the larger occasionally forming synapticulce .
Genera : I[eliastrcea , Confusc1strcea , Agctthiphyllia , Stylastrcea , \amp ; c. ( 4 .
) Faviacece.-The corallum is never branching , always massive .
The edge-zones of polyps are extremely limited , and increase takes place almost wholly by fissiparity within the calyx .
The septa ctre only slightly toothed and the ridges on the lateral surfaces are extremely fin e ; gmnulations , small .
The individual calyces may have a distinct wall ( ex. Favia ) or a wall common to aclJacent calyces ( ex. Goniastrcea , lsastrcea ) .
The epitheca is well developed , limited to the base of the whole colony .
( 5 .
)_ Meandracece ( provisional , until the inui vidua.l genera are microscopically examined ) .
The buds rarely develop individually ; long series of calyces are divided f1\#183 ; om neighbouring series by a cormnon wall .
The septa may be deeply toothed , or very slightly .
This tribe includes at present mea.ndroid forms related to Astneacere and Faviacere , i\#183 ; .e .
, according to their septal structure , the meandroid forms may be ultimately referred to one or other of these two tribes .
Genera : JWeandrinct , Latimeandra , Diploria , C03loi\#183 ; ia , Hydnophom , \amp ; c. There are few corals of solitary habit among the Astrreidre .
At present several are classified as Astrreidre which have not the characteristic features ; for example : Epistreptophyllum , a Jurassic genus which must be placed in the neighbourhood of the Eupsammiche ( vide chapter x. , p. 303 ) .
LeptophylHa , a Jurassic and Cretaceous genus which PR .
A.TZ has proved to belong to the Pseudoastrreinre ( vide chapter v. , p. 18 6 ) .
CHAPTER IV.-FUNGIA AND SIDERASTRJEA . .
FUNGI .
A. .
The so-called " porous " wail of Fungia is homologou , s with the " compnct " wall of Astrceid Types , and is like it , a pseudotheca .
The genus Fungia has many points of interest .
It was to denote the calcareous bars in the interseptal loculi of Fungia that EDw .A.RDS and H.A.IME originally applied the name of : , synapticula."* Again , the developmental history of Fungia is remark* EDWARDS and Hn1 , rn , 'Coralliaires , ' 1857 , vol. 1 , p. 66 , " Sometimes septal granulations become very large , coalesce with those of the continuous septa , and thus give rise to small solid ha-rs , which divide each interseptal loculus into a series of small spaces communicating with one another .
The presence of these solid bars , which we call synapticulce , characterize the family Fungidre , and has not been obserrnd as yet in any other group of the class of Madreporaria .
It is in the genus F , ungia that synapticulre have the greatest development , as in that genus they sometimes extend in the form of almost cont.iimous curved lines from the bottom of the interseptal loculi to within a small distance of the free edges of the septa .
" 1G8 DR. MARIA i\l .
OGIINrn ON THE MICROSCOPIC AND able .
The young polyp possesses an upright , cup-shaped theca , with marked bilateral symmetry of the septa .
But , as it becomes larger , the corallum develops in discoid form , until the theca rests almost horizontal with the base , and the growing exsert tips of the septa are to be found at the periphery of the disc ( cf BOURNE/ Zoe .
cit. , Plate 24 , \#163 ; gs .
3- 8 ) .
No sign of the original .
bilateral symmetry is retained in the mature cornJlum , the septa being radially arranged .
The theca has the same general relations to the crelent , eron and mesenteries as that of other Madreporaria .
According to BouRNE , the theca , or wall , is " formed from the fused ends of contiguous septa , as is statod by G. v. KocH to be the case in the lateml thecre of other Madreporaria .
" This statement is perfectly correct , yet the wall is homologous , strictly speaking , only with a " pseudothecal " wall such as the Astrreids possess . .
A gm1eral impression obtains that the porous nature of the wall of Fungia at once distinguishes it from the compact wall of Astneidre , or even from that of the more nearly related Lophoserinre , E.H. The apertures in the Fungia wall are , however , of small moment .
If we removed the dissepiments between the thick peripheral portions of the septa of many Astrreids ( e.g. , Heliastrcea , Mussa ) , we would have numerous apertures left in the so-called " compact pseudotheca .
" Such apertures would be the exact homologue of those in the Fnngia wall , which are simply occasional interseptal spaces not sufficiently filled up by the basal deposit .
This happens very readily , since the basal deposit in Fungia takes exclusively the form of synapticulce , i.e. , rod-like bars , instead of flat , or archerl dissepiments .
In the Lophoserinre , on the contrary , dissepiments are well developed in addition to the synapticulre .
Fig. 36 presents a diagrammatic section through the corallum of Fungia , designed to show the relations existing between the aboral body-wall of the polyp and the skeletal parts .
Comparison should be made with the similarly designed section through jfilssa .
The Syriapticulce arefo1\#183 ; med within Interseptal Invaginations of the Aboral Body-wall .
BOURNE has given in the paper quoted above a series of figures illustrative of the atta.chment of the mesenteries to the body-wall .
" All are firmly attached by their upper border to the peristome , " the primary and secondary mesenleries being attached at their inner border to the stomodreum , the remaining orders being of " constantly decreasing length .
" The mesenteries radiate freely in the interseptal spaces near the upper margin , that is , above the upper ends of the synapticulrn , but are naturally interrupted in the main portion of the interseptal loculi by the presence of the synapticular bars ( cf. Bo-cRNE , Zoe .
cit. , pp. 302-303 ) .
Exactly how this * BOURNE , " The Aaatomy of t.110 Madreporarian Coral Fiingia , " 'Q .
J. M. S. , ' 107 , January , 1887 .
~ tj 0 0 0 ~ C --1 !
' " ' I p ; N 0 { [ .
Fig. 313 .
e FUNGIA .
The oral ( o. ) and aboral ( a. ) body-wall are doubled round , so that the part correspouding to the edge-zone ( e. ) of ltfussa is here found ou the basal surface syn .
synapticulre ( cj .
pp. 168- 171 ) .
'l'his diagram is to be compared especially with fig. 22 .
Two non-cmnenchymatous types , Mussa and Fungia , are represented in this way for the sake 0\#163 ; comparison with cwnenchymatous types , e.g. , Galaxea , fig. 2 , and Turbinal'ia , fig. 56 .
m t-1 Ul tzj ~ i\gt ; i- , 0 m 8 q t : ; I " 1 0 rzj \#183 ; 8 P:1 tr : ; ~ P--- 0 ( ) l ; i:i t:_'rj 'ti a. 0 l ; i:i P--- l ; i:i H \gt ; --z 'ti : : , ; i m 0 rzj 0 0 l ; i:i ?
-- t : " ' ff.l ... . .
C ) \lt ; :. . ?
J 70 DR. MARIA M. OGILVIE ON THE MICROSCOPIC AND interruption takes place is not clearly explained by BouRNE ; possibly his view is best .
rendered in his subsequent paper ( " The Anatomy of Mussa and Eitphyllia , " loc. cit. , p. 47 ) .
There he writes , " The manner in which the mesenteries may be pierced , as it were , by outgrowths from the walls of adjacent septa is well shown by the form of s_ynapticula in Fungia .
In this form processes arise from the walls of the septa \#183 ; which grow towards similar processes from the contiguous septa , and , meeting them , fuse with them to form synapticulre .
" It is true that the synapticulre grow out as sepfal processes , though their fusion in Fun9ia is ordinarily brought about by the development of a new centre of calcification ; but the important point is that they neither " 1:nterrupt " nor " pierce " the 1nesenteries .
As I show in fig. 36 , the pa.rt of the body wall to which the mesenteries are attached becomes gradually more and more invaginated , raised up on the growing ends of the synapticular bars , but still descending to the base of the polyp between the bars of any one interseptal loculus .
BocRNE did not make a special study of the skeletal , but of the anatomical relations , and his views with regard to the synapticulre rested chiefly on previous observations made by EDw ARDS and HAIME and by DUNCAN.* He says , in fact , at the conclusion of his paper that he " can offer no explanation of the origin and significance of the synapticula .
Physiologically they seem to serve as stays or buttresses giving solidity and coherence to the corallum .
" The structural support which they give is possibly their main function , in this respect they take the place of the \lt ; lissepiments of Astrreid.-:e .
Their growth-relations with the other skeletal parts are precisely similar to those which I have demonstrated for dissepiments in the foregoing Astrreid types .
The first-formed synapticular floor in \#163 ; g. 37 is lettered " S1 , " the last-formed synapticular floor is lettered " Szs\#183 ; " The whole \#163 ; gure represents a , Ji'ungia septum placed vertically , instead of in its natural horizontal position .
This I have done in order to make the comparison with the septum of Astrreids more simple and to bring out the analogy between the successive " synapticular bars " of Fungia and the successive dissepimental floors of Astrreidre ( antea , p. 157 ) .
Growth in the Fungia septum goes on not only at the growing apex of the septum but throughout its entire length ( resp. height ) .
So that the earlier-formed synapticular bars are constantly being added to as fast as the edges of the septum grows upwards .
Their upward extension keeps pace steadily with the growth in height of the sept~ .
The particular curve which is formed , if one draws a line , on the septal surface just above the uppermost ends of the whole series of synapticular bars , is at about the same distance from the septal margin throughout its whole length and is parallel with it .
It corresponds , therefore , to a curve following a row of dissepiments in an Astrreid polyp ( cf Mussa , I-I eliastrcea ) .
What is then the particular advantage to the Fungid polyp of having synapticular * D oNCAN , " Observations on the Madreporarian Family of Fungidre , " 'Journ .
Linn. Soc. , ' 17 .
SYSTKMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
171 hars as a skeletal support instead of dissepiments ?
I believe the main advantage to be that , by this means , an increased endodermal surf ace i\#183 ; s afforded within tl.e visceral cavity .
The synapticular invaginat.ion of the body-wall produces an elongated fold of the soft parts in the course of growth and gives rise to visceral canals on both si\lt ; les of the synapticular prop. If a dissepimenta.1 floor took the place of the synapticular bar at any horizon , the surface of endoderm would be very short , whereas in the synapticular union there is a double endodermal surface from the upper part of the synapticn1ar bar to its base .
The advantage is analogous with that which is attained by highly crimped and folcled or pitted flaps of the septal invaginations as compared with smooth flaps ; or again , by invaginating the basal body-wall over a complicated colume1Ia , spongiose or paliform ; again , by a highly canaliculate cumosarc , as in M. porosa , in place of the smoothly extended crenosarc of Galaxea .
In Fungids , the increase of the visceral surface plays at any rate two roles in the existence of the polyp .
First , the larger assimilative area means more rapid digestion of the food-material introduced into the visceral cavity .
Second , the canaliculnr form of the continuity of the peristornal soft parts with the thecal and basal affords the possil:)ility of a more highly differentiated muscular system within the _polyp , while it necessarily gives firmer attachment of the polyp to the skeleton .
The genus Fungia itself , owing to its unusually large dimensions , shows those features best , but the skeletal structure of the small Siderastrcea or Lophoseris polyp points to a similar t::iignificance of the synapticulre .
The Septa .
The thinner septa of Fungia have a wavy margin with extremely fine teeth , and the surfaces are closely covered with granulations .
Thicker and thinner parts of the septum may be distinguished near the edge and parallel to it , the thicker parts bearing larger granulations .
The thinner parts mark the limit between successive growthsegments .
The thicker septa are deeply toothed and have ridged surfaces ; the ridges thicken as they become more remote from the septa .
I edge and are continued as the synapticular bars .
Each synapticular bar therefore represents coalesced ridges on the adjacent surfaces of two septa .
The important macroscopic features of a thick septum are shown in fig. 37 .
The figure is represented purposely in a position at right angles to its true position in the corallum in order to facilitate comparison with the similar foregoing figures of Astrreids .
The " area " from which the septal ridges diverge may be seen to extend almost vertically in the plane of the septum-that is in the same direction as in Galaxea , Mussa , \amp ; c. ( antea , figs. 3 , 15 , 22 , 28 ) .
It is only the deformation in shape undergone by the growing Fungia polyp ( antea , p. 1G8 ) which renders the true inner edge of the septum the apparent uppe1 ' edge in the mature corallurn , and the true outer edge of the septum the apparent base .
The zig-zag markings , more or less parallel with the septal edges , denote , as z 2 Fig. 37 .
a.d. syn .
28 .
syn .
20 .
syn .
10 .
gr. syn .
I. ps.c .
a.d. Septum of Fungia represented in a vertical position instead of the horizontal position assumed in nature .
The area of divergence ( a.d. ) of the trabecular ridges is thus seen to follow a line in the septal plane rather nearer the outer or " costal " edge than the inner or " septal " edge of the septum ( cf. fig. 15 ) .
sp. = spiniform teeth of the costal eJge .
By1i .
= r. ; ynapticuloo in ascending series .
gr. == granulations .
ps.c .
= pseudo-colurnellar teet , h. SYSTEMATIC STUDY OF THE MADRF]PORARIAN TYPES OF CORALS .
173 before , successive periods of growth , and define the limit of successive growth-segments in the septum .
It may he observed that granulations are scattered alike over the septal surface and the synapticular bars .
The granulations on the synapticulre can only be actually seen at their upper parts , where coalescence has not yet taken the place of the synapticulre on the adjacent faces of two septa .
In fig. 3 7 , the granulations are for the sake of uniformity carried out along the whole synapticular bar .
The general resemblance of the septum of Fungia to th.-1t of an Astrreid , more particularly one with broad ridges and irregularly scattered granulations , like Mussa , is sufficiently demonstrated by the -figure .
The fine structures of the ridge is also in the main comparable with that of a ridge in Mussa ( cf fig. 38 ) .
A diverging system Fig. 38 .
38A .
~llfussa .
tr .
r. tr .
f. r.r. 38B .
Fimg , [ a , c.c. tr .
f. 38c .
Cyclolites .
tr .
tr.f .
Comparison of tr~Lusverse sections ( diagrammatic , cf. fig. 18 ) through three characte1\#183 ; istic septal type : , ; ; Mussa , the septal ridges ( r. ) are elongate , and the centres of calcification ( c.c. ) widely spaced ; Fungia , ridges more elliptical in outline , and the centres ( c. c. ) more closely packed ; Oyclolites , ridges rounded , centres ( c.c. ) in one group , so that six fascicles ( f. ) form here a " compound trabecula .
" of trabecu1re composes the ridges in both cases .
The trabeculre in Fungia are , however , arranged more as a parallel group , only the lateral trabeculre diverging to any extent sideways .
In fig. 38 nine or ten trabecuhe are represented in a single ridg\lt ; ~ of Jllussa and Fungia , and the fascicles of fibro-crystals are bilaterally arranged with reference to a row of calcification centres ( ideal axes ) in the median plane of the septal ridge .
For comparison , the ridge of Cyclolites is given .
This genus is a typical Thamna~trreid , whose septa has a fenestrated porous structure ( fig. 69 ) .
The single " trahecula " ( PRATZ ) is the homologue of a trabecular ridge .
Several fascicles are 174 DR MARIA M. OGILVIE ON 'l'HE MICROSCOPIC AND present , but they are radially a.rrnnged around one main axis , and the ridge assumes rounded contours .
The trabecular structure can be demonstrated in Fung-ia without the use of the microscope .
\#183 ; when a septum is warmed and then thrown suddenly into cold water , the ca1careous laminre flake off in various planes .
The removal of the superficial lamime sho , vs that what appeared a rough , granulate ridge on the surface , is in reality a symmetrical group of traheculre .
The individual trabeculre are marked out by slight grooves from one another .
Successive groups of trabeculre are present , each of which is precisely homologous with the growth-segment of a sAptnm of Galaxea or a " ridge " of Mussa ( figs. 18A , 18B ) .
The only differences a.re , 1st , that the length of the group in Fungia is greater relatively to the breadth ; and , ~nd , that some six or seven of the trabeculre in Fungi\#183 ; a-those in the middle of the whole group-are paralle1 in direction .
Lateral trabeculre bend out somewhat obliquely , and form the means of coalescence with those of adjacent ridges .
The number of trabeculre in one ridge of Fungia is from eight to twelve .
The grouping of the trabecuhe in Fungia is intermediate between that which is observed in Mussa and that characteristic of the genera belonging to the Pseudoastrreinre , PRATZ .
In the latter , each septal ridge is grooved in the manner described for Fungia , but the trabeculre are all parallel in direction .
None bend out obliquely to meet the trabeculre of adjacent ridges ( cf PRATz'figures , 'Palreontographica , ' Zoe.cit .
, 1882 , plate 27 ) .
The absence of obliquely-diverging trabeculre is the cause of the porosity of the septum in those types ( ex. , Cyclolites , Microsofona , \amp ; c. ) .
On the other hand , it is the presence of diverging trabeculre in Fungia and Siderastrma which renders their septa compact like the septum of Astneids .
The very thin septa of Fungia are , however , said to be sometimes porous , apparently due to less regular development of the diverging trabeculre .
But in all the specimens I happened to examine , a.11 the septa were compact .
I represent the transitional stages shown by Mussa , Fungia , Cyclolites diagrammatically in transverse section ( fig. 3 8 ) .
The group of trabecnlm in Cyclolites makes up one rounded , nodulose septal ridge , and it was to this ridge that PHATZ applied the term " trabecula .
" It may be looked upon as a compound large trabecula ( 'vicle antea , p. 141 ) , formed round several centres of calcification massed together in the median plane .
The radial symmetry of the calcareous fibres in the compound trabecula , is distinctive from the bilateral symmetry of fibres in a simple trabecula .
The nodular swellings , with more or less circular outline , in the course of a Cyclolites trabecula , correspond with successive growth-periods , each " knot " representing one group of radiating fibres around crowded centres .
Fungia and Mussa show similar groups , but the centres are not massed together .
They are disposed in a row , and the fibres form a broad ridge with elliptical outline .
Comparison of the Cyclolites " trabecula " with the large compound trabeculre in Heliastrma , ( cf figs. 24 , 26 ) , or in Montlivaltia , shows that it fundamentally agrees SYSTEMATIC STUDY OF THE MADREPORA .
RIAN TYPES OF CORALS .
175 with them in structure .
It will be readily understood that the broadening out of a " ridge " is correlative with compactness of a septum , whereas its narrowing to the nodulose " Cyclolites " form leads to a porous condition of the septum between its ridges .
Microscopic Sections .
The fundamental facts of the fibrous structure are the same as in Astrreidre .
The fibres are laid down in fine growth-lamellre .
The cut edges of growth-bmAllre are well seen iu fig. 39 , a finely-polished surface slab of part of a Fungia septum , under Fig. 39 .
g.l. f.c. Section horizoutal with the septal surface .
g.l , wavy " dark bands " of the growtb-lamellre .
f.c. , fibro-crystals of the la'lrnllre crossing the " bands " in direction .
high magnifying power .
The uneven polishing has exposed successive lamellre and their fibres .
All the fascicles in Fungia take origin in the median plane of the ridge at the ideal axes of the several trabeculre .
They grow upwards for a short distan~e and thAn bend sharply outwards towards granulations on the surfaces of the trabeculre .
That this is the course followed may be deduced from comparing sections made longitudinally , in radial and in tangential direction .
Fig. 40c shows ( 11 times magnified ) a longitudinal radial section in which one septum has been cut for the greater part near the median plane ; but towards the synapticulre ( i.e. , on the right hand of the figure ) the section has passed obliquely towards the septal surface and the synapticnlre .
Fig. 40D give3 another part of the same section where it had passed entirely through the first septum and the synapticulre and entered the adjacent face of the next septum .
The direction of the section through the two adjacent septa is shown in figs. 40A and 40B .
176 DR. MARIA lYl .
OGILVIE ON 'l'HE MICROSCOPIC AND Both of the figs. 40c and 40D show that , wherever surface planes of the septum arc met , or the planes immertiately subjacent , the foscicles are cut almost perpendicularly to their axes .
It may , therefore , be deduced that in those planes the axes Fig. 40.1. .
Fig. 40B .
Small drawing to ~how the positions of the scctio1 : s rrproseuted in fig. 40c ( " x " in drawing ) and fig. 40n ( = " y " ) .
Smn,11 diagrammatic drawing to show how the plane of polishing in fig. 40A passed thi\#183 ; ough two adjoining septa and the synaptioulro between , Fig. 40c .
Section passing obliquely through a septum into the interseptal loculus with synapticular bars ( syn .
) .
Four growth-segments ( g. 0 .
) are indicated by fans of diverging trabecular ridges ( i\#183 ; .
) , a single " syn .
" corresponding to each .
Fig. 40D , r. f. Same section continued through the interseptal loculus to the adjacent septum f. = fascicles cut across .
of the fascicles bend almost directly towards thA surface .
On the other hand , in the inner planes of the septum , the axes of the fascicles are met lengthways for some distance .
They produce here the appearance ( already familiar in :figures of Galaxea , SYSTE : MATIU STUDY OF THE M.ADREPORA .
RIA .
N TYP : H]S Ol !
' conA .
L8 .
177 \amp ; c. ) of a more or less continuous line , representing the obliquely-cut axes of successive fascicles .
We have in this fact a proof that the fascicles develop at first in an obliquely upward and outward direction with reference to the middle line or " ideal axis " of an individual trabecula ( cf. antea , fig. 25 , p. 143 ) .
A tangential section ( 18 times enlarged ) is drawn in fig. 41 , including three septa and two interseptal loculi , occupied in part by synapticulm .
Passing into the synapticulm are fascicles which have their origin in the septum , at a plane near the septal surface .
Those fascicles may either not grow sufficiently beyond the surface , or may Fig. 41 .
s. syn .
s. syn .
s. Tangential section of Fungia .
s. = septa ; syn .
= synapticulre .
Independent centres of calcification are present in the synapticulre .
not happen to meet an oppositely placed fascicle .
In either case , one or more new fascicles develop in the interseptal loculus , and form a means of coalescence between the fascicles growing out from the septum .
The new fascicles are constituted precisely like the septal fascicles and bear evidence , therefore , of their origin in a , specially invaginated pouch of ectoderm .
As growth advances , new interseptal fascicles are formed in the same vertical line as the foregoing , and thus the small invagination of the basal body-wall , in which the first interseptal fascicles developed , lengthens out until the body-wall becomes invaginated over the whole , long synapticular bar .
The continuity of the synapticu1ar bar is in a direction at right angles to the direction of apical growth of the septum ( cf fig. 119 ) .
In a tangential section of the Heliastrma septa ( antea , fig. 27 ) , and again m tangential sections given , by PRATZ , of Pseudo-Astrreinrn genera , pseudosynapticulre and dissepiments are present .
It will also be frequently seen that , where the fascicular outgrowths from septa do not succeed in bridging the interseptal loculus , a small dissepiment is laid down between them by the basal body-wall , as it withdraws from lower to highe1\#183 ; hor\#183 ; izons of the corallurn .
But the basal body-wall in Fungia 11 , ever withdraws itself completely from the lower horizons ; it forms instead these extra , cylindrical invaginations of the body-wall between septal invaginations .
To this ingenious device are mainly due the large proportions which the discoid corallum may obtain , as well as the characteristic and peculiar skeletal feature of the series of synapticular bars in the interseptal loculi .
The Growth Relations of Fungia .
A microscopic surface section through granulations shows that the corresponding fasc icles have a diameter varying from \#183 ; J6 millim. to \#183 ; 25 millim. , and on the synapticulre the fascicles have sometimes a very large diameter , measuring as much as .35 millim. The fascicles of Fungia have , therefore , throughout the corallum a much larger size than those of the Astrreid types , which have been studied above .
Probably the fascicles thus measured represent coalesced fascicles from the central and parallel placed tmbeculre in a ridge .
At the exsert apical portion of the septum , where the additional height ( apparently length ) attained during a growth-period is always greatest , the measurements in Fungia gives fairly definite results .
The growth-segment of a single ridge measures , then , from 1 \#183 ; 2 millim. to 1 \#183 ; 5 millim. in height , the maximum width at the surface of the same growth-segment is 1 \#183 ; 5 millitn .
, its diametric thickness at a nonsynapticular portion a little over 1 millim. at synapticular portion 1 \#183 ; 6 millim. Hence the growth-segment may be regarded as fairly cubical in dimensions , although its actual form is notched at the base , and rounded or pointed at the apex .
The synapticular thickening is at the bases of adjacent growth-segments , hence the curve of any one synapticular bar bends convexly , so as to be parallel with the exsert apical portion of the septum .
In this relation to the growing edge , it simulates the dissepimental floor of the Astrreid polyp .
Compare the last-formed synapticular bar with the exsert margin of the septum in fig. 37 , and the same relations will be observed as those shown by the last-formed dissepimental floor and the septal ma.rgin in figs. 2 , 15 , 22 .
SIDERASTRJEA , The Presence of Synapticulm and the Lateral Coalescence of Septa within the Calyx .
The genus Siderastrcea , although in outward appe,1rance resembling certain colonyfol'miug Astra ; ids , has been ascertained to be like the Fungidre in its tentacles and SYSTEMATIC STUDY OF THE M.A.DREPORARJAN TYPES OF CORA .
LS .
170 soft parts.* A careful examination of the micrOf:icopic skeletal structures would suffice to determine this relationsl:ip .
The calyces are small , united with one another , pseudothecal , have compact radial structures , and a paliform pseudocolurnella , in so far having a general resemblance , therefore , to Goniastrma ( cf below , fig. 42 ) .
But while Goniastrma has dissepimental basal structures and a wall formed by septal thickening at the area of divergence , Siderctstrma has well-developed synapticulm representing both basal and tangential structures .
Dissepiments are also present here and there , but they have not in Siderastrma the definite growth-relations which they have in Astrreids .
On the other hand , the synapticulre are an essential part of every growth-segment , in the same way as demonstrated above in the case of Fungia .
Again , in the inner parts of the calyces , where the septa approach each other more closely , their surfaces actually coalesce ( cf fig. 43A ) , and groups of septa are formed .
The same coalescence takes place in Fimgict and in the Lophoserids , and is a marked feature , above all , in the Eupsammidre .
The Astrreidre , on the other hand , never show marked septal grouping , but only an irregular columellar rneshwork of septal ends in genera .
like Rhabdophyllia , with " spongiose columella .
" The direct lateral coalescence of the septal surfaces belonging to different septal cycles is homologous with the synapticular union of septa .
It is , moreover , a skeletal featnre specially characteristic of synapticulate Madreporarian types .
It may be said in opposition to .this that septal grouping occurs also in some of the Turbinolidre .
Later parts of this paper will show that the Turbinolidre present in their calycinal features a striking resemblance with the Madreporaria perforata , E.H. ; and the lateral , quasisynapticulate coalescence of septa to form Eupsammid-looking septal groups is not one of the least of their features of resemblance .
Fig. 42A is drawn as if looking into the interior of one calyx of Siderastrma and the adjoining halves of two other calyces , seen in longitudinal section .
the septal surfaces are exposed and show parallel ridges on which large synapticmlar granulations mark the successive growth-p:wts of each ridge .
The teeth of the ridges are prominent ; their shape is broad and blunt-edged , like the teeth of the Pseudo-Astrreinre and of Fungia .
Comparing this figure with fig. 37 , a septum of Fungia , the direction of the ridges will be seen to be the same .
In Fungia they di verge from an area in the septum , in Siderastrma the area of divergence is at the wall between two calyce\amp ; .
_The grauulations belonging to any one septal segment in Fungia are those which unite and form a continuous bar on the surface of the septum .
The latter runs therefore in a direction at right angles to the ridges .
In the Palreozoic genus Heliophyllum the * DUNCAN refors to the researches of AGASSIZ and PouRTALES on thi~ genus ; vide " Revision of the Families and Genera of the Madreporaria , " p. 184 .
The same author rightly separates the genrn , Siderastrcea from the two genera Astrceomorpha and Mesomorpha , with which PRATZ had united it under his group of Pseudoast , rreinoo Irregulares .
" carime " on the septal surfaces are structurally homologous with the bars on Fungia , but adjacent bars do not meet across the interseptal loculus in Heliopyllwrn .
In short , the carin::e do not act as synapticulre .
Also in a few Jurassic genera carinre are present which do not form synapticular union of septa .
Fig. 42A .
r-alym .
ps.th .
ps.c .
ps.th .
Skeletal parts of Siderastrroa ps.c .
, pseudocolumella of calyx , ps.th .
, pseudotheca between calyces .
Fig. 42n .
r. Septal surface showing the direction of the ridges ( r ) , and the position of the large Rynapticular granulations ( syn .
) .
In Side1'astr{Ba , there is scarcely any tendency for the granulations to unite across .
the ridgAs of the same septum .
Only at the wall and the centre of the calyx occasional short bars are formed on the surface of neighbouring ridges .
On the other hand , the granulations of adjacent septal surfaces meet in the interseptal loculi and form characteristic vertical series of synapticulre .
This is the typical fo~m of synapticulre , where large granula , te eminences meet at inte1\#183 ; vals .
The continitous synapticular bars or carinre in Fungia are quite exceptional .
Towards the centre of the calyx , where the septa stand very close to one another , direct lateral coalescence of the septa may be said to take place , as the granulations are scarcely perceptible .
Where the distance between adjacent septa becomes greater , i e. , nearer the periphery , the granulations of adjacent surfaces meet without the intervention of a new " synapticular " centre of calcification .
This particular form of synapticular coalescence is called " pseudosynapticular , " and bas already been mentioned as an occasional occurrence in the calyces of Heliastr{Ba .
At the periphery of the Siderastrma calyces where the septa are most distant from one another , the introduction of new synapticular centres of calcification is the rule .
The granulations on the septal surfaces are in short seldom sufficiently developed to bridge the whole space at the mural part of \#183 ; the calyces , and new invaginations of the aboral wall between them give rise to new centres of calcification .
The resulting skeletal bridge is called a " true synapticula , " and is the homologue of the synapticular bars in Fungia .
The transverse section ( 6g .
43 ) of a few septa of Siderastr(JJa shows the various forms of interseptal skeletal support just enumerated ; ( 1 ) lateral coalescence of ' septa , SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
181 ( 2 ) pseudosynaptieular , ( 3 ) synapticular .
In all cases the morphological significance is the same .
It is the building up of a basal support leaving canalicular spaces for downward prolongations of the visceral cavity of the polyp .
In this respect , SideFig .
43A .
CO. Se .
syn .
Transverse section through calyces of Siderastrcoa .
The synapticular " wall , " or pseudotheca ( ps.th .
) is seen between three adjacent calyces : syn .
synapticula ; ps.syn .
pseudosynapticula ; co.se .
lateral coalescence of septa , s. septa .
ps.syn .
Fig. 43B .
ps.th .
8 , s. syn .
s. ps.syn .
s. Fig. 48 r .
r.r. er .
gr. gr. One septum ( high power ) .
gr. granulations ; 8Y11 .
er .
a crack in the section along the mediii , n plane .
Transverse section ; enh , rged drawing of a pseudothecal part uf the same aection as fig. 43A , rastrcea is like Fimgia , and stands in sharp contrast with As.trreids in which the basal support is laid down as a continuous closed platform of dissepiments .
The same section illustrates the microscopic structure of the septa of Siderastnea .
They are composed of a number of trabeculre , in which the fibres are radially arranged around a central axis .
Fig. 43u is a more highly-enlarged septum , showing the t endency of the septa to split along the median planes .
_ When dust accumulates in such an artificial crack , it naturally presents all the appearances of a " dark line .
" This drawing indicates also the cut edges of the series of growth-lamellre as lines running across the radiating calcareous fibres .
Now and then two centres of calcification are present diverging from one another in the median plane .
Then the trabecnla appears donble and is elongated in outline .
The individual trabeculre of Siclerastrcea may\#183 ; be compared with large trabeculre , showing radial symmetry of :fibres in H eliastrcea ( cf :fig .
24A ) , or with the central closely-parallel group of trabeculre in a septal ridge of Fungia ( cf fig. 40 ) .
Fig. 44 .
gr. ps.syn .
Longitudinal section parallel with the septal plane ; pseudosynapticulre ( ps.syn .
) , au enlarged form 0\#163 ; granulation ( gr. ) .
The trabeculre , in any one ridge of Siclerastrcea , may , as in Fungia , be placed obliquely and diverge from one another in the plane of the septum .
This it is which makes the septum both in Fungia and Siderastrcea typically compact , distinguishing them from the septa in the Pseudoastrreinre , Lophoserinre , and Eupsamminre .
In the radial group of fibres ( transverse section , fig. 43c ) , correspon\lt ; ling to individual trabeculre in Siderastra ; a , smaller fascicles of fibres are often distinguishable .
These have the same relation to the large granulations on the lateral surfaces of the septa as I have described above for Fungici .
Fig. 44 is a microscopic view of a fairly thick surface preparation of a Siclerastrcect septum .
The emergent ends of fascicles are seen to coalesce in quite indefinite number at the granulate ( pseudosynapticular ) parts of the septum .
If this view of microscopic structure be compared with figs. 42A and 42.B , it will be seen at once that the series of granulations follow the individual septa ] ridges .
Fig. 44 shows the microscopic structure of granulations and of pseudosynapticulre formed by the coalescence of several fascicles .
As I already took occasion to mention in the historical account ( antea , Introd .
, p. 103 ) , PRATZ observed a certain indication of smaller groups of fibres(= fascicles ) SYSTEMATIC S'l'UDY OF THE .MADREPORARJ .
AN TYPES OF CORALS .
183 passing outwards from the axis of the trabeculre in Cyclolites and allied genera .
He thought they possibly corresponded to the small surface granulations which he called nodules ( " Knotchen " )-a few small granulations uniting to form the " pseudosynapticula .
" PRATZ , however , made no farther research in this direction , and , therefore , failed to recognize the full significance of the successive " nodular " groups as the individual growth-parts of the trabeculce representing successive growth-periods of the polyp ( an tea , p. 15 6 ) .
When that general fact is recognized we can follow the trabecular growth-part in its various forms , narrow and bilaterally symmetrical , large and radially symmetrical , throughout every great group of the Madreporaria .
In Siderastrcea the average measurements of a trabecular growth-part are as follows : Height ( or length ) Width at surface .
Diametric thickness = \#183 ; 22 to \#183 ; 2s millim. = \#183 ; 11 , , '14 , , = \#183 ; 10 , , \#183 ; 16 " Classificatory Relationship of Siderastrcea .
The microscopic structure of a single trabecula is , as we have seen , identical in the genus Siderastrcea , and in genera belonging to the sub-family Pseudoastrreime , PRATZ , Fungia on the other hand shows secondary thickening in the form of new centres of calcification developed on the side surfaces of the septa in connection with the synapticular granulations .
This is , however , a feature correlated with the low form of the Fungia corallum , and the fact that the soft tissues remain in the canalicular portions of the ccelenteron throughout the life of the polyp ; it is , therefore , not a feature in itself of classificatory distinction between Fungia and Siderastrcea .
In spite of the dissimilarity in the form of the corallum in those two genera , the agreement in all essential structural features is undoubted .
Recapitulating briefly , the septa both of Fungia* and Siderastrcea are typically compact , and have granulate surfaces , the miscroscopic structure showing that the septa are composed of trabeculre which may diverge from one another in the space of a single septal ridge ; whereas , in the Pseudoastrreinre , the trabeculre composing a septal ridge are absolutely parallel throughout the whole course of that ridge .
The only mode by which adjacent ridges can coalesce in the latter case is by means of elongation of the fibres in the septal plane .
This elongation takes place at regular intervals between successive growth-periods in the Pseudoastrreinre , and the result is a series of regular pores in the septum .
Secondary calcareous deposits may afterwards fill up the septal pores .
Fungia and Siderastrcea , by means of obliquely set t.rabeculre , form septal segments which actually grow compactly .
* DUNCAN notes that the septa of the ]1igher orders of Fungia are often perforate ; I have seldom seen apertures , they are at any rate never present in the reguliir series characteristic of Pseudoastrwinw .
184 DR. MARIA M. OGILVIE ON THE :MICROSCOPlU ANU , while this difference in the septa } structure is the chief feature distinguishing the two genera under discussion from the Pseudoastrreinre , there is also the second distinction , which PRATZ had observed in his classification of Pseudoastrreinre into minor groups the presen\lt ; le or absence of new centres of calcification to bring about the coalescence of the granulations across septa loculi .
PRATZ noted in SiclerastrC13a that coalescence took place sometimes with intervention of new centres of calcification , sometimes without ; in the characteristic Pseudoastrreime types ( Pseudoastrreinre regulares = allies of ThamnastrC13a and Microsolena ) coalescence in all cases took place directly between granulations , the latter form of coalescence PnATZ called Pseudosynapticular , the former Synapticular .
The fine structure of Fungia was unknown to PRATZ , but it is now clear from the sections given above that SiderastrC13a agrees with Fimgia and differs from the Thamnastrreids and Microsolenids in \#183 ; respect of the coalescence .
In other words , that the basal body-wall of SiderastrC13ct , like that of Fungia , may be specially invaginated between the septa .
Fascicles of fibres are then formed within the invagination similar to the fascicles formed in every small " pit " of the septal invagination of the body-wall .
Those two features lead me to classify SiderastrC13a rathe1\#183 ; in the immediate neighbou .
rhoocl of Fungia than in that of the PseucloastrC13inC13 .
My result differs here both from PRATZ and from DmrnAN ; the latter may indeed be said to have been influenced directly by PnATZ .
DuNCAN 's family of " Plesiofungidce " includes the genera which PRATZ examined and referred to a sub-family , Pseudoastrreinre regulares and irregulares , the only change made by DUNCAN being to break up the sub-group " Pseudoastrreime irregulares .
" He still left , however , the three genera belonging to the sub-group in the immediate relationship of the main group .
The three genera are SiclerastrC13a , AstrC13omorplw , Jl !
Iesomo1JJha .
Siderastrcea is , as I have said , nearly related to Fitngia .
Astrceomorpha and Mesomorpha are both fossil genera , and most probably are related with the Maclreporaria perforata , E.H. They are dealt with in Part II .
, chapter v. of this paper .
True and False SynapticulC13 ( PRATz ) .
The original definition of the term is quoted above ( " Fungict " foot-note , p , 167 ) , and may be seen to refer only to the macroscopic appearance of the coalescence of opposite granulations .
This definition held good until PRATZ made the distinction based upon microscopic investigation of a true synapticula composed of one or more centres of calcification , and a false synapticula in which no new centre of calcification was present , but the fibrous " nodules " of the septa were merely elongated .
Since EDw ARDS and HAIME had defined " synapticula " as an elongation of the granulations ( = " nodules " ) , PRA'rz ' " false synapticula " corresponded to the actual words of EDw ARDS and HAnrn 's definition .
But the " sense " of the original definition was niade clear by Emv ARDS and HAIME in tlieir application of the term in the first place SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
1 85 to the synapticulre of Fungia .
And accepting the " sense " of the original definition and its application by those authors to Fungids , Lophoserids , \amp ; c. , PRATZ invented the term " pseudosynapticulre , " for the manner of coalescence which he observed throughout a number of genera , and which did not agree , microscopically considered , with the manner of coalescence typical of Fungids , Lophoserids , \amp ; c. From that time forward there has reigned a kind of scepticism in the literature with regard to the distinction made by PRATZ and expressed in his terms of " True Synapticula " = " Synapticula " ( E.H. ) , and " False Synapticula " = a microscopic appearance not characteristic of the " synapticular bars " of Fungia which EDWARDS and HAIME had quoted as an example of their use of the term Synapticula .
Although PRATZ himself must have been acquainted with the microscopic structure of the " true synapticula , " he referred to it only in a few passing words in his work , and gave no figure to demonstrate any difference .
Neither was his explanation of the " true synapticula " very clear .
He said a process grew out singly from one septum towards the adjacent side of the neighbouring septum and fused with it , new centres of calcification having built up the calcareous process in the course of its interseptal extension .
This is partly right , such synapticulre occur very often in perforate corals , Eupsammidre , Poritidre , \amp ; c. , but as it happens , the synapticulre of Pungia are quite different , they are formed partly as septal outgrowths , partly as an interseptal continuously-growing calcareous ba1 ' .
ORTMANN did not agree with PRATZ in recognizing a structural difference between true and false synapticulre .
He writes :- " Special centres of calcification do not exist for the synapticulre , hence there are no ' true ' synapticulre \#183 ; in the sense described by PRATZ , and all structures which have hitherto beeri called synapticulre must , according to PRATZ himself , be termed synapticuhe.''* I find , however , that it is highly convenient to have the distinctive term of pseudosynapticula where granulations meet directly .
" True " synapticuhe have various forms of development .
In Eupsammia they attain an independence of development , and a morphological significrmce still greater than they present in Fungict and Siderastrma .
CHAPTER V.-'l'HE FAMILY FUNGIDJE .
PRATZ ' DISPOSITION OF SUB-FAMILIES .
The detailed microscopic investigations carried out by PRATZ on early Jurassic representatives of the family of Fungidre led that author to draw up a table of classificatory features for the various sub-families .
His sub-divisions are Pseudo* ORTMANN , " Die systematische Stellung einiger fossiler Korallengattnngen und Versut ; h einer phylogeneti , ; chen Ableitung der lebenden Steinkorallcn , " 'Neues Jahrb .
fiir Mineral .
u. Geol .
, ' 1887 , II .
, p. 185 .
l\{DCCCXCVI.-l3 , 2 13 astrreime , Pseudoagaricinre , Funginre , Lophoserinre ( Agaricinre ) , Merulininre .
* He bases the c1assificatory features of the three last-named sub-fami1ies upon the data of previous authors ( DANA , VERRILL , EDWARDS and HAIME ) .
The first two subfamilies are , however , treated from what may be called an original standpoint in the classification of fossil corals , and embody the main results of\#183 ; the microscopic and morphological research made by PRATZ himsel\#163 ; Although , as I have said , PRATZ took Jurassic genera as the basis of work , he included also Triassic , Cretaceous , Tertiary , and recent genera in smaller number .
The genera Coscinar(J3a and Siderastr(J3a are indeed the only two recent genera included .
For the features of Coscinar(J3a he relied upon information given him by KLUNZINGER ; while Siderastr(J3a received but a brief examination and was not directly compared with Fungidre and with Fungia .
Again , in accordan\#183 ; ce with the then generally accepted idea that Madreporaria aporosa , E.H. , had little connection with M. perforata , E.H. , PRATZ nowhere institutes a comparison of any type , fossil or recent , with the Eupsamminre .
On the other hand , the trabeculate porous character of the septa in the Poritinre induced him , without any microscopic or closer investigation of typical genera , to place the sub-family Poritinre , E.H. , under his sub-family Pseudoastrreinre , together with Microsolena and other fossil members called Pseudoastrreinre regulares .
I mention this last fact because it is possible it may ha , ve influenced NEUMAYR in writing of " Microsolena , " as building a " bridge to the Poritidre " ( cf below , p. 306 ) .
Proposed Emendation of the Sub-family Pseudoastr(J3in(J3 , PRA .
TZ .
As the microscopic slides of PRA .
TZ were placed at my disposal , I feel in a position to review the table of classification which he founded upon them .
The only modifications which strike me as advisable concern the recent genera , and one or two fossil genera which my examination of recent Eupsamminre leads me to regard as more nearly allied with Eupsamminre than with Pseudoastrreinre .
These fossil genera are classified by PRATZ as Pseudoastrreinre irregulares , and along with them he places the recent Coscinar(J3a .
The type genus Haplarma ( Jur .
) , shows microscopica11y and macroscopically marked similarity with the genus Eupsammia ( , f transverse section , fig. 45 ) , Diplar(J3a is allied with Haplar(J3a ( fig. 45 ) , Meandrarma has , on the other hand , the character of Lophoserinoo ( this genus will be fully diagnosed in my " Monograph of the Stram berg Coral Fauna " ) , while Coscinar(J3a has most in common with Eupsammidre .
On these considerations I propose to annul the Pseudoastrreinre irregulares , PRATZ , * The general features of Merulina seem to agree with those of the Fungidre , KLUNZINGER groups it as an Astrreid in his sub-family Meandrinime .
It is possibly better to include it under the Funginre on account of the porous , echiuulate basal surface as well are the occurrence of true synapticulre in the interseptal loculi ; a more detailed examination is necessary before its systematic position can be stated with certainty .
SYSTEM .
A.TIC STUDY ' OF THE MADREPORARIAN 'rYPES OF CORALS .
i87 and to recognize the Pseudoastrreinre regulares , PRATZ , consequently as the complete sub-family .
I exclude from this sub-family the genus Tlimnnarcea and the Poritinm , E.H. , both of which had been included by PRATZ .
l !
'ig .
45 .
ps .
th .
syn .
'l'ransverse section of Hapla , reea , showing spongiose co1umella , septa with pores ( p. ) and united 1atera11y by true synapticu1oo ( syn .
) , also a pseudotheca1 wall with pores ( ps .
th .
) .
Comparison of Pseucloastrminm , PRATZ ( = Tha11inastrceince ) , with the Astrceidce .
A typical septum representing PRATz ' sub-family is given below ( fig. 69 ) ; a comparison of a transverse section of the con1ponent trabeculre with Fimgia and Mussa , has been already made ( fig. 38 ) .
But the nearer details must be referred to in PRATz ' work ( ' Paheontographica , ' 1882 ) .
The characteristic structural features common to Astrreinm , E.H. ( = Astrreidre , antea , p. 163 ) , and Pseudoastrreinre may be enumerated as follows :(1 .
) In each radial structure there is a " line of divergence , " from which individual trabeculro pass obliquely upwards in opposite directions--inwards and upwards towards the centr8 of the calyx , outwards and upwards away from the centre of the calyx .
( 2 .
) This line of divergence coincides with the position of the peripheral thickening which gives rise to the " wall " of the calyx .
The thickening is pai-tly brought about by outgrowths from the septa , part .
lie by oblique and vertical placing of inter-radial structures .
( 3 .
) The latter are dissepimental and pseudosynapticular in character .
( 4 .
) The septum and costa ( " radial structure " ) is composed of trabeculre , whose microscopic structure is like that described for Goniastrcea , \amp ; c. , above .
Coalescent fascicles of fibres form a large granulation , corresponding to each growth-part of the trabecular ridge .
These large granulations , by meeting adjacent granulations from neighbouring septa , form a series of " pseudosynapticular " bridges across the interseptal loculus .
Occasional pseudosynapticular union is present in Hdiastrcea and other Astrreidre .
Differences from Astrreidre :-(1 .
) In the microscopic constitution of the individual growth-part .
The neighbouring trabecu]re of a single septum do not grow upwards in contact ]88 DR. MARIA M. OGILVIE ON TH1\lt ; ; MICROSCOJ\gt ; IC AND with one another throughout a growth-period .
In other words , not all the calcareous fibres of a fascicle have a fairly equal length , but some grow much longer th.an others , and only the long fibres project suffic , iently to meet those of the trabecular growthparts on either side in the septal plane .
In this way , practically a circular granulation is formed , projecting outwards all round the trabecular growth-part .
Sometimes the pr~jection may be at the upper part , or may wind spirally round the segment .
In some cases the fascicles are su:ffi.ciently large to retain distinctness during the development of the granulate projection , and one can see that the latter is made up of a ring of smaller granulations : each of which corresponds to the emergent end of a single fascicle ( ex. Cyclolites , figs. 3 8 and 6 9 ) .
Those are the " 1\#183 ; egularly grouped calcareous nodules , " of which , as PRA'l'Z says , the trabecula is composed : and he identifies the " nodules " ( " Knotehen , " PRA'l'Z ) with the " sclerenchymatous_ nodule " of EDWARDS and HAIME .
It is clear that in the expression of EDWARDS and HAIME the whole growth-part was meant .
The tiny " knobs " on the surface of a nodule are identical with the single fascicles , two or more of which have been microscopieally proved by me to compose a " growth-part " in all types of Astrreidre and Fungidre discussed above .
The unequal elongation of the fibres in the fascicles of Pseudoastrreinre genera produces the characteristic bush-like appearance of the fibres in longitudinal sections of trabeculre , each bush corresponding to the part added during a single growthperiod .
The group of faseicles in Astrreinre has a more elongated axis and equal development of fibres for a single growth-part .
As a result of the different mode of growth , the septa in Astrreidre are compact , or seldom leave a space between trabeculre , while those of Pseudoastrreinm are originally fenestrated .
Secondary compacting tissue may fill up the original fenestrations left between adjacent trabecular growthparts .
( 2 .
) Difference in the regnlcir_ity of the pseudosynapticulm in interseptal loculi .
The granulate ring of each growth-part which projects into the interseptal loculus is developed so uniformly along the whole horizontal series of growth-parts corresponding to a growth-period , that they tend sometimes to coalesce as continuous bars along the septal surface .
Another still more frequent tendency is for the adjacent granulate projections to meet and give rise t0 a vertical series of interseptal uniting-rods , here pseudosynapticu lee .
Connection of Pse1tdoastrminm with Fungidm .
The features of difference from AstraJinm are precisely those which Justify the systematic position of the Pseudoastrminm , PRATZ , as an ally of the Funginm and Loplioserince , E.H. , and a , sub-fitmily of the Fungidce .
Although later deposit renders the septa in Lophoserinre usually compact , they are for the most part laid down as a series of non-compacted trabecular growth-parts .
SYSTEMATIC STU .
DIE OF 'rHE :VIADREPOR .
ABIAN TYP .
F~S OF CORALS .
189 The same is sometimes true of the smaller septa in Fungia and in a general way of the Eupsammidre septa .
The tendency to the formation of continuous bars along the septal surface is especially characteristic of Fungia as well as of genera in other sub-families of Fungidre .
The interseptal .
uniting-rods in these three sub-families may be produced as in .
Pseudoastrreinre by prolongation of the fascicular fibres belonging to septal growthsegments , or they may be due to the insertion of new fascicles .
In the latter case , as I have pointed out a.hove , they are formed in special invaginations of the aboml body wall and are " true synapticuhe .
" It would seem , therefore , that the fossil sub-family Pseudoastrreinre has a microscopic structure of skeletal parts varying little from that of Astrreinre , but these variations indicate that the living polyp had an arrangement of its body-wall and i~ternal tissues like that of recent Fungidre .
There is no recent representative of the Pseudoastrreinre .
The period of the highest development of the sub-family was Up .
Jurassic : a period in which the recent Fungids , Lophoserids , and Eupsammids were still , so to speak , in the nursery of their existence .
FRECH 's alte1\#183 ; ation of the name of Pseudoastrceinw to Thamnastrceinw .
FRECH recently objected to the names of Pseudoastrreinre and Pseudoagaricinre given by PRATZ , and altered them to Thamnastrreime and Astrreomorphinre respec : .
tively , at the same time referring them to a special family Thamnastrreidro instead of Fungidre , under which PRATZ had placed them ( " Kor .
Fauna der Trias .
, " 'Palaeont .
, ' vol. 37 , 1890 , p. 59 ) .
Both the names Thamnastrreidre and Thamnastrreinre had already been used in the literature , the former by REUSS in 1864 , the latter by BECKER in 1875 .
FRECH made no comment on previous use ; he accepted PRATZ ' systematic grouping , changing only the nomenclature .
The only point wherein he mentions disagreement with PRATZ has regard to the distinction of .
true and false synapticulre .
He writes : " The ( true ) synapticulre have their origin by complete coalescence of pseudosynapticulre ; at the same time the anatomical difference is visible only in the case of relatively few very well preserved specimens .
According to ORTMANN , moreover , the said distinction is untenable , since a line of fusion ( Trennungsnaht ) is always present .
" FRECH , therefore , unites the Pseudoa.strreinre regulares and irregulares , PRATZ , under the title Thamnas~ trreinre , FRECH .
PRATZ , however , did not distinguish the two branch\#183 ; es of his sub-family on the basis of the synapticulre alone .
The main feature to which he drew attention in giving the adjectival names " Regulares " and " Irregulares " was the difference in the microscopic .
structitre of the septa .
He defined Pseudoastrreinre regulares and irregulares as follows :- . .
190 DR. MARIA M. OGILVIE ON TliE MICROSCOPIC AN !
J P. regulares , PRA'l'Z .
" Trabecula built up of regularly ( symmetrically ) grouped calcareous nodules coming into contact with each other at fairly regular distances , by means of which rows of pores are formed , running vertically to the septal edge , and more or less regularly distributed .
Neighbouring septal surfaces bound together by Pseudosynapticulce and traversre ( dissepiments ) .
" P. irregulares , PRATZ .
Trabeculre built up of numerous fine , calcareous nodules , grouped without symmetry , and united irregularly ( regellos ) with one another .
Intratrabecular spaces ( pores ) , therefora , of very different size and spacing ( Anordnung ) .
Neighbouring septa united by true synapticulre in addition to traversre ( dissepiments ) .
\#183 ; If now for the " calcareous nodule " in these definitions the fibrous fascicle be read , and each group of fascicles be understood to represent a distinct growth-part of a trabecular ridge , it will be ser-n that the definitions given by PRATZ have been fully corroborated by my work .
PRATz ' definition of P. irregulares gives , in fact , the main features of the Eupsammidce , by which they may be di'stinguished from Pseudoastrceince regulares ( cf. below , 'Table of Classification , ' p. 252 , " Successive " and " Contemporaneous Growth-parts " ) .
The distinctions so carefully worked out by PRATZ are , therefore , fully justified , and I regard it as a backward step to unite the two sub-groups under the one sub-family Thamnastrreinre .
Moreover , I have already shown that PRA'l'Z had right in distinguishing two kinds of synapticulre the pseudosynapticulre of the " Pseudoastrreinre regulares " and the true synapticulre present in Funginre , \amp ; c. Astrceomo17Jhince , FRECH .
The limits of FRECH 's sub-fami]y Astrreomorphinre coincides precisely with those of PRATz 's sub-family Pseudoagaricinre .
Uncler this sub-family the Triassic genus Astrceomorpha and the recent genus Siderastrcea were included by PRATZ ; two gen~ra which I do not regard as nearly related .
FRECH , at the same time , speaks of Astrceomorpha as the ancestral genus from which sprang three of the four genera belonging to the Triassic family Spongiomorphidre ( loc. cit. , p. 70 ) .
He writes:-- " The most peculiar type ( of the Spongiomorphidre ) is undoubtedly Stromatomoryha .
Precisely this genus is found already in the M1.ischelkalk together with Ast1\#183 ; ceomotpha , , from which last-named genus the three other genera ( Spongiomorpha , Heptastylopsis , Heptastylis ) are descended ( ' sich ableiten lassen ' ) .
" On a later page , however , there is more reserve in the author 's expressions : " The origin of the Spongiomorphidre from the Astrreomorphinre is in so far problematical , since both appear together for the first time in Mitschelkalk .
Probably both spring from the same unknown ancestral type " ( loc. cit. !
p. 73 ) .
SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
191 Fm::cH then gives a rliagram in which he shows the possible connection uf Actinacis as well as of the Perforata related with Turbinaria , with the Spongiomorphidm and Astrreomorphinre .
" The Turbinarias are descended from some form related to Heptastylopsis " ( a Spongiomorphid ) .
The general result which may be drawn from FRECH 's work is , therefore , that the Spongiomorphidre include the ancestral type of Turbinar\#183 ; ia , and are intimately related with the Astrreomorphinre , hence more distinctly with the companion sub-family Tharnnastrreinre , -these terms being used in PRATz 's sense of Pseudoagaricinre and Pseudoastrreidre respectively .
As PRATZ included Porites under the Pseudoastrreinre , we have a very large tree of relationship and descent , including most of the Madreporaria perforata , E.H. and the ancestral forms of the Fungidre , together with a family Spongiomorphidre of an entirely aberrant character .
With this result I entirely disagree .
My reasons will appear during the course of the present paper .
The genera closely related to Cyclolites , Thamnastrcea , and Microsolena are those , which properly belong to the sub-family Thamnastrreinre , and for them it is unnecessary to erect a special family Thamnastrreidre .
They are nothing else than the early Triassic representatives of our Fungidre .
The genus Siderastrcea has a very clear line of descent within this family .
On the other hand Astrceomorpha seems to me to come from an entirely different origin , and to have structurally more resemblance with Porites than with Siderastrcea or any other Fungid .
Lophoserince , E.H. The affinity of the Lophoserinre , E.H. , with the Thamnastrreinre ( BECKER , em , : aut .
) , and with the Funginre , E.H. , is fully recognized .
The considerable development of pseudosynapticulre is a feature of resemblance with the Thamnastrreinre .
But their distribution in the calyx and the occasional occurrence of true synapticulre is a point of difference .
Again , the basal surface of the coral1um has the same smooth or epithecate character in Lophoserinre as in Thamnastrreinre .
In the structural featureR of its septa , the Lophoserinre are intermediate between the Funginre and the Eupsammidre .
The septa may be compact and composed of well-developed septal ridges and regular trabeculre , as in Fungia and Siderastrcea .
Or it may have irregular pores , arising from interruption of the calcification around trabecular axes , as in the case of the Eupsammidre .
The centres of calcification are more closely and less regularly set in such types of Lophoserinre , and in many cases the opposite fascicles are zig-zagged in position .
Fig. 46 is a transverse section of adjoining calyces of Lophoseris , showing the close setting of the fibrous fascicles in the septa , the highly granulate outline of the septal surfaces , also the occurrence of both pseudosynapticulre and true synapticulre .
Fig. 47 is a transverse section made through a thickened pseudotheca , between two rows of calyces in a meandroid type .
A comparison of this microscopic section with fig. 52 , E'upsarnmia , shows the points of agreement in the crowded centres of deposit and pseudothecal thickening .
The figured specimen of Eupsarnmia was scarcely so well preserved as that of Lophoseris .
l !
'ig .
46 .
gr , ps.syn .
Transvesre section of a calyx of Lophoseris-true synapticulre ( syn .
) , and pseudosynapticulre ( ps.syn .
) .
form the pseudotheca ; -septa1 granulations ( gr. ) correspond with the " centres of calcification .
" Fig- .
47 .
ps.syn .
... syn .
Tra:\#181 ; sverse section through a much thickened pseudotheca , continued growth of the synapticulre and pseudosynapticular ~ranulations having blocked up a large part of the interseptal loculi , SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
193 CHAPTER VI.-EUPSAMMIA ; SYSTEMATIC POSITION OF THE EUPSAMMIDAJ EUPSAMMIA .
The Sub-kin , c ; dom M. Perforata , E. H. All the generic types previously described in this paper have been chosen from the sub-kingdom " Madreporaria Aporosa , " E.H. ; in E1tpsamrnia , however , we have a representative of the Madreporaria pe1forata , E.H.* The M. pe,1forata were subdivided by EDWARDS and HAIME into two families , the Madreporidce , in which the porosity of the calcareous skeleton was comparatively little developed , and the Poritidce with a highly reticulate skeleton .
Since their classification , published in " Cora11iaires , " various alterations have been suggested .
The three sub-families of the Madreporidre , the Eupsamminre , Turbinarinre , and Madreporinre , have especially been sorted and re-sorted .
One author groups Turbinarinre with Poritinre , another regards Turbinarinre as distinctly related to Madreporinre , but removed from Eupsamminre , the Poritinre themselves have been classified ( in part ) with the Thamnastrreidre .
Generally speaking , the Eupsammime have been held aloof from much controversy , at the same time that , by reason of repeated research of Astroicl( 's calycular\#163 ; s , t their features were becoming known in the greatest detail .
Canalicular System in the Interseptal Loculi .
The Eupsamminre have porous septa , a synapticular\#183 ; wall , and , in colonial types , a canaliculate crenenchyme .
I examined chiefly the skeletal structure of the genus Eupsarnmia .
Fig. 48 is a drawing of the septa !
surface of Eupsmnmia trochiformi\#183 ; s , E. H. , an Eocene species , with simple corallum .
The septal pores are present chiefly at the peripheral part of the septum and near the columella ; others , which , vere originally present in the septum between these two areas , have been subsequently filled up .
Secondary thickening goes often so far as to completely transform the lower half of the corallum into a solid mass .
The stages of growth are easily observed on this septum , also the relations of the granulations to the growth-segments ( cf p. llO ) .
A transverse section ( fig. 49 ) near the lip of the corallum shows the septa coalescing here and there by means of synapticular cross-bars .
Where the interseptal loculi are narrow towards the columella , the inner ends of the septa belonging to the higher orders coalesce directly by one of their lateral surfaces with the adjacent lateral surface of a longer septum .
Groups of septa are thus firmly united with one another , and skeletal strength is attained in this manner instead of by the development of compact wall and dissepiments as in the Astrreidre .
* See EDWARDS and HAIME , " Coralliaires , " especially vol. 2 , p. 5 , and vol. 3 , pp. 89 and 90 .
t Ref. Introd .
, LACAZE DuTHIERS , v. KocH , v. HEIDER .
Whilst this mode of skeletal interseptal support at once indicates the nearer affinity of the Eupsammids with Fungids ( of .
Siderastrcea ) and Lophoserids than with Astrreids , au important difference may be at once observed between the arrangement of the interseptal supports in Eupsammids as compared with the arrangement characteristic of the previously-discussed synapticulate families .
In Eupsammia the Fig. 48 .
s.p .
g. s. sp. g , \#183 ; .
c.e. s.p. Septum of E , npsamrnia trochif 01-mis , E.H. ; s.p , , septal pores at the inner ( columellar ) and outer ( pseudothecal ) regions ; c.e. , costal edge of septum ; g , s. , a growthsegment of septum with its curving series of granulations ; gr , , granulations m other growth-segments , , \lt ; .p .
s. C. ic.p .
.Fig .
49 .
syn .
c.c. Transverse section Eupsarnmia ; s. , septa united by the synapticular network ; s.p. , septa ] pore near the periphery ; c. , costa ; ic.p .
, intercostal pore ; c.c. , centres of calcification .
synapticulre are not arranged in vertical series , but alternate with one another , 01\#183 ; are quite irregular in position .
The canalicular system in the interseptal loculi is more highly involved thnn in the Fuugids and Lophoserids , it is a branching network of channels instead of simple , smaller channels extending downwards from the ccelenteron into the intersynapticular spaces .
The most complicated development of this synapticular network is found near the costal portion of the simple corallum of Eupsammia .
There the inter-radial loculi are originally widest , and the basal body-wall of the polyp is invaginated in an entirely irregular fashion .
Within an oblique or \#183 ; erect-growing invagination an erect synapticula develops , it is again united crossways with similar synapticula , which need neither be in the same radial nor in the same tangential line as its neighbour .
Fig. 50 is a drawing of the branching synapticular network in an interseptal loculus near the periphery of the calyx .
This is only a more irregular manner of invagination SYSTEMATiC STUDY OF THE MADREPORARiAN TYPES OF CORALS .
195 of the body-wall at the base of the growing polyp than we have hitherto had occasion to observe .
In the Turbinarinre and Madreporinre the ccenosarced part of the bodywall frequently forms a similar irregular synapticular network in the inter-radial loculi of the " ccenenchyme .
" Fig. 50 .
a.d. syn .
c. Ii , a.d. Enlarged drawing of septal surface ( broken at left-hand side ) , showing the branching synapticulre ( syn .
) , and its part in the formation of rows of tubercles betweeu the costal continuations of septa ; s. septum ; c. costa ; a.d. area of divergence .
1\#189 ; e Area of Divergence .
Tn the description of foregoing genera of the Astrreinre and Fungidrn , I have pointed out that all longitudinal ridges observed on the surfaces of the radial structures are directed in opposite ways towards the septum and towards the costa , and also that a pair of ridges take their origin simultaneously from a line at the periphery of the calyx .
The whole peripheral line , or " area " of divergence of the ridges , becomes in those foregoing types much thickened , and helps to form a " pseudothecal " wall .
The farther development of a ridge takes place by the end-toend attachment of new " segments of growth " until the full length of the ridge is attained .
The completed ridge is comparable to a calcareous chain composed of Jinks of various shape and form , viz. , the growth-segments of the spiniform or nodular ridge , as the case may be .
Only in the case of Lophoseris and , generally speaking , the Lophoserinre , we foun\lt ; l that the trabecular chain of individual parts was occasionally interrupted .
In other words , that the successive growth-parts were not always directly joined up end-to-end with one another , although the ridges , otherwise in point of origin and direction , agreed with the Astrrninrn and Fungidrn .
Lophoseris displays in this respect a decided approach to Eupsamrnia .
Referring again to the surface drawing of the septum of Eupsctmmia ( fig. 48 ) , it will be seen that horizontal growth-curves pass over the septum .
The cnrves are more strongly marked at intervals in the septum corresponding to successive growthperiods .
The granulations on the septal surface form , on the whole , curves which follow the intervals between these thicker , more strougly-marked growth-curves .
196 DR. MARIA M. OGILVIE ON 'fHE MICROSCOPIC AND Considered vertically , however , the series of granulations is interrupted at the peripheral and inner areas of the septum by irregularly-shaped pores .
We may look upon the series of granulations as the superficial form assumed by a successive series of growth-parts belonging to a large trabecula .
The irregular shape of the septal pores in Eupsammia is due to the fact that the growth-parts of one trabecula ma.y be disconnected , and also the growth-parts of adjacent trabeculre .
When the spaces thus left run into one another , naturally irregular pores arise .
Moreover , since the secondary outfilling has no definite relation to the structure , the original pores may , in the course of time , be irregularly diminished in size .
New trabeculre are inserted at the periphery , and a distinct area of divergence is there observed ( a.d. fig. 50 ) .
In Eupsamrnia , however , there is not simply the divergence of one septal and one costal trabecula ; but a definite fan-group of trabeculre is formed at each new growth-period .
The new trabecular parts are placed alternately between those of the foregoing growth-period , and two or three new trabeculre take origin at the " keystone " part of the arch .
One might almost say that towards the columella a second area of divergence is present , with alternating position of the successive trabecular parts like that in the main peripheral area .
The trabeculre are directed on the one side towards the columella , on the other slantingly upwards into the prominent exsert expansion of the septum .
This inner expansion of the septum is , however , a generic feature and is associated with the almost turbinate shape of the\#183 ; simple corallum of Eupsammi\#183 ; a trochiformis .
_ H may be compared with the cristiform lobes of the septa in a few Astrreids and in a large number of Turbinolids .
Microscopic Structure .
The radial longitudinal section , fig. 5 lA , is cut parallel with the septum and passing through the area of divergence at the periphery .
It shows that the " septum " is made up of trabeculre with long , radiating fibres , and that the divergence of the trabeculre distinguishes a costal , or outer , from a septal , or inner edge .
These are once more the sa.me essential features as we have already observed in the radial structures of Astrreid and other coral types .
Each trabecula corresponds in width to a single granulate ridge on the septal surface ( cf. fig. 48 ) .
The trabeculre are usually composed of short series of fascicles .
Interruptions in fascicular growth take place and leave skeletal pores , which correspond to intervals between growth-periods .
More seldom a continuous series of fascicles forms , and a compact trabecula is built up through two or more growth-periods .
This occurs specially at or close to the area of divergence , where the septum is always strongest .
The ends of the fascicles bend outward from the median plane into a tangential direction and are then continued as a synapticular structure or into granulations .
Apertures are left in the septal plane above the parts at which the trabeculre ends bend outwards .
Fig. 51B is a radial SYSTEMATIC STUDY OF THE :MADREPORARIAN TYPES OF CORALS .
197 longitudinal section cut very near the surface of the septum and passing through its entire breadth from columella to wall .
The fascicular ends of the trabeculre are cut almost transversely since they bend outwards near the surface .
The synapticular network is drawn in lighter tone against the darker ground of the septum .
This will be understood by comparison with fig. 50 above .
tr .
Fig. 51A .
a.d. a.d. '------y---_J c. s. s.p. s.p. Eupsammia : Rection parallel with the septal plane , meeting trabeculre only at part of their length , the ' tips of the trabeculoo bend outwards from the plane of section ; a.d. area of divergence ; s. septum ; c. costa ; s.p. septal pores-two of a curve of foiw pores indicating the limit of one growth-segment at this peripheral region .
\#183 ; A transverse section ( \#163 ; g. 5~ ) shows the pores occurring m the septal plane , and , corresponding to them , calcareous s:ynapticular Rtructures on either side .
The section includes the peripheral costal area of several septa , but the large amount of secondary thickening has almost filled up the originally canaliculate system of syna.pticulre .
The centres of calcification are extremely close in the median plane of a septum so that each surface-ridge corresponds to several inner centres of calcification .
At the same time fascicular grouping of the fibres is clearly seen , and the axes of fascicles bend outwards from the median plane into the middle of surface granulations .
The growth-lamelhe curve very regularly round .
definite centres at the first laying down of the septum .
But , as successive growth-periods follow , the primary in-andont curving of the innermost layers becomes lost , and larger curves , much less regula1\#183 ; , make themselves evident .
It is this fact which in Eupsammia gives to the innermost growth-layers a certain distinctness from the subsequently developed layers .
They appear in sections as a lighter band on both sides of the median centres , while the fibres beyond them , although having their crystallographical axes generally speaking in the same direction , seem opaquer .
The opacity is due to frequent variation in the particular form of the curves of successive lamellre , and a consequent crossing of the fibres .
The " Dark Line " of the Septum .
The microscopic sections indicate above all the smoothness of the growth-lamellre in contradistinction to the crimped and corrugated lamellre which we have studied in Astrreinre and in Fungidre .
The leaf-like apposition of successive lamellre is much more easily demonstrated in Eupsammia , although here , and indeed in all Madreporarian types , the septum is composed of fibrous layers from the middle plane to the surfaces .
The median lighter band , corresponding to the deposit of the septal margin , represents the " dark line , " the " primary streak , " and " primary septum " of KocH and other authors .
It cannot , however , be structurally distinguished from _a " lamellar stereoplasm " in the sense of those authors , since the so-called primary septum is itself lamellar .
It is the close and at the same time elongate setting of the centres of calcification in Eupsammia which gives the appearance in ordinary thick sections of a median line in the septum .
We have in all hitherto examined cases precisely the same general structural relations in the septum ; the individual differences in the various types consist in a varying distance from one anothe , \#183 ; of the original centres of calcification around which the flaps of the septal fold lay down their first growth-lamence .
The " True " Synapticul(J3 of Eupsammia compared with Pseudosynapticulce .
Correlated with the smoothness of the growth-lamellre is another feature , externally apparent , the smoothness and bluntness of the marginal apices of the septal ridges and also of the synapticular growths .
The septal edge of Eupsammia can scarcely be called toothed in the acceptance of the expression for Astrreinre , much less is it " spined " or " serrated .
" It is knobbed or nodular , the nodule having an elliptical outline with its long axis set across , and not along the septal plane .
Very often ( where the fibrous rows have bent out from subjacent horizons and continued to grow as upright synapticulre ) small knobs are present on one side or both of the septa !
knobs .
These smaller knobs or nodules are synapticula.r growths , each with its own centre of calcification .
And two or three such nodules may bridge the interseptal SYSTEMATIC STUDY OF THE MADREPOR .
ARIAN TYPES O:F CORALS .
199 Fig. 5LB .
s. syn .
col .
Section parallel with the septal plane , but cut nearer the surface than 51A-hence the outward-bent tips of trabeculre are met almost rectangularly .
col .
, columella ; syn .
, synapticular network , seen as if projected against the darker-looking septum behind ( s. ) .
Fig. 52 .
S !
J1t , sp. ps.syn .
s. s. Transverse section Eupsanimia , showing secondary thickening filling"up the interseptal loculi .
s. , septum ; syn .
, synapticula ; ps.syn .
, pseudosynapticula ; sp. , interseptal space .
' space .
As these nodules are frequently connected with each other by this mode of bending of the rows of fibres , now into upright , now oblique , now horizontal direction , one may look upon a synapticular bridge as growing out from one septum to the next .
PRATZ seems to have had this particular form of synapticula in view when he gave his definition of the " true synapticula " ( antea- p. 185 ) .
I am led to this opinion by the fact that these synapticular knobs of Eupsammia outwardly resemble the component " nodules " of a trabecula of Cyclolites : Microsolena , and other genera of the Pseudo-Astrmim.e examined by PRATZ .
PRATZ did not identify the structure of the ?
lodule of Cyclolites , \amp ; c. , with that of the Astrreid granulcttion .
Their identity is however demonstrated in the preceding parts of this paper , since both " nodule " and " granulation " co1\#183 ; respond with one or more f asicular groups of fibres extending towards the lateral surface of a septit1n .
Some of the synapticular knobs uf Eupsamniia are , on the other hand , formed in special invagi'.nations of the basal bodywall of the polyp , and are homologous with the new centres of calcification introduced in the intersepta.l loculi of Fungids and Lophoserids and forming a coalescence of the adjacent granulate outgrowths .
In Eupsammia the invagination does not continue as in the case of the synapticular " bar " of Fungia in a definite vertical direction , but ceases after the formation of a single synapticular growth .
The next invagination arises in any direction obliquely to , or rectangularly outwards from , the foregoing .
Certain points of similarity with the Turbinolidce .
Curiously enough , another Madreporarian family , apparently remote from the Eupsammid::.e , has a very similar microscopic septal structure .
In the Turbinolidce , the axes of calcification in the median plane of the septum are also very closely set and the growth-lamell::.e very smoothly apposed .
An important difference however consists in the fact that a , trabecular axis never bends out of the median plane in Turbinolid::.e .
There are therefore no septal granulations into which the axes of fascicular bundles of fibres bend .
When a granulation forms on the septal surface , it covers a number of small centres of calcification in the median plane , from which practically parallel-placed fibres pass into the granulation .
The fine structure of smooth-septate Eusmilinre is essentially similar ( antea p. 161 ) .
The Turbinolidre often resemble Eupsammia in another characteristic , the coalescence of septa at their inner ends often building a symmetrical figure .
The genus Turbinolia itself shows this feature , and still more the genus Deltocyathus and other genera belonging to the Trochocyathus group .
Such local coalescence can only be produced by the invagination of a part of the body-wall of the polyp , not in the direct radial line of the septal invagination , and is therefore homologous with synapticular invagination .
It affords a means of support to the interseptal portions of the body-wall , and in the case of the short Turbinolid C'orallum , renders the customary dissepimental support less necessary .
We have already studied an analogous case in Fungia .
SYSTEMA'l110 STUDY OF TF.i:13J MADREI'ORA .
RIAN 'I'YPES OF CORALS .
201 Here , however , the points of similarity between Eupsammia and the Turbinolidre end .
Wall and septa of Turbinolidre are compact and the mural ring of calcification is in continuity with the septal centres of calcification .
I shall return to these distinguishing features of the Turbinolidre in discussing the remaining so-called " Perforate " types :-Turbinaria , Mad1\#183 ; epora , and Po1\#183 ; ites .
SYSTEMATIC RELATIONSHIP OF THE EuPSAM .
MIN .
LE ( E. and H. ) TO THE FAMILY OF ' FuNGID-LE ( DANA EMEND .
ZITTEL .
) The Septo-Costal Area of Divergence .
In common with the Astrreinre and the Fungidre , the Eupsarnmime possess radial structures in which a thicker peripheral area is the seat of divergence of the component septal and costal trabeculre .
Farther the extra thickness of this peripheral area is a large constituent in the formation of a calicinal wall area .
Between the adjacent thickenings of the radial structures , a very small invagination of the basal part of the body wall is sufficient to form a compact wall .
According to the particular shape of the invagination , an erect but flat-surfaced dissepiment may be formed or a roundish calcareous rod of synapticular character .
The wall which is formed in this way may be said to be several growth-periods behind the septa in its original development . .
These .
relations of walls and septa are characteristic of Astrreinre , Funginre , Lophoserime , and Eupsamminre , and are absent in the Turbinolidre , Madreporidre ( excl .
Eupsamminre ) and Poritinre .
The Lophoserinre more especially resemble the Eupsamminre in respect of the noncontinuity of the successive growth-parts in the septal trabeculre .
The irregular pores in the septa are thus explained in both Lophoserinre and Eupsamminre .
The nodulate shape of the individual trabecular growth-parts in the Eupsamminre cJosely resembles that characteristic of Pseudoastneinre , of Siclernstrcea , and of Lophoserinre .
In all these types , as well as in Fungia , the coalescence of septal ends with one another is frequent , and is associated with the general fact that the basal body-wall in this group forms small invaginations at any part of the interradial loculi .
As a rule , the branching colonies of Eupsammids tend to the simpler form of canalicular system observed in Funginre and Lophoserinre , and have not the complicated crenenchyme of Turbinwria , Madrepora , and Porites .
Recapitulating briefly the most important features of Eitpsarnmia , we have :(1 .
) The trabeculre composing the septum often incompletely united with one another , as in Pseudoastrreinre , Fungia , Lophoseris , \amp ; c. ( 2 .
) The individual growth-parts of any one trabecula not always attached end to end , although following each other in vertical succession .
Where such an aperture is left in the trabecula , and it happens to be alongside of an aperture left between 202 DR. MARIA M. OGILVIJ\lt ; J ON THE MICROSCOPIC AND two adjacent trabeculre , an " irregular pore " appears in\#183 ; the septum .
The Lophoserinre show the same feature .
( 3 .
) The fascicle of fibres has its axis in the median pfane of the septum , but may bend out in unpaired fashion towards an interseptal loculus , and give origin there to a synapticula .
This feature is , perhaps , the most characteristic for Eupsammia in contradistinction to the LophoserimB , Fungime , and Pseudoastrreinre .
To it is due the occurrence of synapticulre , not in series above one another , but irregularly .
( 4 .
) A great variety in the precise course and position of synapticular calcareous bars gives the skeleton of Eupsammia at its peripheral portion a certain resemblance with the involved mesh~s of ccenenchyme in J.l1adrepora and Turbinaria .
Again , the fact that many synapticuhe are directed obliquely inwards and upwards towards the calyx shows the independence of the synapticula and its analogy with the growthpart of a radial trabecula .
The canalicular system of the Eupsamrninre , E.H. , varies we may say between the less complicated Fungia type and the more highly differentiated " perforate " type , but it is the least constant of all features in the Eupsamminre , and the last that may serve as a systematic basis .
\#183 ; ( 5 .
) From the above microscopic study of skeletal features , the genus E\#183 ; upsammia will be seen to have much in common with the Fungidre .
It is very significant to find that the anatomical relations of another genus belonging to the Eupsamminre show very marked affinities with those of Fungidre .
The species Stephanophyllia .
form , osissima has been described by FowLER ( " Madreporaria , IV .
, " Q. J. M. S. , February , 1888 ) , who writes:- " In several respects Stephanophyllia exhibits a resemblance to the Fungia recently described by BouRNE .
The plano-convex shape of the corallum , and the correlated basal position of the theca , are , of course , characteristic of both genera ; besides this , a view of a complete mesentery shows that a few synapticulrn brace together thA septa of this coral , and that between the synapticulm the muscular pleatings of the mesoglcea are gathered into strong bundles in a very characteristic manner , hitherto described only in Fungia .
The histological appearance of the craspedon is also similar , and characteristic in both forms .
On the whole , Stephanophyllia appears to bring Fungia into closer relations with the Eupsammidm than have been generally allowed .
" According to all the above results , the sub-family Eupsamminm , E.H. , deserves a systematic position as an independent family , the Eupsammidre , allied with theFungidre .
The relationship of the Eupsamminre , E.H. , to the Madreporinm , E.H. , and 'I'urbina-\#183 ; rinm , E.H. , is for more doubtful , as later parts of this paper will show .
Amongst the M. perforata , E.H. , it is the sub-family Poritinm , E.H. , which has most apparent resemblance to the Eupsammids , yet no clear evidence can be given , that these two groups are nearly related .
SYSTEMATIC STUDY OF 'l'HE MADREPORARIAN TYPES OF CORALS .
203 CHAPTER VII.--SOl\1E " PERFORATE " TYPES : TURBINARIA , MADREPORA , PORITES .
TURBIN ARIA .
\#183 ; Fig. 53 .
The species examined , T. mesenterina , LAM .
, has a low foliate corallum , the individual polyps grow very obliquely forming a calycinal tube of turbinate shape .
' ' Fi~ .
53 is the drawing of U : half calyx , presenting the surface view of two septa , and Fig. 53 .
tl , .
p. en .
p. s. th .
str .
en .
col .
sfr .
Enlarged drawing of Ti1rbinaria .
mesenterina , LAM .
; col .
, columellar pali ; th .
p. , thecal pores ; en .
p. , camencbymal pores ; s. , septa ; str .
strim ( septal.trabeculre ) .
the central pa. Ii and columellar area .
The septa are compact and smooth , reminding one of thA Turbinolid and Eusmilinid type in general .
On the other hand , the wall is ess~n.tially synapticular in character , apertures being left in regular series and affording the means of communication between the canalicular crenenchyme and the polypal cavities .
During the life of the polyp the crelenteron , or body-cavity within the theca , is continued into the extrathecal crenosarcal tissue lining the complicated crenenchymal network of skeletal canals .
t The general relations of the hard and soft parts in Turbinaria may be studied in fig. 56 .
* A species nearly allied to T. mesenterina is figured by EDWARDS and HArnE in the " .Atlas of Hist. Natur .
des Coralliaires , " Plate E , fig. lA and lB , Turbinaria cinerascens .
t The anatomy of the polypal parts has been worked out by Dr. G. H. FOWLER , " The Anatomy of the :Madreporaria III .
, " 'Quart .
Journ. Mier .
Sc. , ' August , 1887 .
2 D ~ The Skeletal Parts of the Calyx .
Septa.-The lateral surfaces of the septa are striated , and on the strire are occasional small granulations .
The strire are 11ot arranged in the form of a divergent fan of ridges as in Galaxea , , neither do they diverge from a peripheral area in the radial structure as in Astrreids , Fnngids , and also , if with a slight difference , in Eitpsammia .
But the strire are placed almost horizontally with the upper edge of the septum so that the apex of the stria is directed towards the centre of the calyx .
In a transverse section of the calyx usually only one or two strire are cut in each septum , and the microscopic structure presented must therefore be compared , not with the transverse sections of Astrreidre , \amp ; c. , but rather with any longitudinal , t angential section through a few strire or trabeculre in these groups .
en .
th .
p. th .
s. t. Fig.:54 .
Tnrbinaria : transverse section showing the microscopic structure of septa ( s. ) , their prolongations as pali ( pa. ) surrounding the columella , the theca ( th ) , thecal pores ( th .
p. ) , and coonenchymal interspaces ( en .
) , radial in directio ; n ; t. , an accidental t ear in the section .
Fig. 54 shows the agreement of the microscopic structure of the septa with that already described for Euphyllia , in a less degree with that of Eupsammia .
The dark band in the middle is composed like that in Euphyllia of very closely set centres of calcification .
Similarly there are no fascicle-groups of fibres passing out.wards from the median plane to the surface granulations .
The growth-lamellre are smooth , and only the outermost lamellre show a less uniform deposit ; the fibres become lengthened at some parts and not at others , and thus superficial granulations of small size occur on the strire , bearing however a definite relation to particular series of calcification centres .
Septa , Pali , and Colitmella.-The paliforrn prolongations of the septa have the SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
205 same microscopic structure as the septa .
They surround in buttress-like fashion a central lamellar columella , which is itself the continuation of one of the primary septa .
A diagrammatic ground-plan of the septa , pali , and columella is given in fig. 55 .
The position of the columella in the long diameter of the usually somewhat Fig. 55 .
Ground plan of calyx ( T. 1nesenter , iua ) .
c. , columella , a continuation of one of the primary septa ; s. , septa , primary and secondary septa unite as couples to form pali ( pa. ) .
The theca is formed by tbecal and costnl bars alternating with pores ( th.p .
) ; similarly the coonenchyme ( en .
) .
See p. 209 .
elliptical calyx , as well as its crest-like prominence amidst the pali , giyes a marked bi-symmetry to the skeletal parts , ( cf FowLER , the position of the " directive " mesenteries is at the ends of the dividing plane of bi-symmetry in the polyp , Zoe .
cit. , p. 87 .
) An important feature represented in the figure , is the origin of each palus from the coalescence of two of the principal septa .
This at once distinguishes the Titrbinct , ria pali from the occasional paliform prolongations of single septa in Goniastrcea and a few other Astrceid genera , while it shows their resemblance with many of the Turbinolidre in which rings of pali alternate in position with the septa .
The coalescence of the septa to form the pali takes place according to a definite system .
One primary and one secondary septum unite in couples all round the calyx .
An exception to this is at the ends of the long diameter marked out by the crest , where the primary septum either remains solitary or unites with the secondaries on both sides .
The primary which remains solitary bends sidewards , and then returns to the mid-line and forms the columellar crest .
A long , leading septum with columellar continuation may be said to be one of the most characteristic features of the Eusmilinre , En .
and H. , and it is among genera of the same sub-family that we find lamellar pali in front of the principal septa , e.g. , Dendrogyra , Pectinia .
Those are also , as I have already said , familiar in Caryophyllia and other Turbinolidre ( see also p. 272 ) .
If we now bear in .
mind the similarity of the fine structure of the septa in Turbinaria with that of Eiiphyllia or of Ca1\#183 ; yophyllia , it will be admitted that the 206 .
1 DR. MARi .
A M. OGILVIE ON THE MICROSCOPIC AND skeleial parts of the calyx of Turbinm\#183 ; ia have many points of affinity with similar parts in the calyces of Eusmilinid and Turbinolid genera .
A featur~ of smaller moment is the s1anting away of the pali from the radial septal plane as they grow upwards and inwards .
The same slanting growth is observed very often in the inner ends of septa ( e.g. , Euphyllia ) .
It has in Turbinaria a certain significance , as the pali in winding round , become coalescent with one another by small synapticnlar bars , and considerable strength of the amphicolumellar structure is attained .
Between the septa and the pali a trench-like hollow is observed , which has its origin in the deep scallop between the inner ends of the coalescent pairs of septa and the corresponding palus .
The mesenterial filaments descend into this skeletal trench within the polyp , and I have little doubt that the Turbinaria " trench " is the .
functional equivalent of the " fossula " of Palmozoic Madreporaria .
To this I shall return in chapter ix .
Conflicting reports are often given in specific descriptions of pali-bearing forms , with regard to the actnal number of rings of pali present .
A section made near the growing lip of the Turbinaria calyx yields two rings of pali , a section at some distance below yields only one .
The explanation is easily observed in a longitudinal section of the calyx .
The pali starting from any given growth-horizon have already grown inwards and upwards to the main columellar crest when the next series of pali begin to develop .
At the origin of the next series , the two rings of pali are clearly apparent , but subsequently the pali in the same radial line are united by secondary calcareous deposit , laid down by synapticular invaginations of the_ basal body-wall as the soft tissues are being slowly withdrawn from lower to \#183 ; higher growth-horizons .
Similar observations may be made on the septal pali in Poritidre and on the spines of Madrepora .
The so-called Synapticular"~ Wall .
Iu discussing the synapticular bars of Fungia , , I referred to the attachment of the mesenteries to the polypal body-wall at the oral , lateral , and basal body-surfaces of the polyp .
In Fungia , the lateral and basal body-wall of the polyp to which the mesenteries are attached is invaginated subsequently to the septal invaginations , the interseptal invaginations then keeping pace with the septal , but at a lower horizon of growth .
A continuous synapticular bar thus takes origin over which the bodywall bearing the mesenteries is doubled into long folds lining canalicular continuations of the ccelenteron between adjacent synapticular bars .
* The term " synapticular " bas been applied to the porous wall of " Perforatt s , " no doubt \#183 ; because of the general resemblance between the skeletal bars of the wall and the synapticula:r : bars between the septa in Fungidre .
But I think the term misapplied .
The original use of " synapticula " was limited to a , bar 11-nitiriy septa ( seep .
167 ) , and necessarily of later origin than the septa .
\#183 ; SYSTEMATIC STUDY OF THE MA:bRl !
JPORARlAN TYPES OF CORALS .
'l,07 The process in Sidei\#183 ; astrcea and Lophose1\#183 ; is is similar , but no continuous synapticular bar is formed .
Separate synapticulre are laid down in parts corresponding with the growth-periods of the polyp .
Thus apertures are left unfilled , although the synapticular invagination continues to be withdrawn upward in a definite vertical direction .
In Eupsammia , still another variety is presented-new synapticular invaginations are made at each growth-period , and they are not above each other , but either alterriate or may be irregularly placed .
Such ap arrangement denotes a much greater complication of coolenteric canals than in the case of the vertical synapticular series or continuous bar .
One structural feature is , however , shown in common by all those types .
, It is that the invagination of the mesenterial portion of the body-wall takes place at some distance below , and is at first of posterior date to that of the septal and costal portions .
Hence the parts of the mesenteries enclosed in the synapticular channels are continuous in the open interseptal loculus above the synapticular invagination .
In other words , the mesenteries in the synapticular channels are merely the prolongations of the mesenteries in the upper non-synapticular part of the coolenteron .
Further , the mesenteries pass continuously over the synapticular wall into the " Randplatte , " or edge-zone , of the polyp .
Turbinaria , on the other hand , at once distinguishes itself in the fact that , at the peripheral portion of each polyp where mural invaginations are formed , they form contemporaneously with or even zJ1\#183 ; ior to the invaginations in which septa are produced .
This fact I have deduced from a series of sections .
In all types mentioned above , the synapticulre form subsequently to the septa , and microscopic sections show that the calcareous deposit around the septal centres of calcification serves as the starting-point of a synapticula , the prop on which the mesenterial body-wall temporarily supports itself in the course of its further invagination .
In the converse case represented by Turbina1\#183 ; ia , the sections show a mural ring of calcification centres around the polyp , from which at certain points processes grow out and serve in their turn as the starting-points of septa .
I describe here merely the appearances of the section .
A comparison between transverse sections of Siderastrcea ( fig. 4:3 ) or Eupsummia ( fig. 52 ) , on the one hand , and Turbinaria ( fig. 53 ) and Madrepora ( fig. 57 ) , on the other , will explain the difference .
The same difference may be observed between the transverse sections of the wall and septa in an Astrreid like Heliastrcea ( sub-family Astrreinre ) and those of Galaxea , Euphyllia , or , still more , Turbinolia , Flabellum , and other genera belonging to the Turbinolidre .
There the invaginations homologous with the synapticular invaginations in the previously-mentioned types are continuous and give rise to a compact wall ; yet the priority of the septal invaginations in the Astrreiure , and , on the contrary , the contemporaneity or the priority of the wall invaginations in the ~08 DR. MARIA M. OGILVIE ON 'l'HE MICROSCOPIC ANO Turbinolidre , produce similar appearances in transverse sections , as have just been described for different synapticular types .
I find no statement in FowLER 's paper on Turbinaria with regard to the timerelations of the development of radial and inter-radial structures of the polyp .
But my own microscopic sections of the skeleton seem to me sufficient basis for my statement that the inter-radial invaginations at the mural area are at least not later than those of the radial structures .
The wall " synapticulre " are , therefore , at the same horizon in the calyx as the septa ; they are no longer basal in position .
In analogous cases a number of researches give clear evidence of a difference in the time-relations of the radial and inter-radial invagination ( vide Kocn 's various papers on development ) .
The contemporaneous time-relations of the septa !
and mural invaginations in Turbinaria , together with the alternations in position of the mural and costal invagination at successive growth-periods , are , in my opinion , the main reasons which explain the " cutting-off " of the polypal mesenteries at the mural portion of the polyp from the camosarcal inter-polypal portions of the colony .
The Perforations in the } Val occ-w ; \#183 ; between successive Mural Bars .
Two successive stages of growth are represented in the diagrammatic ground-plan , fig. 54 .
In both stages the cCBlenteron of the polyp is continuous with the cCBnosarcal canals above the septa and between the adjacent mesenteries of two neighbouring pairs of mesenteries ; in other words , the inter-mesenteric chambers of the polyp are extended continuously into the cCBnosarc .
At one stage represented in the figure , the mesenterial parts of the body-wall are invaginated at the periphery of the polyp , and mural supports are produced ; at the other stage , the mesenterial parts of the body-wall continue freely across the mural area , but new costal invaginations are formed , corresponding in position to an ectocCBle , i.e. , the part of the cCBlenteron between adjacent mesenteries of two neighbouring pairs of mesenteries .
This alternating mode of support at one stage by mesenterial invaginations , at the next by additional costal invaginations , produces the characteristic double series of apertures in the " wall " of Titrbinaria between any two septa , and at the same time is seen to be necessarily correlated with the formation of costre unrepresented by corresponding septa .
I may at once call attention to the similarity which the genus Turbinolia presents in the possession of additional cost(JJ , as well as in the appearance of a double series of dimples on the outer surface of the compact wall between every two septa .
An allied Turbinolid genus has a double series of pores in the wall ( see below , p. 249 ) .
SYS rEMATlC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
30 !
' ) TURBIN ARJ A. A Ccenenchymatous Reef-building Ally of the Turbinolirlm .
FOWLER writes ( ' Q. J.M. S. , ' August , 1887 , " Madreporaria , " IIL ) : " If we are justified in crediting the appearance of the tissues in Stylophorci , McidrPpora , and Turbinarict , which implies that the body-wall is supported upon the echipulations , it is not perhaps too mueh to infer that t.hese relations are of secondary significance and have arisen contemporcmeously with the development of ccenenchyme for the support oj the external body-wall , owing to the inadequacy of the peripheral sections of the mesenteries to effect this support elsewhere than immediate1y round the theca ( where this is exsert from the ccenenchyme ) .
" I may also refer to BouRNE 's General Remarks on the porous and aporose skeleton ( Zoe .
cit. , 'Q .
J. M. S. , ' August , 1887 ) which so for anticipate my results .
The latter show that the presence of a porous ccenenchyrna ( of which the porous wall is simply the proximal calycinal part ) in Turbinaria is itself of only secondary systematic importance .
The features of the calyx are the basis on which classificatory relationship must rest , and the result of my investigations would go to prove that Turbinaria 1:s a specialized " Pe1forate " Turbinolid , specialized in that part of the po1yp which forms the thick wall and epitheca of its non-crenenchymatous relations .
Fig. 56 .
TURBJ : TARIA.--Ccenenchymatous type ; letters as in fig. 2 , p. 107 ; a columella ( col .
) , and pali ( p. ) , are present ; th.p .
thecal pores , which opn into thoi ; e ( black in the figure ) of the ccenenchyme ( en .
) ( cf. p. 203 ) .
This diagram is given for comparison with the siIL larly-treat , ed figs. 2 , 22 , 35 , and 36 .
The sh letal parts of the Ccenenchyme agree in Mic1\#183 ; oscopic Structure wi'th those of the Calyx .
One can generally distinguish in the crenenchyme radial or costed and i11ter-racl\#183 ; :ol or synapticular centres of calcification ; although , owing to the subsequent continuity of the original invaginations , the two are very often coalescent .
Then the crenenchyme may appear to be concentric in its arrangement , or variously irregular .
The greatest irregularity naturally obtains at the parts most remote from adjoining calyces , where the soft parts rest more or less loosely on the trough-like depressions of the camenchyme .
The details of microscopic structure prove that each so-called " echinulation " of the ccenenchyrne , wheth er in a costal or synapticular position , is itself identical with a septum or a pal us of the calyx .
There is a dark line medially made up of closely set axes of calcification and ori0ainally of organic constitution and on both sides of it or .
' ' arou nd it ( according as the echinulation is elliptical or circular in section ) , are smooth F ig .
57A .
s. en .
~ .\#183 ; I ' Transverse seclion t hl'ougb camencbyme between two adjacent calyces of Turbinaria .
Median " darker line " ( calcification centres ) seen in the septa ( s. ) and also in the several t rabeculro of the camencbyme ( en .
) .
Fig. 57B .
Section longitudinally cut th'Ough coonenchyme , showing the varying direction of trabeculre ( tr .
) and the occurrence of pores .
tr .
growth-lamellre in which the fibres are parallel with one another .
Usually the " echinulation " of Turbinaria is laminate in form , and one can often distinguish that the lamina is composed of two or three strire , and repeating , in fact , the septal structnre .
Fig. 57 A is a transverse section through the ccenenchyme between adjacent calcyces ; the trabeculre are met in all possible directions .
Fig. 57B is a longitudinal section through the camenchyme , showing the irregular slanting growth of the trabeculre and the consequent irregular occurrence and shape of intertrnbecular pores .
The echinulations of successive growth-periods are not necessarily above one another in position , hence the system of canals is very irregular , analogous , speaking generally , with the syst em obtaining in the mural area of Eupsammia .
The 'JCBnosarcal parts descend to very unequal depths within the canalicular system .
In SYSTEMATIC STUDY 01 !
' THE MADREPORARIAN TYPES OF CORALS .
211 the diagrammatic representation of the camenchyma given in fig. 56 , it is impossible to show more than the general relations of calyx and crenenchyma .
AcTINACIS .
The Cretaceous genus Actinacis has a fine structure of its skeleton , identical with that observed in Turbinaria .
The transverse section , fig. 58 , may be compared with fig. 54 , a corresponding section of Turbina1\#183 ; ia .
The small part of the calyx at the lefthand corner of fig. 58 shows especially well the connection of septa and wall .
Instead of a septum being continued through the wall as a costa , it appears to divide in two halves , which bend away from one another in the direction of the wall .
Between the Fig. 58 .
S\#183 ; th .
p. en. .
P\#183 ; Transverse section AcLinacis ; s , septa ; th .
p. , thecal pore , where a septum splits into two , to form mural bars ; en .
p. , camenchymal pore in a line with the thecal pore .
two halves a thecal , virtually a crenenchymal , pore is present .
These relations of structure in wall and septa are the same as in the Turbinolid : B , and , as I have already explained , can only take place when the upper edges of the wall and the septa keep equal pace in growth , remaining at one and the same height , or where the wall is prior in growth to the septa .
In my " Monograph of the Stramberg Corals , " I have discussed the resemblances between Actinacis and the Jurassic genus Tharnnarcea .
The latter has hitherto been placed among the Pseudoastrreinre , PRATZ , I believe erroneously .
MADREPORA .
The ans , tomy of the J , !
f.adrnpora polyp , and the relations of the polypal and camosarcal soft parts to the coral skeleton , have been made the subject of a special memoir by FOWLER.* Since , moreover , the coral skeleton is especially familiar in its general appearance to all , I shall at once enter into detailed structural features .
A longitudinal section made near the surface of a branch is presented in fig. 59 .
A number of calyces are cut obliquely , still the general arrangement of six interseptal loculi can be observed in the calyces .
The skeletal structures , calycinal or ccenenchymal , are seen in this drawing to be built up of trabecnlre with radially arranged fibi \#183 ; es .
Fig. 59 .
c , i. t. ; ; , C\lt ; t. Cl/ .
en .
Longitudinal section passing obliquely through a branch of Madrepora ; ca .
, the six-septate calyces distorted by section ; i.s. , an interseptal loculus ; en .
, camcnchymal trabeculre , I have explained under Turbinaria how the development of the mural invagina.tion previously to the septal invaginations is correlated ( 1 ) with the introduction outside the poJyp of new costal invaginations not represented by septa , and ( 2 ) with the invagination of those parts of the aboral body-wall of the ccenosarc which correspond to the mesenterial parts of the body-wall of the polyp .
In these ccenosarcal invaginations synapticular structures form , which are practically side-growths of the costal structures ( antca , p. 209 ) .
The same explanation may be extended to l , fadrepo1'a , with the additional fact of a much more rapid tangential expansion of the polypal and cmnosarcal body-wall , and consequently the introduction at the mural part of the polyp of a greater number of costre without relatiou to septa .
Unlike Eupsammia , no " area , of divergence ' .
' is present between the septal and costal parts of the radial structures ( " tiepta " ) in Turbinaria or Madrepora .
Its absence is another feature correlated with the contemporaneity or priority in development of the wall-ring of calcification relative to the radial structures .
* Fowum , " The Anatomy 0\#163 ; the Madreporaria , " II .
, 'Q .
J. M. S./ August , 1886 .
SYSTRMA'rIC STUDY OF THE MADREPORARIAN 'l'YPFJS 0.1 !
- , CORALS .
213 While the above observations are founded on the study of microscopic sections of the skeleton , they are in harmony with an important comparison which FowLER has drawn between the anatomical relations of the camenchyme in Rhoclopsamrni\#183 ; a , a ccenenchymatous Eupsammid , and Madrcpora ( Zoe .
cit. , 'B .
Anatomy ' ) .
In one of his best-known papers on Madreporaria , KocHi , remarks of Madrepora , " The skeleton shows already considerable thickening close to the calycinal opening , and it may readily be observed that the secondary introduction of peripheral secondary septa is of importance in the conrse of enlargement of the branches .
The later thickening distinguishes itself clearly in cross-sections by its layers , and it fills ont large \#183 ; spaces .
" The first part of this statement concerns the skeletal feature just discussed ; the second part expresses again the general conception of a primary skeletal tissue , differing in structure from a secondary , viz. , the sternoplasm , only the latter being looked upon as lamellar .
I have already had occasion , under Astrteidte [ l , ncl Fungidte , to prove that this conception is incorrect , and now , in Mallrepora , although the appea , ranr.e of the deposit towards the surface of any skeletal part varies in a higher degree than in any other Madreporarian coral which I have hitherto examined , I hope to show by microscopic sections that the pr\#183 ; irnary deposit as well as the " later thiclcenin_g " is la1nellar in origin and structure , and that the g1\#183 ; 01.uth-la1nellm of .Macln ; porci from first to last are composed of coalesced calcififd ccdicoblusts , which once represented living ectoderrn .
]11.icJ'Oscopic Sections .
A transvJrse section ( fig. 60 ) through calyx and ccenenchyrne shows at once ~he strong contrast to which KocH referred .
As in Eupsammia , a clearer band iH apparent on both sides of a very dark " middle line " or " centre of calcificatior~ " in every skeletal structure ; but the calcareous deposit beyond the clearer band has no lunger its fibres oriented in the same direction as those of the clear band , they Hurround it in every possible direction .
The continuous thecal ring of calcification , and its coalescence with septal and costal centres of calcification , is well shown in the same figure .
\#183 ; I shall at once compare longitud/ nal sections ( fig. 61 ) .
The calyx is very obliquely cut , showing the main septum and the smaller one opposite it , met in longitudinal section .
The contrast is again observed , and the fibres of the central area of calcification are seen to be more horizontally oriented than those of the so-called " stereoplasm .
" The central areas are continuous in the septa and in the thecal commissures , bearing out the statement above that the invaginations , both tangential and radial , in the calyx and out of the calyx , are contemporaneous in their origin as well as in the formation of their calcareous deposit .
The outer lightet-looking fibres belonging to * KocH , " Uebcr das Verhaltniss von Skelet und W eicbtheilen bci den Madreporim , " 'Morph .
Jahrb .
, ' vol. 12 , 1886 , p. 159 .
the " later thickening ' ' wind themselves in and about all the original centres of calcification , and their orientation becomes more oblique , in some parts it is almost vertically upwards .
Fig. 62 is a longitudinal section less obliquely cut than fig. 61 , and shows better the variation in the direction of fibres at the original centres of calcification and at the outer parts corresponding to later periods of growth .
At the right side of the figure the section bas met the central part of the skeletal structures , and one sees the dark-looking areas at which the first layers of deposit formed .
At the left side of the figure the section has passed deeper , the darker areas are polished away , and only the outer deposit is exposed .
It is clear from the sectio~s that the difference in the orientation of the calcareous fibres is associated with the rapid growth of the Madrepora cora.llum as well as with the constantly changing apical direction of growth .l !
'ig .
60 .
sec. c.c. th .
c. s. Transverse section ( partly injnred inside the calyx ) .
The fl.bro-crystals close to the centres of calcification are met obliquely by the section .
Those of the secondary deposit ( sec. ) are met almost perpendicularly .
followed by the individual polyps of the branching-stock .
The arrangement of the individual polyps around a leading mother-polyp in the colony is well-known , also the twisting of the polyps during growth .
Little need here be said of the Madrepora septum .
Its agreement with Turbinaria is seen in the possession of a well-marked , darker central axis of deposit directed from the porous wall inwards ( cf. also Part IL , pp. 270 , 279 ) .
The peculiarity of the septal apparatus of Madi\#183 ; epora ) as compared with that of Turbinan:a , , is in the fact that a number of the smaller cycles of septa are represented by vertical series of spmes .
These " septal spines " of Madrepora haiVe e'lch a single centre of calcification which SYSTEMATIC ' STUDY OF THE MADREPORARIAN TYf\gt ; ES OF CORALS .
215 in some species appears as a ring at its base , in others has itself a broad base and narrow tip .
Its variation in form is always correlated with some external difference of shape in the spine itself .
The spines are developed in pocket-like invaginations extending from the aboral bodyftwall of the polyp which is apposed to the inner wall- Fig. 61 .
th .
sec. 1/ ' !
,8 .
Fig. 62 .
syn .
f .c .
syn .
sec. c.c. c.c. Longitudinal sections-fig .
61 obliquely cut , fig. 62 vertically cut : m.s. , main septum 0\#163 ; calyx ; th .
, thecal centres of calcification ; sec. , secondary deposit ; f.c. , fibro -cryst , als 0\#163 ; secondary deposit met across their axes ; c.c. , centres 0\#163 ; calcification in costre ; syn .
, synapticulre .
surface , and from the appearance of sections , I am inclined to think that the 1nvag1nated pocket of polypal flesh is not withdrawn , but dies out upon the skeletal strncture to which it gives origin .
The next spines may or may not be exactly above each other ; they are all directed horizontally .
This feature is very different from the type of perforate radial structure observed in Lophoserids and Eupsarnmids .
In those the successive spines ( = trabecular growth-parts ) are vertically or obliquely directed .
When not iu contact end to end , they still keep to a fairly definite vertical line of growth , and can be distinguished as a trabecular chain of individual growth-parts . .
Mctdrepora may be said to displu.y a kind of haphazardisn .
in respect of its skeletal invaginations , more especially in its camenchyme , as if at any time and at any place a new invagination might appear .
Sometimes an invagination shoots up in a radial line of costal spines , sometimes it is an invagination between two adjacent lines of costal spines , and again a flat smooth surface of the body-wall is busy laying down laminar thickening around and between the original invaginations .
It is impossible to ,9nd in the ccenenchyme of the Madrepora corallum regular \#183 ; relations of growth governing radial , tangential , and basal stn1ctures , such as the growth-relations which I have elucidated for the Astneidre , \amp ; c. One can , ho , vever , in a single costal series find nodular , pseudo-synapticular thickenings of the spines indicating the pauses in the growth of the individual series .
But these pauses occur at different horizons of two aclJaccnt costal series and even of adjacent spines in a single costal series , hence no regularity can arise in the position of pores in the skeletal strnctures .
It will readily be seen how completely the unequal growth of a skeletal system of support replaces for the ccenosarc the support afforded in Astrreidre and Fungidre by mesenteries extending from the polyp into the edge-zones or extrathecal parts of adjacent polyps .
At the same time this unequal spinate system of grm , vth outside the polyp is necessarily independent of the system of growth within the polyp , since the polypal wall is contemporaneous in its development with the septa and both wall and septa are anterior in development to the parts outside the polypal walls .
For the same reason , viz . .
, the contemporaneity of mural and radial invagination in the polyp , the mesenteries are shut within the polyp .
The wall grows upwards as rapidly as the septa , and the rnesenterial loculi have therefore no opportunity of extending between septa over a lower-placed lip of the wall and forming an edge-zone continuous with the polyp .
There can be no edge-zone , in short , where wall and septa are contemporaneously developed .
The Calicoblasts .
The calico blasts at the growing apices of skeletal spines in ilfad1\#183 ; epora must , to judge from the rapid growth of the corallum , undergo the changes of calcification very rnpidly .
They appear also to be less completely calcified than in the case of corals where the invaginated layer retreats slowly and uniformly , and rests for a longer period in any particular position .
This I take to be the reason why the original fibrous deposit at the central areas of calcification in Madrepora are so much intermixed with organic matter , and seem so opaque in sections .
SYSTEMATIC STUDY 0:1\lt ; 7 'l'HE MADRl !
JPORARIAN 'l'YPES OF CORALS .
217 The vertically directed fibres have a chequered appearance , not unlike striated muscle fibres , and measure \#183 ; 006 millim. in diameter .
Each such fibre represents a section through growth-lamellre , the chAquering of the fibre indicating the " bands " of contact between successive calicoblasts .
This is proved by fig. 63 , which shows the scale-like appearance of the calicoblasts composing the long fibres .
Frequently finer Fig. 63 .
c.s. S.S. s.t. cf. Longitudinal section of Madrepora ( partly decalcified ) ; c.s. , calicoblast scales in decalcified portion ; c.f. , calcareous fibres ; s.s. , septal spines covered by the constituent scales of the .g-rowth -l amellre .
fibrils are seen within the main fibres .
The ngure represents a thin section of .Madrepora which has been slowly decalcified round the edges .
In the decalcifi ed portions the fibres are seen to break up into the constituent cells or calicoblasts .
At the right hand of the figure a longitudinal series of septal spines are- seen , projecting horizontally into the lumen of the calyx .
Th e rest of the figure passes vertically through the crenenchyme .
'l'his seetion is of greaL importance , as it wholly confirms what I explained in G1_lfoxw above , that I.he calcifieJ calicoblasts belonging to successive 218 IJR .
MARIA M. OGILVIE ON ' 'l'Hl\lt ; l MICROSCOPIC AND lamellre unite to form the long and apparehtly continuous fibres , which radiate outwards from the centres of calcification to the skeletal surfaces .
In structure and origin , therefore , the " later thickening " is identical with the primary deposit .
But it remains as a typical feature of Madrepora , that the primary fascicles are laid down in a nitmbe1\#183 ; of ectodermal pockets entirely separated from one another .
Around those small calcareous fascicles , the body-wall is enveloped and the ccelenteron is drawn out into a labyrinth of canals ; secondarily , the body-wall winds itself upwards on the primary skeletal supports , during which process some parts may remain inextricably caught , and continue for a time only in living connection with the polyp , while others may be completely withdrawn .
In either case , the direction of the calcareous deposit , which is , in the primary fascicle , obliquely outwards from the centre , veers more and more into an upward direction , and gives rise to the long fibres parallel with the direction of growth at different parts of the Madrepora branch .
The relations of the porous wall and ccenenchyma are essentially alike in Madrepora and in Turbina , ria ; the small size of the calyces in Mctdrepora , and the diminution of the septa frequently to mere spines projecting inward from the wall , a.re features probably correlated with the branching , and at the same time twisting , growth of the Madrepora coraJlum .
Well-developed septa in Madrepora have the same structure as septa of Turbinaria . .
Above all the almost constant feature in Madrepora of the continuation of one of the primary septa to form a columella , yielding a plane of division in the polyp between bilaterally symmetrical halves , is a strong point of resemblance in the calyces , of Madrepora and Turbinaria .
The development of an efficient ccenosarcal system of canals seems to have done more towards the degradation of the calyces in Maclrepora than in Twrbinaria .
Yet there can be little doubt that both types are specially-modified reef-living Turbinolid allies , especially adapted for the fluctuations of tide and depth , and having replaced the compact wall-thickening of the deep-sea Turbinolid by a light framework of interpolypal skeletal canals .
PoRITES .
Its General Structure .
I shall be able to curtail my description of Porites by frequent reference to the foregoing pages on Madreporci and Tilrbinaria .
The relations of wall and septa are \#183 ; the same in Porites as have been described for those two types .
Centres of calcification a.re continuous in wall and septa , and the deposit is laid down in both at the same hotizon of growth .
The shape of the wa11 in Porites is more zig-zag in form than in 1'urbinaria or Madrepora ; in other words , the mural bars are not exactly tangential to the septa in position , but form acute angles with them .
The other ends of the mural bars meet usually with , sometimes without , the intervention of a cost.a .
The costa then props the body -wall at that part during the course of a SYSTEMATIC STUDY OF THE MADREPORARIAN TYP:1\#163 ; S 01\lt ; ~ CORALS .
2Hl growth-period ; at the close of the period a new pair of mural bars develop , and so on .
The pores of the wall in Porites are therefore produced according to the same process of growth as in Turbinaria .
A comparison of the ground-plan of Porites ( fig. 64 ) with that of Turbinaria ( fig. 55 ) , will show the similarity in the structural Fig. 64A .
Fig. 64B .
c. th .
col .
p. s. th .
c. Ground-plan of two .
different types of Porites .
th .
, theca , zig-zag in shape ; c. , co3ta ; s. , septum ; p. , palus ; col .
, columella .
plan of the two calyces .
This is especially clear with regard to the calycinal pali in both forms .
In Turbincwia , the pali grow upward from the fused ends of a primary and a secondary septum , the same is true of Porites .
But the pali in Porites are small and round and not lamellar , hence a transverse section may cut one , two , or even three pali in the same radial line , the individual pali having taken origin successively and attained various stages of growth .
As explained under Turbinaria , subsequent calcareous deposit may block up the spaces originally present between the pali and septum in one radial line and give the appearance of one continuous compact radial structure extending from the mural area to the centre of the calyx .
Then the radial structure looks very like an Astrreid septum , and the resemblance is still more heightened by the fact that the individual pali appear on the lateral surfaces like ridges jutting out from the compacting calcareous tissue .
On the other hand , even after a secondary deposit of calcareous tissue has been laid down , it may happen that apertures are left unfilled between the septum and pali .
Thus , in systema.tic descriptions , the radial structure in -Porites has usually been regarded as perforate , and PRATZ placed the Poritinre , E.H. ( loc. cit. , 'Palreontographica , ' 1882 , pp. 122-123 ) , amongst the Microsolenid members of the Pseudoastrreinre .
According to my examination , the septum of Poritidm is small , either rounded or lamellar in shape , and its relation to the pali is precisely the same as the relation observed in Turbinaria , or that which may be observed in pal\#163 ; -bearing Turb\#163 ; nolidm .
Moreover , the zig-zag nature of the wall in Porites is paralleled by the genus Pctracyathus and others belonging to the Trochocyathacere , E.H. Tangential expansion of the mural area is again in Po1\#183 ; ites , as in Tu1\#183 ; bina1'ia and Madrepora , accomplished by the introduction of new costal invaginatio11s ( see fig. 64 ) .
It will be seen from the ground-plan of the calyx that any two mural bars coalescing 2 :r 2 with a radial structure bear the same relatio~ to it as any two adjacent\#183 ; septa bear to the palus between them .
The tangential expansi~n of the calyx follows in fact the S[Ltne system throughout , and a mural ring of thickening might in some cases arise round the pa. Ii , in which case the " septa " would be looked upon as " costre .
" The costre present between adjacent polyps are placed opposite every a1igle of the zig-zag mural ring .
Secondary thickening not only tends to solidi(y the wall , but frequently produces iri addition an outer ring of thickening , halfway between adjacent polyps .
Its }tficroscopic Structure .
Figs. 65 and 66 are made from the uppermost portions of the corallum of Porites incrustans ( DEFR .
) .
Looking at a transverse microscopic section , first with hand lens only , a loose tissue is seen in which radial lines of structure are much less pronounced than in Astrreids .
The cause of this is that we have not in Porites one continuous broad calcareous structure , but a series of compound trabeculre varying little in size and shape .
It may be observed ( 1 ) that the series is apt to deviate from a definite radial line , ( 2 ) it is thinner and thicker , much like a string of beads , ( 3 ) between adjacent radial series numerous cross-bars ( synapticulre ) are present , whose fine structure exactly repeats that of the septal trabeculre themselves .
It follows that such a section represents a closely-knit , spongiose skeletal ti8sue , built , however , on a regular trabeoular system .
The network of the corallum is all the more striking , since the skeletal parts of the radial series grow vertically upwards , and the synapticulre are placed horizontally between them ( cf fig. 65 ) .
\#183 ; Stronger magnifying power shows short , dark axes of calcification , both in the radial series of trabeculre and in the Rynapticulre .
Fascicle~ of fibres are gathered around the axes of calcification and radiate horizontally outwards to the surfaces of the skeletal parts ( cf fig. 66 ) .
The synapticulre are seen in transverse sect.ion to have almost the same areal dimensions as the cross-sections of the trabeculre .
The longitudinal section ( fig. 66 ) shows the general arrangement stated above , of vertically-directed trabeculre and synapticulre placed at right angles to them .
The position of the short , dark axes of calcification corresponds to the different directions of growth ; those in the trabeculre of the radial series being vertically placed , in apical succession , those in the synapticulre being horizontally placed , :1nd the whole synapticula including one axis of cal ?
ification of equal length with any one of the axes of calcification in the radial trabeculre .
The fascicles of fibres in the radial trabeculre lie horizontally , since they are placed at right angles to the axes of calcification .
In the synapticulre , the fascicles are directed also at right angles to the axes , that is , they radiate upwards and downwards in the section .
It may be concluded , from a comparison of the\#183 ; transverse section of the synapticulre , that in the actual synapticular bar , the fascicles radiate in all directions , from the axes of calcification to the rounded surface .
SYSTEMATIC STUDY OF THE M.ADREPORARIAN TYPES OF CORALS .
~21 Fig. 65 .
cn.p .
th.p .
th .
cn.p .
Transverse section Porites incrustans .
th .
, theca ; th.p .
, thecal pore ; cn.p .
, coonenchymal pore .
Pali are seen in front of septa , or in various stages of fusion , Fig. 66 .
c.c. f. s. .
, yn .
s. syn .
s. Lono-itu linal section showm( ) ' the precise similarity of structurP- in septa ( s. ) , and synapticulro ( syn .
) .
0 I 0 In both , the fascicles ( / .
) bush out circularly around the calcification centres ( c. c. ) .
It is readily demonstrated in the sections that the individual fascicles are essentially the same in origin and structure as the fascicles in other types , for example , Galaxea .
Alternating light and \lt ; lark bands cross the fascicles and indicate successive growthlamelke .
The chief difference is that in Galaxea the long axis of the fascicle is mainly in the\#183 ; median plane of the septum , the uppermost end fascicle then bending out laterally towards a granulation .
Whereas in PoriteR , the fascicle takes merely its origin in the middle line , its whole length is in a rectangular direction .
The darkerlooking " centre of calcification " in Porites represents the organic remnant , not of one fascicle or pair of fascicles , originated during a single growth-period , but of a number of fascicles formed radially around an axis during a single growth-period of the trabecula .
So much of a trabecula as occurs between two successive synapticnlro represents one growth-part of a radial structure .
The interesting relation is therefore demonstrated that all the radial skeletal structures of Porites , pali , septa , cost(E ( all of which are here merely special names applied to trabecul(E of the same structure , according to their particular position in the calyx ) , as well as the synapticul(E , are composed of growth-segments prncisely similar in size , shape , and structure .
Each synapticnla is one ho1\#183 ; izontal growth-part , developed for the most part contemporaneously with the vertical growth of the radial trabeculre .
This is shown by the position of the synapticulre with reference to the series of growth-parts in the trabeculre .
Each new synapticula is interposed at the interval between two successive growth-parts in a trabecula , and during their simultaneous growth the fibres become mixed at the coalescent areas of the horizontal and vertical deposit .
I look upon the cross-bars of Porites as tangentially placed growth-parts of equal rank with radially placed growth-parts , meriting the name of " syuapticulre , " inasmuch as they are regularly disposed along the interradial loculi and coalesce with the radial Rtructures .
Yet , unlike the " mural bars , " they are not placed opposite each other in a\lt ; ljacent loculi so as to form a continuous ring of thickening .
Its Systematic Relationships .
If we now compare Po1\#183 ; ites with Eupsammia , a certain resemblance obtains , which is due merely to the fact that the growth-parts of the septal trabeculre are clearly defined throughout the whole corallum of Porites .
This gives regularity to the general structure , and , when secondary deposit renders the septa and pali apparently one continuous surface , partly porous , partly compact , the resemblance to the Eupsammia septum is very great .
A comparison of the microscopic sections , however , shows a very different fibrous arrangement within the trabeculre , that of Enpsammia being undoubtedly much nearer the Fungidre than it is near Porites .
The latter agrees with Turbinaria and Maclrepora and differs from Eupsammia in the contemporaneous time-relations of the development of mural and septal structures , ltnd in the absence of an " area of divergenc~ " i:\ , t the periphery of the septum .
SYSTEMA'l'IC STUDY OF 'l'HE MADREPORARIAN 'l'YPES OF CORALS .
223 Again , comparing Porites with any typical genlls of the Pseudoastrcm\#183 ; nce , to which group the Poritirne were referred by PRA .
Tz , several features of difference are presented .
( 1 .
) The fine structure of the trabecula is different , in so far as the fascicles of Porites are more horizontal in position , and spring from different horizons of the individual ' ' growth-parts , " instead of radiating out like a bush from the base .
( 2 .
) The Pseudoastrreinre have their trabeculre in a definite radial line corresponding to a continuous radial fold of the invaginated body-wall .
( 3 .
) The canalicular system produced by trabeculre and pseudosynapticulre in the Pseudoastrreinre is solely of secondary origin , the pseudo-synapticular bars depending on elongation of the trabecular fibres during withdrawal of the soft parts .
In Porites , on the other hand , the canalicular system .
is determined at once by the primary deposit ; septal and synapticular ''growth-parts " form simultaneously at the uppermost horizons of the coral .
If we could eliminate the factor of secondary coalescence , we should have in Porites a primary reticulate skeleton , of which the individual pieces had been .
enveloped in polypal flesh , firmly envf'loped , in de.finite 1\#183 ; elative positions crosswise and lengthwise .
It is the perfection of this primary reticulate skeleton which gives Porites the palm among the " Perforate " Madreporarians .
I regard it as the most st.rongly specialized generic type among reef-builders .
In spite of transitional forms among the Poritinre and the ancestral type presented by Litharma , it seems only right to place Porites and its congeners systematically in the rank of a special family .
The Poritidre need not be far removed from the Turbinarinre and Madreporinre , and all three must be looked upon as mixed types , at any rate related in polypal characters with various members of the compact Turbinolidre .
CHAPTER VJII.-GENERAL REVIEW OF MACROSCOPIC AND MICROSCOPIC SKELE'r AL FEATURES .
CONCLUSIONS RELATIVE TO CLASSIFICATION .
The above comparative study of the skeletal parts and of their microscopic structure indicates that the sub-kingdoms of Madreporaria aporosa and M. pe'r-Jorata \#183 ; recognized by EDWARDS and HAIME , draws an arbitrary live of sub-division .
If we were to ba8e sub-kingdoms of the Madreporaria on the compactness or porosity of \#183 ; the skeletal structures , the families Turbinolidre and Astrreidre would alone* be " " Iu giving tho following summary , I limit my remarks here to the families whioh I have myself examined ; the Stylophoridce , E.H. , and Ociilinidro , E.H. , for example , are not specially discussed , although thoy are families belonging to the M. aporosa , E.H. The Damiidw may be regarded as degraded forms of Turbinolidre .
correctly called Madreporaria aporosa , while the great majority of the Fungidre would have to be included , \#183 ; together with the Madreporidre , E.H. , and the Poritinre , E.H. , as Madreporaria perJorata .
But such sub-kingdoms would be also entirely artificial .
Certain genera ( e.g. , Microsolena .
) , placed by EDWARDS and HAIME among the M. p erforata ( sub-family Poritinre ) , were removed by PRATZ to the Pseudoastrreinre as a sub-family of the Fungidre .
The same author showed that , in spite of the regular fenestration in the original laying-down of the radial structures in those genera , the characten , of the sub-family , as a whole , were intermediate between the Astrreinre and the Agaricinre ( Lophoserinre , E.H. ) .
Although PRATZ rightly appreciated the main features of the Pseudoastrreinre , he\#183 ; erroneously included several genera , e.g. , Haplarcea , Thamnar\#183 ; cea , Porites , Siderasfrcea in side-branches of this sub-family .
The Jurassic and Cretaceous genus Haplarcea may be regarded as an ancestral type of the simple Eupsammids .
In Haplarcea the union of adjacent septal surfaces is both pseudosynapticular and synapticular .
A certain tendency to ~orm septal groups is also apparent , but is not a definite feature as in the Enpsammids .
In the latter the synapticular union is mostly restricted to the peripheral area , while nearer the centre of the calyces the pseudosynapticular union ( i.e. , union of adjacent granulations ) is often replaced by the closer coalescence of the actual ~eptal surfaces at nongranulate parts .
This close coalescence it is which produces the septal groups , often symmetrically arranged , in the Eupsammids .
Similar septal grouping is characteristic of the families Fungidre and 'Lophoseridre , both of them likewise " synapticulate " .
families .
The ancestral Jurassic and early Cretaceous types of these three families are naturally less specialized than the later types , and such forms as Haplarcea are important witnesses of the originally common features of the stock .
From the point of view of the pala:\gt ; ontologist the mos't important " commo~ feature " would be expressed as the rich development oj both synctpticulce and dissepiments .
From the zoologist 's point of view the same feature could be expressed as the development of canaliculate downward prolongations of the visceral cavity of the polyp-within the polyp itself in simple forms , and both within and withoitt the polyp in colonial forms .
At the same time these canaliculate prolongations are definitely confined to the interseptal and intercostal loculi , i.e. , they occurred in distinct lines radiating outwards -from the central parts-of the polyps , and extending ( in colonies ) without interruption into the camosarcal or interpolypal areas .
It is well known that no such canaliculate extension of the ccelenteron is present in \#183 ; the h1eshes of the skeleton in the\#183 ; Astrrei\lt ; lre and Turbinolidre , nor indeed in the Oculinidre or Pocilloporidre .
But it might be thought that the Madreporaria perforata of EDWARDS and HAIME had a similar system of visceral canals .
This is not the case .
Excluding the Eupsamminre , E. H. , from the M. perforata , E. H. , there remain the Madreporinre and the Poritidre .
The typical genera of the Madreporinre SYST .
EMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
225 have been discussed in the preceding pages .
Within the polyps there are no long cann1icular prolongations , but a continuous deep visceral chamber .
In the camosarcal areas there is a richly-developed canalicular system , which , however , is not restricted within definite uninterrupted radial lines , but constantly changes the relative position of its ramifications .
With regard to the Poritinre , E. H. , I can say less , as no sufficient anatomical research has been made ; to judge from the skeleton they seem to have a canalicular system equally developed within and without the polyp , but nowhere restricted to definite radial limits-in short , a system more complex and more highly differentiated than in any Eupsammid , Fungid , or Lophoserid .
Thus the canaliculate system in the skeletal meshwork of the Madreporinre , E. H. , and Poritinre , E. H. , is not homologous with that which characterises the Fungidrn and their allies ( cf Part II , pp. 261-263 ) .
The same end is reached for the vital processes of the coral polyps , but by different means .
A " porous wall " is a necessary part of the means adopted by the Madreporinre and Poritinre , and is always developed contemporaneously with the septa in these groups , and according to one unvarying plan of structure .
On the other hand , in the Fungidre and their allies , the radial structures ( continuous septa and costre ) are first formed , and they determine the subsequent development of synapticulre .
At some particular part a wall-thickening is formed , partly by the septo-costre , partly by the synapticulre .
But a wall is an afterthought , as it were , in these corals , and mn , y be entirely absent ( ex. Pseudoastrreinre ) , irregularly and incompletely developed ( ex. Funginre ) , regularly porous ( ex. Funginre ) , compact or porous ( ex. Lophoseridre ) , regularly or irregularly porous ( ex. Eupsarnmi\lt ; lre ) .
The two important skeletal features in the Fungidre and their allies are therefore ( 1 ) the continuity of the radial structures from the centre of the cal ices into the costal or inter-calycinal areas ; ( ~ ) the free upward growth of the radial structures previous to the faying down of any wall deposit .
These two features are the correlatives in the skeleton of an anatomical feature in the living polyp , to which zoologists have lately been devoting more attention , viz. , the presence of an edgezone ( " Randplatte " ) round the polyp .
The edge-zone is the continuation of the interradial locu]i of the polyp above and beyond the lip of the calyx .
The rnesenteries of the loculi are continued into the edge-zone and support the oral surface of the polypal body wall .
Naturally this continuation of the soft parts of the polyp into interpolypal areas is almost inevitable when the radial structures rise up prominently from the wall-thickening as " confluent " septo-costre .
On the other hand , if the wall is contemporaneously developed with the septa , or developed before the septa , as it is in some Madreporas , there is far less chance of the rnesenteries of the polypal loculi being continued above and beyond the wall .
Then it is-in the absence of distinct interradial canals outside the polyp , and of mesenterial support for the oral body wall-that an irregular C(J3nosarcal system of canals develops , and a light skeletal framework supports the aboral body-wall of th0 polyp and pushes it upon costal spmes or teeth against the oral body-wall , thus supporting it also ( cf. fig. 5 6 , p. 209 ) .
We find , therefore , that one of the most characteristic features in the anatomy of the colonies of Turbinaria , Madrepora , as well as colonies of Stylophora .
, Pocillopora , , Seriatopora , is the absence of an edge-zone round the individual polyps.* If we compare the Astrreidre and the Turbinolidre in respect of an edge-zone , we find that the Astrreidre , like the Fungidre and their allies , always have an edgezone round the individual polyps of colonies , except in such colonies where the adjacent polyps are in direct communication with one another above a common calicinal wall ( ex. Goniastrcea ) , whereas the Turbinolidre have primarily no edgezone .
Only after a polyp has attained its maximum of height does an edge-zone begin to develop in certain genera .
Then also the radial structures grow upwards freely from the wall , and septo-costal lines of structure divide a limited edge-zone into narrow intercostal loculi , corresponding with the interseptal loculi of the polyp .
I have frequently pointed out that the relations of the septa and wall in the skeleton of the 'l'urbinolidre are identical with those of Madrepora , Turbinaria , and Porites .
They are also identical with those in Stylophora , Pocillopora , and Seriatopora .
On the other hand , the relations of the septa and wall in the skeleton of the Astrreidre ( ident .
Astrrninrn , E. H. , vide antea , chap .
iii .
) are the same as those in the Fungidre , Lophoseridrn , and Eupsammidre .
The Oculinidrn show extremely variable relations of the septa and wall , and I have not yet had time to make a special study of this group .
The majority of the Oculinid genera seem to develop septa and wall contemporaneously at first and then the septa outgrow the walJ .
At the same time the polypal edge-zone extends in its proportions .
The genus Lophohelia , examined by FOWLER ( Q. J. M. S. ) is a very good example .
It shows , as described above for the Turbinolidrn , how first the primary septa and then the snialler cycles gradually grow higher than the wallthe centres of calcification , originally continuous in septa and wall , becoming in * FOWLER has shown the following genera to be without an edge-zone , and has pointed out that all possess a well-developed coonenchyme :-Madrepora , Turbinaria , Stylophora , Seriatopora , Pocillopora , Heteropsammia multilobata .
Other genera which have been proved to possess an edge-zone al'e .Astroides , Rhodopsammia , Dendmphyllia , Fungfo , Oladocora , Oaryophyllia .
FOWLER inclines to think the possession of the edge-zone the more primitive condition , while its absence and the development of camencbyme is a secondary condition .
The " edge-zone , " as at present understood , is in my opinion itself a " secondary " condition ( cf. p. 262 ) and arises in independent types .
It may well be , however , that in the Eupsammidre , the presence of a pseudotheca and of a well-developed edge-zone is the primitive condition , and that retrogression of the edge-zone may bring about the appearance of camenchyme characteristic of certain forms ( Heteropsammia ) .
This is possible , but the anatomy and skeletal i::tructure of Heteropsammia is not yet known in detail .
In all the typical ccenenchymatous forms just quoted , there is no foundation for supposing retrogression of an edge-zone ( cf. pp. 265 , 312 ) .
On the other hand , the ccenenchyme ( = " peritheca , " E. H. ) of Galaxea is very possibly a retrograde form of tho primitive edge- ; , .
; one of Stylinids , S'fSTEMATiO STUDY OF THE J\IIADREPORARIAN TYPES OF CORALS .
227 consequence interrupted .
Galaxea shows similar variations ( cf p. 121 ) .
The septa in such types have the tendency also to be toothed at their upper edges , corre1ati ve with their freer upgrowth and exsertness above the wall .
The skeletal relations which I have observed in Galaxea led me to unite that genus with the Stylinidre , a family which from its earliest origin has shown considerable resemblance to the Oculinidre .
The fossil group of colony-forming genera allied with Stylina show also variable relations of septa and wall .
Very often they seem to be more akin to the Astrreidre .
Yet the microscopic structure of the septa , the character of the basal structures , and the presence of an epithecal deposit round each individual polyp in several genera , indicate also a near relationship with Galaxea and with the less branching of the Oculinid types .
For the complete proof of these relations a detailed examination of various types_ of living Oculinids is needed ( ref .
" Stylinid " type , fig. 35 ) .
We arrive at general systematic results by the careful study of the growth-relations observed in different parts of one polyp , or one colony , as the case may be .
Throughout my description of the foregoing types I have frequently emphasized the correlation of the detailed microscopic structure in the various skeletal parts with the relative time of their development .
Thus , differences in the microscopic structure go handinhand with the differences in the relative time of development of the radial , tangential , and basal structures .
The most highly differentiated radial structures are those of the Astneidre and Fungidre .
Each septal and costal ridge in those types is a firmly-knit group of trabeculre arranged with reference to an ideal axis in the ridge .
The whole series of septal and costal ridges are again definitely arranged in a fan-shaped group with reference to an ideal axis in the radial structure .
The firmness thus attained by the radial structures is correlated with the absence of tangentially placed thecal structures .
The septa grow up from the embryonic " basal plate " withoitt any thecal snppott between them .
This is true of the Astrreidre , Fungidre , Lophoseridre , and Eupsammidre .
The septa themselves develop in most cases a " pseudothecal " thickening , which forms a lateral connection between them .
In others , interseptal pieces are also introduced in the calyx , either in fragmentary fashion ( synapticulre ) or continuously ( the synapticular bars of Fungia ) .
In other ca.c ; es no pseudothecal thickening is apparent , and the calyx may be said to be entirely without a theca .
At the same time the retreating lappet of the polypal disc lays down outside the ealyx a fragile epitheca , which completely encircles the septa and costre , and has no structural connection with them .
This epitheca is formed always , whether a pseudotheca is present or not .
Its structure is more brittle than the ca1icinal skeletal structures , and there never are in it " centres " of deposit , for the simple reason that the part of the polypal wall which deposits it is never pouched up many wn.y .
Montlivalti\#183 ; a , for example , is an Astrreid genus and Microsolena is a Fungid ( Tham228 DR. MARIA M. OGlLVIE ON TE'E MICROSCOPIC AND nastrreid ) genus with an epitheca and no pseudotheca .
Both these are examples of simple corals .
Thecosmilia , Cladophyllia , and other branching Astrreid colonies have an epitheca round each branch , and seldom develop a distinct pseudotheca .
In massive Astrmid colonies the epitheca envelops the whole colony , and is not present around each individval polyp .
Usually the individual polyps have pseudothecre ; again , among the typical Pseudoastrminm , no pseudothecal wall is developed between adjacent calyces .
Among the Stylinidre genera occur in which the calyces have each a pseudotheca and an epitheca .
The latter is common to adjoining calyces .
The pseudotheca and epitheca go in these cases usually under the name of an " inner " and an " outer " wall respectively ( vide below , p. 262 ) .
It is noteworthy that throughout all those families in which a " pseudotheca " occurs , the pseudotheca occupies a position , as it were , within the calyx .
To one side of it are septa , to the other costre , in the usual accept ation of these terms .
As , however , the septa and costre ara continuous , I have used the more general term " radial structure , " and the pseudothecal ring of thickening may occur at any part of the radial structures-near the periphery , ex. Mussa ; nea1 : the middle , ex. Heliastrma ; or at the very periphery , ex. Goniastrma .
In the last case , the epitheca of the whole colony closely invests the pseudothecre of the calyces which happen to be at the outer rim of the colony .
In all other cases the pseudothecre are free structures , not coalescent with an epitheca , and practically within the calyces .
In the living polyps the part of the polyp which extends beyond the pseudotheca is called the edge-zone of the polyp ( " Randplatte " ) .
These remarks may be said to exhaust the possible relations existing between septa , " wall"(= pseudotheca ) , and epitheca in the above-named families-Astrreidre , Fungidoo , Lophoseridre , and Eupsammidre .
The fundamentaJ feature in all of them is that the embryonic basal-plate does not develop farther .
The ectodermal folds in the embryo are solely radial folds , and , as they grow upward , septa a.re developed within them .
No tangential folds rise up simultaneously .
The first connection between septa is made when the first , floor of dissepiments , or the first series of synapticular bridges is laid down in the interseptal loculi by the basal parts of the polypal wall .
A marked contrast is presented by the families Turbinolidre , Madreporidre , Poritidre , Pocilloporidre , Oculinidre .
The mature calyx in them always possesses a theca .
It will be remembered that " theca , " as distinct from " pseudothecci , " indicates a peripheral wall in which tangentially-placed lamellre occur .
These are deposited quite independently of the septa , in special wall-folds of the polypal flesh .
The wall-folds form and grow upward , simultaneously with the septal folds .
Thus the centres of calcification are continuous in wall and septal folds .
The septal structure is the same as the thecal structure , except that it is placed radially in the calyx , and the calyx is placed tangentially .
One may even see on the lateral surfaces of the septa , fine , fragile , transverse folds , which exactly correspond to the " epithecal " folds SYS'rEMATIC S'l'UDY OF THE MADREPORARIAN TYPES O.b ' CORALS .
229 outside the theca .
From the very nature of its growth , the " epitheca " in Turbinolidre is not separable from the theca , jt is in other words the outermost part of the theca .
Its place is taken by the crnnencltyme between adjacent calyces of the colonial families , Madreporidre , Poritidre , etc. ( see below , p. 267 ) .
The concentric folds of epitheca forming the outer surface of the Turbinolid .
theca , are structurally the same as the concentric skeletal rings around the naked septa of Montlivaltia , Microsolena , etc. Hence it is desirable to retain the term epitheca for both , although in these Astrreids and Fungids the " rings " in question are not placed " upon a theca " but on the outer edges of the crowded , numerous septa .
It is readily seen that where a theca is developed Rimultaneously with the septa , there is no need for the building up of a pseudotheca .
And a free pseudotheca is never present in these thecate families .
Still , the continued deposit of growth-laminre on the inner surfaces of the theca and on the lateral surfaces of the septa is sometimes so great that it fills up the interseptal loculi near the periphery , and lateral surfaces of the septa come into contact .
We may say then that the characteristic condition in all these families is the presence of a theca in structural union with the septa and of an epitheca as superficial rings on the outer surface of the theca .
Variations , of generic value merely , may occur in the relative development of septa , theca , epitheca .
The septa have a tendency near the lip of the calyx to grow more rapidly than the theca , and the theca in its turn to grow up without much in the way of epithecal rings .
This tendency on the part of septa and theca is correlated with the development in the mature polyp of an edge-zone in those particular genera .
Otherwise an edge-zone is absent in those families .
Before going further , I shall show diagrammatically the main points in which I agree or disagree with the classificatory system of EDw ARDS and HAIME .
The classification given by those authors is as follows : Madreporaria aporosa .
ED .
and H. ( Lo .
cit. , 'Corall .
, ' IL , p. 7 ) Dasmidre I ITurbinolidre .
I Oculinidre !
~ Stylophorinre ( Echinoporinre J Astrreidre See foot-note , p. 223 .
Merulinacere ( foot-note , p. 186 ) . !
_Fungidre J !
:~::::.~:= } Madreporidre .
E. H. Madrepor~ric~ perfor~ta .
ED .
and H. Tur~~narinre . .
_ ( Lo .
cit. , Coral .
, III .
, p. 90 ) l ' Pontmre .
E. H. ( = Pont1dre Dana ) .
Montiporinre ( included by KLUNZINGER m Madreporidre ) .
The Seriatoporidre , E. H. , originally grouped by EDWARDS and HAIME under " .M . .
tabula ta , " have since been removed to M. aporosa .
The chief Madreporarian families enumerated by EDWARDS and HAIME under these two main sub-divisions are represented in the following diagram :This we may take as the basis of classification in Madreporaria , although in the course of time several important alterations have been made .
I have represented it in circular form for readiness in comparison with my own general results .
In the next diagram I follow KLUNZINGER in placing the sub-family Poritinre , E. H. , in the rank of a family , in accordance with DANA'S classification .
My reasons are founded on the characteristic microscopic structure of the skeletal parts .
Again the genera grouped by PRATZ as Pseudoastrreinre regulares are recognized as a special sub-family Thamnastrreinre under the Fungidre DANA .
And I propose to remove the sub-family Eupsamminre , E. H. , entirely from the Madreporidre , E. H. , and place it as an independent family next the Lophoserinre , E. H. I have considerably altered the sub-families Eusmilinre , Turbinolinre , Caryophyllinre of EDWARDS and HAIMl~ for rea8ons into which I shall enter more fully in the second part of this paper .
The fissiparous Eusmilinids ( type : Euphyllia , p. 161 ) are placed under the new family of .Amphiastrreidre ( cf. p. 334 ) ; the Trochosmilia group are included under the Turbinolidre , and the Stylina group are erected into a separate family .
In the accompanying circular diagram I have entered the Madreporarian families which are represented in Recent time .
SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
231 , ,--- --------- ... ... ... ... . .
, ' ' ' ' ' ' ' \ ' ' \ \ It will be seen that I do not accept the main sub-divisions of Recent Madreporaria made by EDWARDS and H. AIM , viz. , the M perforata and Aporosa .
My reasons have been already stated .
The simplest polyps are those belonging to the Turbinolidre ; in the Pocil1oporidre , Madreporidre , and Poritidre the polyps remain simple , the calycinal structures of the two first-named agreeing with the Turbinolidre in microscopic details .
The extrapolypal flesh lays down a camenchyme of various form .
The Amphiastrreidre and Oculinidre show a greater complexity in the microscopic details of the calycinal skeletal structures , and the form of the polyp is distinctly intermediate between the typical Turbinolid and the typical Astrreid .
The Astrreidre are distinguished by strongly-marked regularity in their processes of growth .
The septa are built up step for step ; one floor of dissepiments is laid down after another until the whole corallum presents one of the most striking examples among the invertebrata of a perfectly-constructed shell .
The Fungidre and Eupsammidre can only be properly studied when the details of their structures are compared , part for part , with those of the Astrreidre .
They agree with the latter in the general plan of structure in septa and wall ; their distinctive feature is in the frequent synapticular union of septa in the middle of the calyx , i.e. , above the horizon of the floor of the polyp .
This introduces irregularity in the arrangement of the soft parts of the polyp , and the uniformity so remarkable throughout the Astrreids is entirely wanting in those two related families .
I shall conclude Part I. by giving a table , iu which the classificatory features of the main Madreporarian families can be readily compared with one another .
But , before doing so , it will be necessary to recapitulate the chief features of structure in the Madreporarian skeleton , and the meaning of certain terms which I have introduced or made use of throughout the first part of this paper .
Summary of Skeleton Formation .
The Madreporarian Skeleton is a complicated shell wholly built up of successive calcareous laminre or lamellre .
These laminre are virtually growth laminre , the equivalent of so-called " Anwachsstreifen " in the molluscan and other shells .
The Calcareous Laminm of ll{adreporaria are composed of coalescent calicoblasts which gradually lose their cell-wall and organic contents as calcification proceeds .
The Calicoblasts are ectodermal cells in which , by a process of chemical change , lime is secreted and at once laid down within the cell in the form of a more or less connected group of aragonite fibres .
An organic remnant of the cell may be left , especially at the part of the cell remote from the polyp and next the skeleton .
Alternate bands appear in each lamina , a narrow band , dark by transmitted light , representing the proximal organic ends of calicoblasts , a broader lighter crystalline band representing the fibrous deposit within adjacent cells .
The breaking down of cell-walls and cellular tissue leaves often an irregular intermixture of disintegrating carbonized products in the laminre .
At the upper edges of all invaginations of the polypal wall , septal and mural , growth and cell-multiplication takes place very rapidly .
There the calicoblasts are rapidly calcified and shed off from the living ectoderm .
In those calicoblasts a relatively large amount of cell material is present in unchanged condition .
It may disintegrate quickly or slowly , and gives rise to various appearances in microscopic sections .
The upper edges of the septal invaginations necessarily correspond to the median plane of the septum , and the presence originally of a larger amount of organic material at this part causes it often to have a somewhat different appearance from the rest of the septum .
One of the most frequent appearances is that of a " dark line " in the middle of the septum .
This line is in many cases a fissure filled up by carbonized debris , in other cases it is the first-formed laminre in which the large mixture of organic cellmaterial has given rise by disintegration to impure , mixed products .
The actual fibrous deposit within the calicoblasts is also in some genera denser in the early deposited growth-laminre than in those laid down at later periods of growth , and this fact alone causes a greater opacity at that part of a thin section .
Again , the growth-lamime laid down at the edge are sometimes very smooth , and those of later growth-periods are less uniform , the cells being active only at certain areas of the lateral surfaces of deposit .
This local activity of the calicoblast cells of the side-flaps produces granulations on the septal surfaces .
Sometimes the early growth-laminre are not smooth .
That is , their calicoblast cells form active groups of deposit , and granulations are marked out at the first laying-down of the septum .
These differences are of minor importance .
The chief point is that organic cellscalicoblasts- are converted .
into small groups of fibres , and that as the cell-walls SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
233 break down , the small groups of fibres coalesce and form a calcareous layer which may be compared with a layer of living cells .
Following the stages in the building-up of an Astrreid septal segment:(1 .
) It is laid down at the upper edge of an invagination of the aboral body-wall of the polyp .
During one growth-period some five or six laminre are laid down by both of the invaginated flaps , and the septal segment has a thickness of 10 or 12 calcareous laminre .
( 2 .
) A second period of growth finds the same septal segment no longer at the edge of the invagination , but moved as it were one tier below the edge .
Deposit goes on less actively , granulations become more marked .
The calicoblasts are , in fact , developed more fully at some parts of the invaginated flapB than at others , and the lamina bulges out at those parts .
( 3 .
) A third period follows in which local activity is still more pronounced , a fourth and fifth period follows , and then , as a rule , the soft body-wall is withdrawn from the particular segment owing to the upward growth of the polyp .
The primary septa in an Astrreid usually has a large number of calcareous lamime , from 25 to 35 on either side of the median plane of the polyp .
The septa of the smaller cycles have few lamina:- , from 5 to 15 on an average .
Clearly the activity of the calicoblast deposit varies immensely , being such as to allow , say , five times as many lamime to be deposited on the primary septa than on the septa of the smallest cycle in an Astrreid calyx .
Owing to the method of growth , the successive growth-lamellre are parallel with the surfaces of the septum when seen in transverse section , but when seen in tangential longitudinal section , they form an acute angle with the surface , the laminre of the two halves meeting at a point in the median plane of the septum .
These investigations disprove the distinction which has hitherto been made by palreontologists beb , Yeen a " primary septum " and " stereoplasm .
" The structure of the primary septum has , indeed , never been described satisfactorily , although the term has been in general use as signifying the " dark line " in the median plane .
The stereoplasm has been spoken of sometimes as fibrous , sometimes as structureless .
W bile certain authors have limited its application of an " Ausfollungsmasse , " or secondary compacting tissue , which was said to be lamellar .
I find no need for all these terms , since the first lamina laid down at any part of the aboral wA]l is exactly like the last in essential structure .
And the elemental unit of structure is a calcified calicoblastic cell shed off from the living ectoderm of the polyp .
I have brought forward proofs from typical genera representing the recent families of Madreporaria that the lamina has this structure .
In many specimens the actual outlines of the cells could still be clearly traced in the growth-laminre , since they retain their organic nature after the cell contents have been entirely changed .
The changes which render the fibrous deposit in the cells of successive lamime MPCCCXCVI.-B , ~ Ii apparently continuous are purely of crystallographical nature .
The subsequent orientation of the fibres is dependent on the original form of the deposit .
For example , where the laminre bulge out into a granulate eminence , the fibres form a fascicle or bundle radiating from the centre of the hemispherical eminence .
NOMENCLATURE OF SKELETAL p ARTS .
The chief skeletal parts may be classified as follows:( 1 .
) The Radial structures-septa , costre , pali .
( 2 .
) The Tangential structures-theca , pseudotheca , epitheca .
( 3 .
) The Basal structures--tabulre , dissepiments , solid laminar deposit , synapticulre , columella .
( 4 .
) The Bxtrathecal structures-rugre , rootlets , crnnenchyme .
The Radial Structures .
In some cases different parts of the same radial structure may be distinguished as pal us ( cf Goniastr(13a ) , septum , costa .
The pal us is then an elongated septal tooth growing obliquely upwards from the inner edge of the septum .
The septum is the part of the radial lamella within the calyx , while the costa is that part of the lamella external to the wall of the calyx ( cf figs. 7 4a , b ) .
Both pal us and costa may take origin independently of the septum , alternating in position with two neighbouring septa ( cf fig. 55 ) .
In all cases , the radial structures are laid down within radial folds of the aboral body-wall of the polyp .
If the radial folds are smooth , a flat-shaped radial structure is formed with two uniform halves laid down by the opposite flaps of the fold ( fig. 68b ) .
If the radial folds are pleated , a ridged radial structure is formed , and a more or less complicated system of septal trabeculre is developed ( figs. 68c-68f ) .
The Growth-Period ; Surface Growth-Curves .
the Madreporarian skeletal structures are not added to throughout the whole existence of the polyp , hut growth takes place from time to tirrie at definite periods .
The interruptions in the laying down of the deposit are , as a rule , very easily distinguished in the skeleton .
Any single period at which the calicoblast deposit is actively formed may be called for convenience a groivth-period ( ante , pp. 121-124 ) , and the part added to a skeletal structure during one period , a growth-segment ( fig. 67 ) .
The thinner lines or curves between successive growth-segments are fitly termed growth-lines or cur\#183 ; ves ( cf fig. 3 with fig. 67 ) , SY-STEMA'rIC STUDY 0"1 !
' THE :MADREPORA .
RI.AN TYPES i:O.F CORALS .
2tl5 Fig. 67 .
fr .p .
tr .
g.c. g. c. Rad\#183 ; ial structure ( equiv. whole septum of primitive types ) marked out in types wi Lh exsert septa , e.g. , Astrreids , into so-called " septal " and " costal " wings to left and right of the area of div ergence , at which the trabeculre take their origin and diverge .
The trabecula ( tr .
) signified in the figure , is composed of seven trabecular parts ( tr.p .
) , and represents , therefore , seven successive gowth-periods .
The trabecular parts between two growth-cur1 ; es ( g.c. ) form a septal growth-segment , and represent one growth-period .
The Septal 0rowth-Segment in its relcttion to the " trabecular part " and the " fascicle .
" A septal growth-segment is the part of a septum which is added along its upper edge during a single growth-period of the polyp .
In fig. 67 one complete growthsegment is marked out from the others by means of darker shading and thicker limiting-lines between it and the segments formed just before and after it .
A single trabecula is also denoted in the figure , in order to demonstrate how it is here made up of a row of the individual parts belonging to successive growth-segments .
Each such part may be fitly called a trabecular part .
I have elsewhere shown the " tra.becular part " to be a composite of " fascicles , " and thus we arrive at the elementary " group " or " fascicle " of calcareous fibres which I have discovered to be the unit of skeletal structure throughout the Madreporaria ( cf pp. 118 and 124 ) .
In other words , an isolated needle or fibre does not exist per se in the Madreporarian skeleton as it does in related groups of the Crelenterata ; on the contrary , the calcareous fibres are formed primarily in bunches of various size and form ; each bunch is a clear structural unit , distinguishable from its neighbours , sometimes bound to them only during the life of the polyp by the fleshy substance , more often bound to them in the actual skeleton side by side--in part or wholly .
Contemporaneously-formed bunches radially placed 2H2 236 DR. MAR !
A M. OGILVIE ON TltE M !
CROSCOPIC AN : D within the calyx of the coral skeleton build up a single growth-segment of the septum .
These bunches may form one or more " trabecular parts " ; if the latter , they may be attached or unattached .
Septal segments may be classed in two varieties , which I propose to call the monaxial and the polyaxial types .
The Monaxial Septal Segment ; Septal Spine .
Here the septal growth-segment is represented by one trabeciila of a compact septum , or by an isolated " septal spine , " directed horizontally , sometimes at a low angle Fig. 68A .
'l'r .
Diagrammatic drawing of a compact septum composed of monaxial septal segments .
Tr. = trabecula from wall to centre of calyx .
The lowest trabecula is shown in section ; those above are drawn with their surface granulre .
inwards , from the periphery towards the centre of the calyx ( fig. 68A ) .
The calcareous fibres laid down during the growth-period are aggregated around one axis of calcification .
The figure represents a compact septum with smooth upper edge , the inner row of teeth being apices of the successive trabeculre .
The lowest trabecula in the figure is represented in thin section to show the continuous axis of calcification ; those above are repreF !
ented in surface view bearing the granulations .
Examples of " septal spines , ' ' with precisely the Rame structure and relative position as these " septal trabeculre , " are found chiefly within the groups of Madreporidre and Pocilloporidre among recent corals .
The Polyaxial Septa.l Segment .
Far more frequently the septal growth-segment is represented by a la1\#183 ; ge number of trabecula1\#183 ; parts at various angles , belonging to as many trabeculre attached more or less completely side by side .
Such a segment is indicated in fig. 67 , and it is manifest that here the calcareous fibres aggregate around numerous axes of calcification in the SYSTEMATIC STUDY OF ' THE MADREPORARIAN TYPES OF CORALS .
237 median plane of the septum .
According to the completeness of the lateral attachment of trabecular parts , the polyaxial septal segment is compact or porous ( cf. figs. 67 and 69 ) .
A row of erect " septal spines " in a radial line ( ex. the ancient Gystiphyllum or Triassic specimens of Stylophyllurn ) , projecting upwards from the base into the lumen of the calyx , is , in my opinion , simp]y one of the primitive forms of a polyaxial septal segment .
The Compact Septum.-The compact septum is composed either of a succession of monaxial septal segments ( ex. Turbinaria ) , or of a succession of polyc"xial septal segments ( ex. Galaxea ) , each of which is in and for itself compact ( fig. 68 ) , and is united at every point with its predecessor and successor .
The Porous Septum.-The porous septum is always composed of a succession of polyaxial septal segments , each of which is in and for itself porous .
In many cases where the septal fold of polypal flesh is deep , the originally porous septal segment may be gradually thickened and compacted .
This takes place during the several growthperiods which ensue between the first formation of the segment at the edge of the septum and the ultimate complete withdrawal of the polypal flesh from the lateral surfaces of the same segment as the polyp grows upward .
Porous septa occur only in the families of Fungidre ( cf. fig. 69 ) , Eupsammidre Poritidre .
In all of these families synapticular union of the septal surfaces is present in addition to the deposition of dissepiments in interseptal loculi .
In some subdivisions belonging to these families the septa are from the first compact ; and it is , therefore , necessary in identifying the generic features to make vertical sections parallel with the septal surface .
Such sections betray at once whether any individual septal segment was originally porous and subsequently had the apertures filled up by secondary deposit ( ex. Tharnnastrcea ) , or was originally a compact segment ( ex. Siderastrcea ) .
The porosity of the septum is a feature which is certainly of frequent occurrence in the families named , but is in nowise constant .
KLUNZINGER , and others , have shown that in many genera the septa of the first cycles are compact and the smaller septa porous .
It is to other skeletal features , such as the synapticular , and most of all , the microscopic structures of the septa , that we must look for a satisfactory means of classifying the genera within these families .
The systematic importance of the septal porosity is its correlation in these families with the canaliculate nature of the interseptal loculi .
At the same time synapticulre are formed in other families in which the septa are wholly and primitively compact .
Thus , neither the one feature nor the other justifies the erection of a special group for these families .
By virtue both of the general habit of the corallum and of the microscopic structure of the skeletal part the Poritidre are probably more closely related with the Madreporidre than with the Eupsammidre .
Again , the Fungidre are , on the one hand , related with the Astrreidre , on the other with the Eupsammidre .
Macroscopic Appearances on the Septal Surfaces and their Relation to the Microscopic Structure .
The surfaces of septa may be either marked with fine lines or striC13 , or they may be strongly ridged .
Granulations are merely a secondary feature of these septal markings .
Perfectly smooth septa do not exist to my knowledge .
Examination with a hand-lens will demonstrate in all those cases of so-called " smooth " septa , the presence of extremely fine " strire , " like these in Galaxea , fully discussed in the first chapter of this work .
Striated Septa .
The strire on the one surface and on the other of any striated septum are usually placed exactly opposite to one another , but they may also alternate with one another in position .
Each pair of strire on the opposite surfaces culminates in a common apex Fig. 68n .
tr .
'l'ype of septum in a 'l'urbinolid coral ; the strire on the surface correspond to " simple " trabeculre , whose elliptical axes give rise to a " median dark line " in transverse sections .
at the edge of the septum .
This apex is , in many cases , visible to a careful observer as an extremely fine " serra , " or tooth ; in other eases again , magnifying power must be used to demonstrate the distjnct apices of " strire .
" As a matter of fact , a great number uf genera , in which serrre are readily visible , have been described ~y former authors as smooth-edged , e.g. , several in the sub-division of Eusmilinre , E.H. It is only the precise microscopy of such genera which can permit of their proper description and classification .
Now , after the microscopic research of this paper , one may compare a Turbinolid type ( fig. 68B ) and any Eusmilinid or Stylinid type ( cf fig. 35 ) .
Comparison will show the fundamental similarity in their microscopic structure , and the contrast they present to the various types of ridged septa below .
I need scarcely repeat again how each pair of strire on the opposite septal surfaces is the external expression of a " trabecula " in which " fascicles ' ' of fibres are gathered round successive opaque-looking " centres of calcification " in the median SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
239 Types of .Astrceid Septa .
Fig. 68c .
pa. tr .
Septum with fine granulations .
The trabecula ( tr .
) shows four successive fan-groups of fascicles = four " trabecular parts .
" Any trabecula may be elongated as a palus ( pa. ) , or to form part of a " spongiose " columella .
( The two lowest trabeculre a1\#183 ; e drawn in section to elucidate the microscopic structure .
) tr.p .
Fig. 680 .
tr .
a. d. a. d. Septum with regularly placr , d , large granulations marking successive " trabecnlar parts " ( tr .
p. ) on the septal ridges .
At the " area of divergence " ( a. d. ) the granulations coalesce and help to form the pseudotheca between septum and cost\amp ; , , Fig. 68i ; ; .
s. r. C. r. sr .
fr .
gr. tr .
Septum with broad ridges studded with granulations ( gr. ) ; s. i\#183 ; .
, septal ridge ; c. r. , costal ridge ; s.r. , serr::e at the tips of the trabeculie ( tr .
) .
septal plane .
Each " centre " is , in reality , lens-shaped and the trabecular radius extremely short .
Hence , in transverse sections across a septum , the " centres " of adjoining trabeculre , which are met variously in the line of section , become fairly continuous with one another , to all appearance , and give rise to a well-marked " dark line " along the middle of the section .
Ridged Septa .
The septal ridges are sometimes broad and flat , as in the Mussa type ( fig. 68E ) .
They are composed of fan-shaped systems of fairly uniform-sized strire , easily observed on the surface .
The ridges on opposite surfaces of the septum culminate in spiniform teeth .
The teeth may be themselves finely serrated , corresponding to the striated structure of the ridge .
Granulations of very various size are scattered over the surfaces of the ridge .
Sometimes the ridges are narrow and highly bulged , as in the Heliastrma type ( fig. 68D ) .
In this case there is usually one large trabecula in the middle of the ridge , and smaller trabeculre form lateral union of the larger trabeculre in the septal plane .
The fan-shaped systems of trabeculre are not readily observable as strire on the surface , but are demonstrable in all microscopic sections .
The smaller trabeculre have lensshaped centres of calcification and bilaterally placed fibres ; the larger have the fibres radially arranged around two or more centres of calcification , closely grouped , and corresponding to the trabecular axis ( cf. fig. 24B ) .
The marginal apices of those ridges are conical smooth-edged teeth .
The surfaces of the ridge are granulate .
Coalescence of small granules takes place , and one large granulation marks each growth-segment of a ridge .
Rows of granulations are thus produced , following the course of the ridges .
Fig. 68c represents a similar , but finer type of septum , characteristic of Ji'aviacere and certain meandroid Astrreids .
When one imagines a circular group of simple trabeculre forming one compound trabecula , and the latter coming only at intervals into contact\#183 ; with its neighbours , the characteristic " trabecula.te " septum of PRATZ is obtainerl ( fig. 69 ) : Thus we pass from a compact Astrmid type to a porous Fungid type .
In all these cases of ridged septa , the " centres of calcification " are small , and the trabecular radius is relatively long , so that in transverse sections the appearance of the " dark line " in the middle of the septum is neither so continuous nor so well marked as in the case of striated Repta .
Superficially , the trabeculre of a septal type belonging to the Fungi'dm ( as above defined ) are marked by their extreme uniformity .
Fig. 681 ?
presents a compact LophoRerid type , practically the same as the Fungia type ; fig. 69 , on the other hand , a porous Thamnastrrean ( Pseudoastrreid ) type .
These two figures display again the broad , flat variety of ridge and the narrow , highly bu]ged variety { antea , figs. GSE and SYSTEMATIC S'I'UDY OF TE : E MADREPORARIAN TYPES OF CORALS .
241 68D ) .
One large granulation on the surface denotes each successive growth-part .
or growth-segment added at the apex of the trabecula during a single growth-period .
Microscopically the arrangement of the fibres is perfectly radial in each trabecula .
At the same time , distinct fascicles of fibres can be traced passing from the individual centres of calcification outward to the surface , and giving rise there to small granulations .
These then coalesce during growth and form the one large granulation corresponding to a trabecular growth-part .
The large trabecula with radial symmetry of fibres is a particular case , where several pairs of strire may be said to be arranged circulctrly around a comnwn ctxis Fig. 68F .
tr .
yr .
tr.p .
:b'ig .
69 .
tr .
Types of Fiingid Septa .
Septum compact , composed of t11ree main " fans " of trabeculre ( tr .
) shown in section at the lower part of figure .
Synapticu]ate granulations mark trabecular growth-parts .
Septum porous ; adapted from a drawing by PRA'IZ ( ' Palreontographica , ' Zoe .
cit. , Taf .
14 ) ; tr .
trabecula ; tr .
p. , a trabecular growth-part ; gr. , synapticulate granulations coalescing rn parallel lines between the pores .
instead of in plane series .
One can frequently distinguish on the surface of the trabeculre fine strire exactly like those on the surfaces of a striated septum ( ex. Galaxea ) , or on the surfaces 0f a striated septal ridge ( ex. Mussa ) .
The la1'ge , radially symmetrical trabecula is indeed nothing more than a rounded ridge , and there are aJl transitional stages between the broad : flat ridge of )J ; Jussa , with its long , elliptical outline and the squarish or circular ridges of Cyclolites , Microsolr\gt ; na , etc. The individual trabeculre of a trabeculate septum culminate at the septal edge in teeth of various form , most commonly rounded and irregularly knobbed .
Any individual trabecula is made up of successive trabecular growth-segments or trabecula1\#183 ; parts .
These " parts " may firmly adhere to one another lengthways , when they M:DCCCXCVI.-B .
2 I 242 DR. MAltIA :M .
OGILVHJ ON 'fHE MICROSCOPIC AND fonn compact trabecula , or they may be not very closely attached to one another , when the individual trabecula is in and for itself porous , as in Eupsammids .
Again the trabecular parts of any one septa ) growth-segment may be laid dovm at the edge of the septum as a compact row ( ex. Siderastrcea ) , or as a porous row ( ex. Microsolena ) .
In the family of Eupsammidai , the pores in the radial structures are of extreme irregularity , for the reason that they occur originally both in the individual tra.beculai and in the septal segment .
Granulations .
Great variety obtains in the actual form of granulations .
Many of the septa which have hitherto been described as having smooth surfaces are in reality studded all over with minute pin-points of granulations , whose microscopic Ettructure may , according to my investigations , tally with much larger and more prominent-looking granulations .
On the other hand , I have found that all the large granulations have not one and the sctmc microscopic struct1irc .
In short , the word " granulation , " as at present used in Madreporarian literature , has received no exact definition .
It is merely a general , descriptive term used for any kind of hump on the septal surfaces .
Whatever the form of the hump-hemispherical or irregular , smooth or knobbed , tooth-like or conical the name applied is always the same .
One or two noteworthy attempts have been made to make some systematic nse of the occurrence of these granulations ( I refer to EDWARDS and HAIME and to MILASCHEWITZ ) , but these attempts have confined themselves to the positio\#183 ; ri of the granulations on the surfaces , in short : to the occurrence and external anangement of prominent granulations , and have overlooked all cases where the granulations were not of i:mfficient size to obtrude themselves on the observer 's notice .
I should define a granulation as a protiiber\#183 ; ance on any part of the skeletal surface , caused by the emergence at that part of one or more fascicles of fibres .
Such a definition would include the so-called echinulai on various types of " crenenchyme , " and the adornmeti.ts on the wall and costre of many Turbinolids , as well as the " granulations " on the septa .
The fundamental fact in the definition is the relation of the granulation in all its forms to fcu ; cicles of fibres-and , therefore , to one or more definite centres of active calcijfoat-ion , already fixed at the laying-down of the very first growth-lamella at that part of the skeleton .
It will be remembered that the " growth-lamella " is simply the term I have applied to the completely calcified layer of " calicoblast " cells .
Hence we may describe the building-up of granulations in terms of the calicoblast .
A group of actively-dividing calicoblasts , at any part of the ectodermal layer of the polyp or crenosarc , will give rise ultimately to a corresponding group of calcified cells , whose fibrous elements arrange themselves with respect to a definite centre or centres of calcification .
Each new calicoblastic layer behaves in the same way until a thick 6helly growth has accumulated over the original centre or centres of active calciSYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
243 \#163 ; .cation .
This growth forms the protuberance known as a _qranulcttion .
If there be but one centre of calcification in the group only one fibrous fascicle is formed , and the granulation looks simple and has symmetrical proportions .
If there be two or more centres of calcification , as in the trabeculre in the middle of the septal ridges of Heliastr\#183 ; ( J3a , Mussa , Fungia , etc. , several small fascicles are formed and grow so closely crowded together that a large knotty-looking or irregularly-shaped granulation is the result .
Once more I repeat the important observation that the same microscopic structure holds good for all protuberances on the skeletal surfaces of Madreporaria , and I , therefore recommend the use of a single term , preferably " granulation , " for all cases .
The typical appearance of a " granulation , " as seen in surface section , is shown in fig. 33 .
This figure yields also unmistakeable evidence of the fundamental lamellar structure of the Madreporarian skeleton .
Full-grown coral skeletons often show much larger granulations at the uppermost part of the calcareous framework than lower down .
This is simply due to the fact that growth in thickness of the septa , costre , etc. , is continued after the coral has reached the full height which happens to be characteristic for that particular species .
During this continued growth in thickness the " granulations " are the parts of the skeleton most added to , and they assume a variety of shapes .
They very often lose at the same time their primitively clear arrangement in rows along the course of " trabecul(13 , " since they keep spreading sideways , crossways , or lengthways , and become confused with neighbouring granulations .
It is in this condition that the granulations are apt to become " pseudosynapticular , " i.e. , to form calcareous bonds between adjoining septal , costal , or indeed cCBnenchymal parts of the skeleton .
Again , " pseudosynapticulre " ( vide fig. 71 ) was a term applied to the appearance of uniting bars in a few particular cases studied by PRATZ .
The very general occurrence of such bars is demonstrated by my observations ( antea H eliastrcea ) , and is discussed below ( p. 246 ) .
We simply have a general law of the correlation of parts to deal with .
Where the coral is one which grows to a considerable height under its normal conditions of life , the ectodermal layer expends its building energy on adding to the height anrl not to the thickening of the skeletal parts .
And the large number of our common Astrreids are tallish corals .
But where the coral is one of lowgrowing habit , as the most of the Fungids are , it is predestined as it were to the development of pseudosynapticular granulations .
The occurrence of " true synapticulre " in the Fungid group may be considered , I believe , 011ly a specialized mode of interlocking skeletal parts which have a " pseudosynapticular \#183 ; ' habit .
Hence I have not allowed this to influence my system of classification in any great degree , but hold that the Astrreidre and Fungidre are very nearly related to one another , more nearly than either family with any other members of the Madrepomria .
From the Astrocidre , however , I have removed the Stylinid and Eusmilinid groups on several grounds .
From what has been said it will be seen that the systematic use of granulations according to their superficial appearance is for the most part misleading .
It must be combined with an exact microscopic study of the trabecular relation of the granulations .
My meaning will be at once understood by referring to some of the foregoing figures ( cf figs. 68 , 69 ) .
BASAL STRUCTURES .
Solid Lamina ?
' Deposit.-Smooth horizontal laminre or growth-lamellre may be successively apposed on one another without interruption .
In this way a solid Fig. 70 .
cliss .
Dissepiments belonging to different " floors " between two septa , and met obliquely in a t-ransverse section .
Fig. 71 .
S !
J11 .
ps .
syn .
co .
True synapticulre , " syn .
, " pseudosynapticulre , " ps .
syn .
, " and synapticular coalescence of septa , " co .
, " seen in transverse section ( p. 240 and cf. p. 176 ) .
laminar deposit occupies the base of the calyx , added to during each growth-period , hut leaving no free space between the group of laminre deposited in two successive growth-periods .
Solid basal deposit is typical chiefly of low small corals , ex. Turbinolidre , as well as branching colonies in which the individual calyces are very short- , ex. Oculinidre .
Tabul03.-A tabula represents a horizontal group of growth-lamellre deposited in the calyx at the close of an active growth-period .
The base thus formed spans the calyx completely in its central part .
The tabula may be smoothed or arched ( cf. fig. 12 ) .
Tabulre occur in most Palreozoic corals , and in certain of the Pocillopor id re , Stylinidre , Oculinidre , and Turbinolidre .
In a word , the great majority of ancient types of Madreporaria which we know possessed tabulre , but , as time went on , tahulre b8came almost a sign manual of " old fashion , " and has now become limited to these recent families which I , in my systematic table , have derived from the ancient Zaphrentoids .
Dissepiments are small tabulre present in the interseptal loculi , but not in the . .
SYS'rE~1ATIC STUDY OF THE MADREPORARIAN TYPES OF CORA .
LS .
245 lumen of the calyx .
Whereas a tabulate floor is understood to have only one group of growth-lamelhe , a dissepimental floor is made up of many groups of growthlamellre , usually highly arched and resting at their rims upon one another .
The floor of the interseptal loculus assumes therefore an uneven appearance , and if the dissepiments are very small and delicate , it may be described as vesicular .
In most cases the floor is wholly compact , leaving no part of the interseptal loculus unoccupied by calcareous tissue .
Transverse sections ( as in fig. 70 ) however do not pass through the plane of the dissepimental floor , but meet usually the upper parts of successive floors .
The development of regular dissepimental floors at the intervals of growth-periods is particularly characteristic of the Astrreidre ( antea , p. 157 ) .
Columella .
Colurnellar Style.--The so-called " true colmuella " is sty liform or fasciculate in character , and is laid down in an arched or pouched central part of the aboml bodywall of the polyp ( cf : figs. 35 and 56 ) .
I have , in the second part of this paper , traced back the origin of the true columella to direct modification of the tabulre in the central part of Palreozoic corals .
In this case the colume11a is a pnrely basal structure .
But the longer septa often invaded the central tabulate area of Palreozoic corals , and shared therefore in subsequent modifications which that area underwent .
Curious examples of a partly septate , partly tabulate columella , are common among carboniferous genera , more especially the Clisiophyllinre .
In them , small septal spines are caught up as it were by the bent , vesicular tabulre .
And possibly in this we have a primitive form of the irregular " trabeculate " or " spongiose " pseudocolumella in many younger genera .
Pseudocolumella.-A pseu\lt ; locolume11a may be defineJ as an irregular skeletal tissue composed of the inner or central ends of the septal trabeculre , bound together by intermediate calcareous deposit of various character , sometimes vesiculate , sometimes bar-like .
A pseudocolumella is of extremely frequent occurrence in the families of Astrreidre , Fungidre , and Eupsammidre ( cf figs. lB , 23 , 45 ) .
The Larnellar Colurnella.--The most primitive form of a Jamellar columella is the elongation of the main septum into the middle area of the ca.lyx in many Zaphrentid corals .
Occasionally , union of main and counter septum occurs with the formation of a complete mid-line acrmis the calyx .
In many Jurassic corals one still finds a " main septum " continued as a columella ( cf Part II .
, p. 292 ) ; in others again , a lamellar columella is present in the middle of the calyx , which cannot be observed to be actually in continuity with any of the primary septa .
Yet there can be little doubt that the free , axial lamellar columella present in certain Trochosmiliacere , etc. , is due to retrogression of a main septum .
This origin is farther indicated by its microscopic structure , which is identical with that of a septum .
The lamellar columella has usually been spoken of as a " true columella , " like the styliform columella , but its origin and structure is in reality " septa ] , " like that of the pseudocolurnella .
Synapticulce.-The name of " synapticulre " was introduced by EDWARDS and HAIME .
These authors wrote : " A certain number of the septal granulations sometimes meet those on the adjacent septal surface ; thus transverse prolongations of the septa result in the form of rnterseptal bars which we have called synapticulre " ( ' Coral .
, ' I. , p. 35 ) .
" The presence of these solid bars characterise the family of the Fungidre , and have not as yet been observed in any other group of the class of corals .
It is in the genus Fungia that , the synapticulre are most highly developed , " etc. ( ' Coral .
, ' I. , p. 66 ) .
PRATZ distinguished later two kinds of synapticul::e , according to their fine microscopic structure .
He found that the solid interseptal bars were not always " prolongations of the septa , " but that they might be in part or wholly independent structures .
Synapticular bars formed by the fusion of granulations on the adjacent surfaces of two septa , were called by PRATZ " pseudosynapticulre .
" While synapticular bars , whose microscopic structure showed 11ew centres of calcification for the calcareous fibres , were regarded by him as " true synapticulre .
" Macroscopically , it is scarcely possible even for a practised eye to see any distinction , yet microscopic study proves that PRATZ was right .
In most of the Fungidre , both kinds of synapticul::e are present .
Fig. 71 , shown above , side by side with fig. 70 on p. 224 , displays the occurrence of " true synapticulce " at the peripheral part of a calyx , where the interseptal loculi are , vide , and of " pseudosynapticulce " about the middle of the calyx where the interseptal loculus is narrower , while still nearer the centre the end of a shorter septum is shown to be coalescent with the lateral surface of a longer septum .
This figure may be confirmed by reference to figures of Siclerastrcea given above .
It is made somewhat diagrammatic here , in order to bring out ( 1 ) the " centre of calcification " in a true synapticula , , ( 2 ) the absence of any such centre in a pseudosynapticula , and { 3 ) the gradual passage to the actual coalescence of septa centrally .
The genus JJ'ungia itself also shows similar variety in septal union according to the width of the loculus .
In Lophoseris again , the septa are very closely placed throughout their whole length , and the interseptal loculi are easily bridged by psendosynapticul::e .
" True synapticul : : e " are formed only when the interseptal loculus , is fairly wide , and the granulations are not prominent enough to meet each other .
Then part of the body-wall is compelled , as it were , to form a new interseptal invagination between the granulations , and to lay down fibrous lamin::e around a new interseptal centre of calcification .
At the same time the body-wall draws itself up locally , the unsupported intersepta.l parts on either side of any synapticula hanging freely downward between it and its neighbours .
Thus the body-walJ in the interradial loculus has an interrupted synapticular series of basal supports instead of a continuous , if uneven , dissepimental SYSTEHATIC STlJDY O.l !
" i THg 1\IIADREPORARIAN TYPES 01 !
' CORALS .
247 floor .
I have already pointed out that the openness of the synapticular floor is specially adapted to low-growing forms of Madreporaria ( antea , p. 237 ) .
Previous illustrations ( e.g. , fig. 37 ) have demonstrated their essential structural relation to the septa !
trabeculre and to the growth-segment of a septum .
Their distribution in a calyx in consequence entirely corresponds with the course of the main trabeculre in the septa .
The distinction made hy PRATZ , of true and false synapticulre , is of use in classifying Fungid genera .
An important Mesozoic and Tertiary group of Fungids , the Thamnastrreinre ( = Psendoastrreinre Regulares , PRA'rz , p. 107 ) possess only pseudosynapticulre , whereas all the recent Fungids ( Funginre ) possess both true synapticuhe and pseudosynapticulre .
The Lophoserinre have mostly pseudosynapticula:i , the Eupsamrnidre mostly true synapticulre .
In all three families , the synapticulre may be defined as interradial calicinal structures , which afford a partial support for the aboral body-wall of the polyp in interseptal and intercostal loculi .
They replace dissepiments to a large extent iu these families , a few synapticulate genera being entirely without dissepiments .
Typically , the synapticulre are merely transverse rods between septa , one tier of \#183 ; " rods " being formed at each growth-period .
Carince.- " Carina , " the name which has been given to vertical bars marking the s11l'faces of septal ridges in Heliophyllum , and one or two other Pala:iozoic genera .
The carina , in this case , represents the coalescence of the granulations on any one ridge of a septal surface .
In many Triassic and Jurassic genera , horizontal bars occur running across the septal ridges .
These bars represent the coalescence of the granulations at the same horizon of growth in the neighbouring ridges of a septum .
The name of " carinre " may be very well applied also to these horizontal bars .
The carinre on the adjacent surfaces of two septa meet occasionally in the interseptal loculus .
The same specimen may show adjacent septal carinre in parts coalescent , in parts not touching ( cf PRATZ , loc. cit. , " Horizontalleisten " ) .
The genus Fungia may be regarded as presenting an example of horizontal car\#183 ; ince running transversely across the septal ridges , which are always coalescent with one another across the interseptal loculus .
Coalescence of carinre takes place just in the same way as the coalescence of a pair of granulations .
In short , one may have synapticular ca.rime just as one may have synapticular granulations .
Again , the carinre may sometimes be united by the intercalation of new centres of calcification in the interseptal loculus .
The figures given of Fungia above ( chap .
iv .
) , show that the carinre of Fungici are united in both ways .
Carinre may be defined as bars running either lengthways , i.e. , parallel \Yith , or transversely across the ridges of the septal surfaces .
They are formed by vertical or ]ateral coalescence of the septal granulations .
They may meet across an interseptal loculus , either as pseudosynapticulre or true synapticulre , or both combined .
When 248 DR. MAR !
A M. OG !
LVIE ON THE MtcROSCOPIC AND such interseptal coalescence of adjacent carinre takes place , the latter may be called synapticulate carince .
Although synapticulre are characteristic of the calyces of the Fungidre and Eupsamtnidre , I have observed the occurrence of synapticulre in Astrreldre .
Fig. 27 above shows interseptal coalescence of granulations , i.e. , " pseuclosynapticul(JJ , " m 1-Ieliastrcea .
Tangential Structures-Epitheca , Theca , Pseudotheca .
Epitheca.-Whereas an Epitheca is a primitive structure in Madreporaria , thecal structures are of secondary origin .
The epitheca is an external basal structure , laid clown at the angle of the aboral wall , where it bends towards the oral or peristomal Fig. 72 .
t:Jept .
.l !
Jpith .
Transverse section showing the relation of Epitheca to Septa , where neither Theca nor Pseudotheca is present .
The interseptal loculi are filled by basal supports , omitted for clearness in the figure ( tabulre , dissepiments , or synapticulre ) , which coalesce directly with the epitheca at their external edges .
Fig. 73 .
Hp .
0 .
'l.'h .
8 .
'l'ransverse sectiou showiug tho structural relation of " S. " = Septa , " O. " = Costa , " Th. " = Theca , and " Ep. " = EpHheca in a typical 'l'urbinolid .
region of the polyp ( figs. 22 , 36 ) .
It is the continuation upwards or outwards of the embryonic " basal plate , " and may be well-developed or remain rudimentary .
Fig. 72 represents part of a calyx in which the septa are directly surrounded by an epitheca , no theca being present .
This case may be observed in Montlivaltia , Microsolena , etc. , also in several Palreozoic genera .
Theca.-Tangential folds of the aboral wall of the polyp may rise at the peripheral part of the polyp , simultaneously with the septal folds .
A theca is then developed within these tangential folds .
A " dark line " or " centre of calcification " in the theca corresponds to the edge of the tangential fold just as that in the septum corresponds to the edge of the septal fold .
When septal and tangential folds keep pace with one another in growth , their centres of calcification are continuous ( fig. 73 ) .
In many genera the septa grow more rapidly than the theca , and then the thecal centres of calcification appear between septa ( antea , figs. 4 and 5 ) .
In solitary thecate corals ( e.g. , Turbinolidre ) , the epitheca encircles the theca , and is coalescent with it .
In colonial thecate corals the epitheca may encircle the theca of SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
249 each polyp ( Amphiastrreidre ) ; more frequently it is limited to the base of the whole colony .
In the latter case , the parts of the aboral wall around the thecre of adjacent polyps form ccenenchyme ( ex. , Oculinidre , Madreporidre , Pocilloporidre , Poritidre ) .
When the ccenenchyme is of compact structure , the theca is also compact , e.g. , Oculinidre , Pocilloporidre ) ; when the ccenenchyme is porous , the canals in its meshes are rendered continuous with inte.rseptal loculi in the calyx by means o " pores " in the theca ( e.g. , Madreporidre , Poritidre ) .
Pores are not altogether unknown in the theca of Turbinolids .
They are present in the genus Trematotrochus , WOODS , in the same position as the intercostal dimples of the Turbino1idre , i.e. , in the same position as the pores in the thecre of Turbinaria , , Porites , and Madrepora ( figs. 55 , 64 ) .
Two rows of pores occur between adjacent septa on either side of a small thecal costa , which is placed midway between the same septa .
I take this thecal costa to be the same as an " epitheca1 costa " in many of the Palreozoic genera .
Pores also occur in the thecre of Turbinolid genera which have rootlets ( ex. , Rhizotrochus ) .
Similarly in certain Palreozoic genera which characteristically retain the rootlets .
The thecal pores are , in those simple types , always a means of communication between the interseptal loculi and the canals of the rootlets , and are , in my opinion , the exact homologue of the thecal pores in ccenenchymatous colonial types which have a porous ccenenchyme .
Note on the Porous Theca.-ORTMANN has , in his systematic arrangement of the Madreporaria , treated the solid parts of the porous theca in Madreporidre and Poritidre ai : : the homologue of the synapticulre in Fungidre and Eupsammidre .
( I use these titles of families in the sense in which I have already defined them in foregoing pages .
) .
He clasi:iifies all these families as " Athecalia , " ORTMANN , and places them in coutradistinction to the " Euthecalia " ( HEIDER , ORTMANN ) , represented chiefly by the Turbinolidre , Pocilloporidre , Galaxea , Euphyllia , etc. , and the " Pseudothecalia " ( HEIDER , ORTMANN ) , represented by the Astrreidre ( vide below , pp.310--314 ) .
I entirely disagree with the main sub-groups erected by ORTMANN .
Reference to my figure of Madrepora ( fig. 60 , p. 214 ) will be sufficient to prove that the theca of Madrepora corresponds in structure precisely with ORTMANN 's own definition of an Eutlteca , differing only in the fact that it is porous .
To call the solid parts of the Madrep01\#183 ; a theca " synapticulre , " is to deviate wholly from the original application of the term by EDWARDS and HAIME ( ant ea , , p. 167 ) , in so far as no part of them is made up of septal gramulations .
The synapticula would , in this case , have to be defined as a thecal bar , having its own centres of calcification set at right angles to , or very obliquely to , the centres of calcification in the septum .
These thecal bars are , moreover , always built 1tp exactly opposite each otha , and at one de.finite radial distance from the centre of the calyx .
The homologue of such uniting bars is found in the theca ( = .Eutheca of HEIDER and ORTMANN ) of Turbinolidre , and not in the occasional irregular rings of synapticulre and pseudosynapticulre in the calyces of Siderastrcea , Loplwse1'1:s , etc. ( of figs. 42B , 46 ) .
Hence I would apply to the wall in Madrepom , Turbinaria , and Porites exclusively the name of a porous theca , thereby indicating ( I ) the fact that a t angential thecal ring of calcification is present , even if it is not , continuously developed during upward growth of the polyp ; ( 2 ) its structural homology with the usually compact theca of Tnrbinolidre , Pocilloporidre , and Oculinidre .
The same relations obtain in all these cases between the theca and the septa , whether the theca be porous or compact .
The interrupted development of a skeletal structure is nothing unusual .
An analogy may be , for instance , made with the septa of 'frfadrepora .
Some are con .
tinuously developed , and appear as a compact septa } plate ; others , in he same calyx , are interruptedly developed , and appear as a vertical uncompacted series of septal spines .
Yet the structure of the septal spine is precisely the same as that of successive growth-parts in the flat septn.l plate .
Thus , we must seek for the .
structural homologue of the " porous wall " of Madrepora , without paying any respect to its interrupted development , and need not feel surprised to find it , in u compact form of theca .
Extrathecal Structures , 01\#183 ; Structures external to the " true " 11/ ieca .
\#183 ; The extrathecal structures which I shall enumerate are forms of ccenenchymal and epithecal deposit .
The epitheca may be present as occasional rings round the theca , or round the exposed edges of septa .
In many Turbinolidre the deposit of calcareous matter on the outside of the thecal centres of calcification becomes very great , and may be either compact or may show hollow spaces .
The thin epithecal rings in some genera are probably the homologue of this thick concentric extrathecal deposit in others .
The latter simulates " ccenenchyme , " but is an entirely .irregular structure , and its calcareous composition seems to me to be utterly distinct from that of all interpolypal ccenenchyme in coral colonial life .
The following are the chief forms of extrathecal structures :a .
Occasional epithecal rings , not forming a connected pellicle .
b. Pelliculctr epitheca , solid or vesicular in structure .
c. Ornanicntations-bosses , echinulat.ions , crests , ridges .
Such ornamentations may be solid , and are then of secondary origin .
More frequently they are hollow , and represent original concavities and irregularities of the bodywall , or the scars of buds and embryonic rootlets .
The costre on the " rugose epitheca " of many Palreozoic genera alternate in position with the septa ; they find their counterpart in the extra costre introduced in the wall of Turbinolidre and of many ccenenchymatous types .
d. R ootlets-de finit e , hollow , calcareous tubes at the lower part of a theca , by which the young polyp attaches itself , and which may or may not be retain ed clurin O ' its fre e existence later . .
" '.\gt ; SYS'rEMA'l'lO STUDY o~~ TH:1\#163 ; MADRE : PORARIAN TYP ] ; S O ] ' CORALS .
2b1 e. Cwnenchyme-( l ) compact : a system of calcareous superficial channels between adjacent thecre in a colony ; ( 2 ) po1\#183 ; ous : a complicated system of intrnskeletal and superficial channels between adjacent theme in a colony .
The skeletal channels open into the calyces by means of pores in the thecre .
Pseudotheca.-'l'his term is a very useful one , which has been applied by HEIDER ep .
ps.th .
c. s. Fig. 74A .
c. Fig. 74n .
s. ps.th .
ep .
Transverse sections showing the structural relations between " s. " = septum , " c. " = c0sta , " ps.th " = pseudotheca , and " ep .
" = epitheca , .
The pseudotbeca is shown rather uearol ' the centre of the calyx in fig. 74u than in fig. 74A ; all gra\lt ; les occm\#183 ; between a position of the pseudotheca very near the centre and very remote from it .
, until it may be a , most coalescent with the epitheca ; the costa is respectively longer 01\#183 ; .shorter .
and by ORTMANN to the ring of septal thickening which may arise at any part of the calyx in the Astrreidre .
Those authors have defined it as a compact " false wall , " formed entirely by the lateral coalescence of septal surfaces at a definite zone in the calyx .
During the course of my microscopic investigations I have found that oue must slightly modify this definition , since in many cases the laying down of the basal structures matArially helps the compacting of the false wall ( cf Heliastr(l}ct and 111.ussa ) .
Even then it may not infrequently be that tiny apertures are left unfilled .
According to the position in the calyx at which this thickening t:tkes place there will be a narrower or broader costal area around the inner part of the calyx ( see figs. 7 4A and 7 4B ) .
As the epitheca is external to the costal area , the appearance of " two walls " may be produced .
In colonial Astrreids , however , budding goes on in the inner or outer zone of the calyx , and the epitheca surrounds all the polyps of a colony in common .
The pseudotheca distinguishes itself structurally from the theca by being form \lt ; l within radial , i.e. , septal , folds of the calyx .
It will readily be seen that the synapticular rings in Siderastr(l}a , Lophoseris , etc. , may be regarded as the homologue of the pseudotheca , although not of the true theca .
If the pseudotheca forms very near the periphery , the epitheca may be coalescent with it ( ex. Goniastrceci at the edge calyces of a colony , and frequently .Zlf.ussci ) .
Usually , however , there is a fairly-developed costal or edge zone between pseudotheca and epitheca .
25:Z DR. .
MAR[A .
M. OGILVlEl ON 'rHE MICROSCOPIC AND NoTE .
Before closing Part I. , I wish to refer again lo the series of figures , 2 , 22 , 35 , 36 , 56 , which are longitudinal sections of various coral types , treated diagrammatically .
They show the investiture of the hard skeletal parts ( printed black ) by the soft polypal flesh ( printed red ) .
In one portion of each figure , the polypal flesh is represented covering the septum and costa ; in another portion , it is seen clothing the basal structures in an " interradial loculus " between adjacent Repto-costre ( cf. pp. 107,108 ) .
The edge-zone ( " Randplat.te " ) of polypal flesh is specially indicated , and its relation to the epithecal deposit .
In JJ1.ussa and Fungia , representative types of my proposed subdivision , Septocorallia ( p. 322 ) , the prominence and shape of the septal wing of the " radial structure " determines the particular curvature of the edge-zone along the costal wing .
Stylina , Galaxea , and 'Tiirbinaria all belong to my group of Camenchymata ( p. 320 ) , and may be said to represent three very distinct grades of cranenchymatous corals .
( 1 .
) The fossil Stylinids , to judge from their skeletons , supported an edge-zone upon the broad costal wings of the radial structures and the intercostal vesicular tissue .
But the costre of adjacent calyces frequently did not quite meet , and cranenchymatous deposit filled the gap .
( 2 .
) The living Galaxea has an edge-zone supported on the narrow costal wings of radial structures , but a very considerable gap between adjacent calyces is filled up by cranenchymatous deposit supporting a " common cranosarc .
" ( 3 .
) In Turbinaria , a " common cronosarc " entirely covers the space between adjacent calyces .
The rings of the cronenchyme nearest the calyces show the same stmctural relations with the " radial structures " as are to be found between the primitive wall ( eutheca ) and the radial structures in many non-cronenchymatous corals .
Henee I look upon the " cranencbyme " of perforate corals simply as a particular form of deposit laid down by the wall area of the polypal flesh .
Correlatively 1 regard " camosarc " as of miiral origin in contradistinction to the " edge-zone " which I have traced to an intraca.lycinal origin ( p. 262 ) .
Both are adaptations of special parts of the polypal flesh ; and neither , so far as I see , has any right to be considered more primitive than the other .
( exc .
foot-note p. 226 ) .
List of Abbreviations itsed in lettering the Figures , This " list " will be found at the end of Part II .
on p. 345 .
To face page 252 .
TABLE of Classificatory Features .
[ Phil. 11mns .
, Og ilvie .
------------------------------- ; ------------- ; --------------- , ------------------------------------ Centres of Oalcifi-cation , with especial reference to the nrrangement of the mic1\#183 ; oscopic calcareous fibres and the appearance of a so- call ed " dark line .
" Sever al centt-cs of calcification are gatherer !
m groqps corresponding to the i11div id nal t.rabecular ridges .
Each group is clearly distinguishab le m trn n sverse section from ueighbouring gronps , :md as the trabeculoo meet the septa .
I edge at a hig\#183 ; h angle , no appearance of a con- '\gt ; tinuous " dark line " in the median plane of the septum is presented .
The calcareous fibres pass upwards and outwards from the axes of the t rabeculoo to the surface of the septum .
:Fascicles of longer fibres ernm\#183 ; ge as scptal granulations .
_ Radial structures ( septa , costoo , pali ) .
Compactness or porosity .
Successive growth-parts in indiYidual trabeculoo .
l 'l'hc indivicl ual trabccul::o I consist 0\#163 ; a continuou~ , Con tern poranoou slyformt ) d growth-parts oi ' adjacent trabeculre ( the septal segment ) .
------ A compact ti~r of trabocu-l lar growth-parts is I formed at t be eclgo of \gt ; the radial skn cLu res 1 daring each succc~si vc I grov ; th-periocl .
j coinpacf , vet\#183 ; t ical series ; , -------of trabecular growth- I " " '''\#183 ; j l _________________ 1 The individual trabecnloo l are not primitively I compact ; the series of growth-parts a r e not ~ joined up encl to cud , I succ2sve partR may \#183 ; , sometimos alternate in 'l'hc t icl ' of trabcc u l , n\#183 ; growth-part : ; at the edges of septa and costre are of fen not compacted within one period of growth .
Apertures may remain I unfilled between the neighbouring growthparts or they may be s ubseqncnLly filled up .
'l'hc first-form ed fibrous deposit isl mo r e horizontal in position than I the lat , te r. The groups of " centres " are again present , but alwn , ys in \gt ; close contiguity with one anotl1er , 1 and in a straight median row , so I as to give the effect of a continwms position .
_j median line of calcification .
j , __ j I Characteristic wheel- like arrange- } ment of fascicles of fibres directed horizontally outwards from one common axis or centre in each h-ahecula .
No appeara , nce of a " dark line " in the median septal plane .
Slrongly-111a1\#183 ; kcd " dark line " in tho ' " ) medirm plane of the septum , owing I to the almost horizontal posilionJ1 of the trabccnla or spine .
'l'be fibres pass obliquely upwards and outwards .
Successive growth-part-s I arc always directly I above each other i11 I posit- , ion , but pores are r somcti111es left in the I trabecnla , opposile f\lt ; ) ' llapticuJai .
j Compact septu .
l The indi viclnal trabcculoo I nrc compacted ( 'l'nrb iuaria and Maclrcpora , ; , ( pp. 21:{ , 214 , ) or re- I m"in free from one anothc1 \#183 ; as septa ] spincs.j Tlic fibres vary , horizontal or obliquc.1 The indi vidnal tra becnlm " ) Strongly-marked " dark line " here , consist , o f a cont.inuons , I owi~g usunlly lo tho very close compru\#183 ; t , ye1\#183 ; tical series t set Ling ( i11 serial line ) of the ccnt1\#183 ; es I of trnbccL1lar growL h- I of calcilicatiou .
j parts .
j Usually porous .
l The individual trabeculm I arc very distinct , and their parts at the edge \gt ; complete .ely en veloped 1 in polypal flesh dm\#183 ; ing I any one period of deposit .
j Compact .
l Since the trnbccuh m\#183 ; I spine is horizontal with the growing septa ] ; , edge , ouc IJl'Olctl , -Jmrt I represents the whofo seplal seyment .
j Compact .
l The trabccnlm arc cx-1 trcmclyna1Tow and adjoining growth - parts ; , in any one sc plal , cg-1 meant arc iutimately comptteted .
j Direction of the r-rabcc ulre or trabecular ridges in the radial structure .
-\#183 ; - The trabcculro or ridgcH d i verge obliquely upwards , in regular fanlike form , to right and r ie ft , from one or more areas of divergence in the rarfo , l structure .
FAMILIES .
r AsTR./ ElD\#163 ; .
Relation of the md , iol strucfores ( septa anu .
costre ) to the wall .
Intcrradial and c\#163 ; Xio-polypal basal snppurls :-\#183 ; ( 'l'abulre , disscpiments , synapticuloo , pseudo-synapticul re , lateral septa !
coalescence , terminal septa ] coalosceucc , coluwella .
{ I Diasepimeuls .
, : B.arc occm\#183 ; :\#183 ; cncc of p se ndo-synapticul ro I 1'erminal septal coalescence produces very often a pseudo-columella .
TlIAMNASTHiEINA ' : .
the pseudotheca and the interradial loculi correspond in the adjacent polyps of colonyI 's endo-synap1icnlm in vertical series .
Dissepimcuts alw p1\#183 ; csc11t .
Occasional occurl'ence of a pseudo-col um ella .
I j The radial slrncli , rrs pass throi , g h { building form !
'\lt ; .
---- ~ I All costre are e represented by ' Syna .
, ,ticulce ( both true , and pseudoF uNG1N\#163 ; .
1 r ~ ~ septa within the calyx .
I s , rnapticulm ) in Yertical series .
S I Disscpiments also present except in the I j ~ l ' ~ genus Pungia , where the synapticulm~ 1 form continuous bars .
I LOPIIOSERINJ\lt ; : . !
I Lateral septal coalescence charac !
I teristic ; terminal septal coalescence ~ l also present .
---------~---------- ------------------------------l The t.rabcculre diverge in I irregular fan-shapc , l \gt ; figures in the radial 1 structure .
I j ,1 'l'hc trabccnlm arc cl i- I rected almost vrrti- ; crtlly n pwarcb , more ; , 01\#183 ; Jes~ paralld with on e auother .
i I j The irnbecnlm :1rc di-1 reclPcl almost hori- I wnLa ll y inwards from the i h('C , t towal'ds the ; , centre of the calyx , I and outwards towards the cumcnchymc .
j The tralicculn1 vnryl greatly in cl irccLion .
At the Ycry base o[ n ' seplum an area of I divergmtec is usually ~ present , afLerwarns tlil ' I trauecul(( ' !
]/ 'OW inward : , I from the 1rnll as in the Madreporiclm .
j EursA1n11n..i : : .
P ORIT lD .
E. r.i I M ADREPORIN./ E. ~ I r:H ]I l'2i l TutrnlNAIUNA\gt ; .
'l\ , RLJJNOLI 11./ E. r I I ~ I L l , I The radial structures pass throngh I the pswdotheca , but the inter- Idern , but the syna.pticulro arc nor placed radial lo~uli d ?
not alwnys cor- j in regul_ are .
Yertical series , they are l r espond m a\lt ; lJacent polyps of ' ) mostly lnmted to t h e wall are ea and colonies .
A ll costoo arc rept\#183 ; c- 'I form an irregular network .
sented by septa within the c , dyx .
I '- ( " 8ynaptimlce ( only true synapticulre ) I in vertical series .
\lt ; Dissepiments also present .
I Colurnel la frequen t. I Lateral roa !
e. , cenre of s eplal pairs to L form pali .
Th e radial structures ?
1W !
J or may , not pa,1:s through the hrca ; the I intcnadial loculi never cone- I spoon \lt ; l in adjacent polyp~ of .J colonies .
'\ Alternate cosr-m occm\#183 ; i11 tlic l wall which arc n ot r epresented by sepia wilhin tl1c calyx .
E.tfrcmely rnrio1 , le natnre of Ofl\lt ; al deposit , reprrso11ting an irregular combination of tabulate and synapticuhte structures ; it is almost absent in the cal !
Jl'CS , 1\#183 ; idily develu1)(JJ in the cwnenchym\lt ; ' .
A main septum is often continued as columclla .
foterradi'al deposit almost absent ; reprcsentccl at the base of tho cn.lyces by ( t solid tnbulate deposit .
Lntcr , il RcpLal coalescence may occu1\#183 ; .
Oolnmella usually prcscnL cithet\#183 ; as a n indepcnd , \#183 ; nt , structure 01\#183 ; connccLcd , rilh a main septum or wiLh pa , li .
( 2 ) Wall .
( 1 ) Compact or porous .
'l'hc " centres of caleification , " their relation to septal " centres .
" { I The theca is pact .
cmnr Outer ( true ) thocaj very rarely prnThe theca 1s sel dom co mpact Lhrougho ut , it is usualiy p:u\#183 ; tl.v porous , partly com pa.ct .
sent ; a pseudo theca is formed by a thickening of the laf , eral surfaces of the septa at the pet\#183 ; ipheral area , together with1 a m od ification of the inter- I septa !
basal sup- L I ports .
-- r I j The tLeca 1s nTe- , I gularly pol'ous .
l1 I ]1 'l'lio Lhcca is regn .
'\ lftt " lie ))Ol'OllS j in !
form 1 7.igzn.gged .
0 utcr ( lrue ) theca l is the only one present .
It has its own centres , of calc:ircons de- I posit , indepcndent I The ll1c ca is reguof those in t.1 10 .J I lai : lie porous ; in radial strucLurcs .
' form pe rfectly 'l'bc thecal deposit I circular .
usually seen to L Lend tanr , ent ially rou11d and form continuity between adjncenc 1 septa .
I I LI I The t hcC'a is compact , either zigza ) Iged or circu1:ir in shape .
---------- SYST .
KM.A'rtc STUDY OP THE MADRl !
JPORARI .
AN TYPES OF CORALS .
253 PART !
!.-PHYLOGENY OF MADREPORARIA BASED UPON SKELETAL STRUCTURE .
CHAPTER IX.-EVOLUTIONARY CHANGES .
Page .
1 .
KocH 's Description of the " Septal Spine " of a Palreozoic Genus . . . . . . . . .
253 2 .
Similarity in Microscopic Structure of the Septa in Palreozoic and Recent Madreporaria .
259 3 .
The Coonenchyme of " Perforates " . . . . . . . . . . . . . . . . . . .
261 4 .
The Tabulate Area of Pal !
eozoic Madreporaria gave place to the Columellar .Area of N eozoic and Recent Types . . . . . . . . . . . . . . . . . . . . .
268 5 .
Precocity of the Radial Symmetry in Cyathophylloids . . . . . . . . . . . .
273 6 .
The Change from Bilateral to Radial Symmetry was an adaptation to the increase in number of the " Reproductive " Mesenteries . . . . . .
276 7 .
The old-fashioned Character of the Recent Genus , Seriatopom . . . . . . . . .
279 8 .
The Fossula and the Pinnate Arrangement of the Septa . . . . . . . . . . . .
280 9 .
Correlation in the Position of Directive Mesenteries with that of the fossulre ( axial , abaxial , alar ) in Palreozoic Madreporaria 286 10 .
Brief Review of Madreporarian Evolution . . . . . .
292 11 .
Recapitulation of Phylogenetic Changes . . . . . . .
295 12 .
Influence of Geographical Changes on the Palreozoic polyps 296 1 .
Kocn 's DESCRIPTION OF THE i : SEPTAL SPINE " OF A P ALJEOZOIC GENUS .
The term " stereoplasm , " as first used by LINDSTROM , was applied to a , " structureless " calcareous deposit on both sides of a narrow stripe , much darker in appearance , in the middle of the septum .
This narrow stripe or inner septum of LINDSTROM is not however the same as was meant by VON Koen when he introduced the term " Primarstreif " ; farther , the identification by FRECH of LINDSTRoM 's and Kocn 's nomenclature has been the cause of considerable confusion .
Kocn 's description of a longitudinal section of the septal spine of Pholidophyllum , LoYENI , is possibly the most exact which is to befound in the literature ( cit. ant .
, :j : p. S 9 ) .
It reads as follows : " The central mass appears in quite thin sections to be composed of small crystals , showing no particular arrangement ; in this mass one can obserYe small dark spots and points crowded together in greater or less number , and glancing brightly in some positions of the microscope . . . . . .
The second , clearer zone , consists also of small crystals , between which spots and points occur , and , indeed , more numerously and generally larger than those in the middle area ; but the crystals appear radially arranged ; this is most plainly seen near the periphery , where they become rather larger . . . . . . .
That mass which unites the individual spines together and forms the " wall " is composed of crystals , the same as the crystals in the spines themselves , but they are arranged in rows which run nearly parallel , and in somewhat thicker sections appear separated from one another by strong , dark lines .
These lines seem to run concentrically when seen in transverse sections of the spines , but in longitudinal sections they are parallel with the vertical axis of each individual spine and pass into the similar lines of neighbouring spines with a slightly crooked course .
" My reason for quoting the entire description is twofold : ( I ) because of its agreement with a longitudinal section of a septal spine of the recent coral Madrepora spicifera , DANA ( antea , fig. 61 ) , and to which I shall presently return ; ( 2 ) because it is important to understand clearly the nomenclature used in the literature of the day .
KocH makes no mention of the word " stereoplasm " neither here , nor when , a few pages farther on , he describes the transverse section of Cyathophyllum , sp. The latter is , he says , composed of a darke1\#183 ; median area like the central mass in the Pholidophyllurn spine , and two clearer lateral areas on opposite i : ; ides of the median line in which the crystals are arranged radially .
The median area in Cyathophyllum takes the form of a narrow band which he calls " Primarstreif " ; to the lateral areas he gives no special name .
I shall now represent in tabular form the comparison made by KocH between Cyathophyllum , sp. and Caryophyllia cyathus on the one hand , and Pholiclophyllum on the other .
I add a third column , which will show how KocH 's description of Pholidophyllum may be compared with mine of Madrepora .
I.-KOCH .
OyathophyllU ?
n , sp. ( Pale\#183 ; ) .
Oaryophyllia cyathus ( Rec .
) .
Septum , long in radial direction , and compact .
Septum .
KOCH'S distinctions .
a. Dark " Primarstreif " as = Central area a continuous median band b. Lighter-looking , radiat .
= Clearer zone ing crystals forming lateral bands c. Absent " Wall mass " II.-KocH .
Pholidophylluni Loven-i ( Pale\#183 ; ) .
III.-AU'l ' .
Madrepora spic1fera ( Rec .
) .
Septum = a vertical interrupted series of short " septal spines .
" = Dark , central mass , circular in section = Lighter-looking , radial ] y arranged crystals , around the central mass = Crystals arranged m rows parallel with the axis of growth ; uniting the spines at their basal parts , corresponding with their points of origin at the periphery of the calyx Idem ( called " core " er " axis " of calcification in my paper , regardless of its appearance , light or dark , homogeneous or crystalline , filled or empty !
) .
Idem ( called above , " primary growth-segment , " or first-formed deposit of similarly-oriented calcareous fibres .
) Ideni ( described above as larger fibres , irregularly swathed round the primary growth - segments , in the course of the withdrawal ordyingoff of the soft parts at any particular horizon of polypal growth ) .
--------'---------'------------------- ---SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
255 KocH 's " Septal Spine " in Pholidophyllum ident .
Au , thor 's " Septal Spine " in Madrepora and , generally , Author 's Growth-segment in the Marlreporarian Septum .
When KocH suggests that the septum of Cyathophyllum and Caryophyllia is structurally similar to the septal spine of Pholidophyllum , he expressly states that the suggestion is based on the agreement in appearance of " a " and " b " in the respective columns .
I have given actual evidence in the cases of Madrepora and Turbinaria , that a series of horizontal septal spines , may , when in contact at the successive\#183 ; horizons of deposit , make up a compact " Turbinolid " -looking septum .
In the genus Madrepora itself , the " principal septum " is often just a closely-set series of " spines , " and thus approaches the condition of the septa in Turbinaria , as well as in Turbinolia , Ca.ryophyllia , etc. Whereas the less-developed septa have their spines originally placed at wide intervals , and only get compacted at their basal parts , just as KocH has described for Pholidophyllum .
Summarizing my results , there are certain Madreporarian genera in which the septal growth-segments are laid down , at first in the form of free , separate spines , and the spines are subsequently compacted either in part or wholly .
The greater number of such genera is to be found amongst Perforata , in the former signification of this group .
During the process of compacting , the fibres may be deposited in the same direction as those in the free spines , or they may slope more in one direction and another entirely according to the local axis of growth during this secondary deposit .
Some of the Fungid genera show a different orientation of fibres according to the period of deposit .
Again , the Eupsammime , E.H. , show an even greater difference in the orientation of the fibres nearer or farther from the median plane , so that an inner band of fibrous deposit may be distinguished from an outer .
Both hands are composed of \#183 ; radiating fibres , but those of the outer band very obliquely inclined , while those of the inner form a very small angle with the horizontal .
The Poritinre , on the other hand , have the fibres of the secondary deposit oriented in exact continuation of those of the primary ; moreover , the fascicles are placed horizontally , i.e. , at right angles always to the upward axis of growth .
The Madreporinre show the most marked difference in the orientation of the primary and subsequent fo , years of deposit .
The primary fibrous deposit is horizontal and regular , the subsequent deposit follows no definite system .
The whole foundation for the term of stereoplasm rests , I believe , on those differences in the orientation of the calcareous fibres , differences which are observable only within some of the Mn , dreporarian groups .
Among recent corals , these differences are .
only well-marked in the Turbinolidre , Madreporidre , and Eupsamminre , the Lophoserinre also showing indications of it now and then .
In short , difference of orientation is a feature in many cases correlated with the laying down of the septal spines in isolated pockets at the edge of the main septal invagination , in other cases correlated with rapidity of growth and the quick formation of a mere scaffolding of " septal spines " instead of continuous septal structures ; in still others it is a result of slow growth , which allows crystalline changes of disintegration to go on very considerably during intervals of growth .
" Primary Streak " ( KocH ) ="Primary Septum " ( FRECH ) , " Crystalline Deposit " ( KocH ) = " Stereoplasm " ( FRECH , non LINDSTROM ) .
Summarizing my statements above , I may say that the primary streak ( " Primarstreif " ) of v. KocH signifies the median dark band , and wherever it is present it is the only true , lie structureless part of the skeleton .
Any crystals which may appear in it are due to later infiltration of water .
The fib ?
" Ous deposit may be from first to last , i.e. , from median plane to surface , similarly oriented and similarly constituted , in which case there is no appearance of " stereoplasm .
" On the other hand , some difference in the actual constitution of the layers of deposit may give rise subsequently to differences in the crystallographical properties of the fibres , in which case one may speak of " primary " bands opposed to " stereoplasm .
" In KoCH 's description of Pholiclophyllum the " wall mass " may be looked upon as stereoplasm , but in his description of Cyathophyllum there is no such mass , and nothing therefore to represent stereoplasm .
Yet FRECH has drawn the following comparison between KocH and LINDSTROM:"According to LIND8'l'ROM 's description , a structureless calcareous deposit , which he calls Stereoplasm , occupies the sides of the septa in certain 'Tetracoralla ' genera ( Oyathophyllum , Pycnophyllum , Oyathaxonia , Zaphrentis ) in such a manner that the septa are distinguished from the surrounding mass only as thin , shadow-like ( ' Schatteniihnliche ' ) streaks .
Transverse sections of Oyathophyllurn mitraturn , a species cited by LnrnsTROM as an example , entirely corroborated this description , and agree , on the other hand , with KocH 's figures ( of Oyathophyllwm , sp , , referred to in column l above , p. 254 ) .
Accordingly , we may call the calcareous deposit at the side of the 'Primarstreif ' in KocH 's figure also ' Stereoplasm .
' Extending the sense still farther , I feel justified in also calling the abovedescribed compacting mass ( Ausfii.llungsmasse ) 'stereoplasm ' which unites the septal spines .
This mass appears similarly structureless and has the same significance as the stereoplasm , namely , to strengthen later the septa previously laid down"* ( Zoe .
cit. , pp. 931-932 ) .
Those two conceptions of " stereoplasm " are not identical ; the first of them applies the " stereoplasm " to the middle zone of KocH ( " b " in comparative columns above ) , while the second is word for word that which KooH indicated in describing the " wall " or outer concentric mass of Pholidophyllum .
After the explanation I have just given of the actual difference in orientation of the fibres , it is easy to interpret the limit between KocH 's middle and outer zones , and also to identify the " middle , ; zone with the inner layers of lighter crystalline deposit so often depicted in sections of various Palreozoic genera .
These are indeed the same layers * F RE CH , " Ueber das Kalk Geriist der Tetrakorallen , " " Zeitscbrift d. D.G.G , " SYSTEMATIC STUDY OF 1'HI'1 l\IADREPORARIAN TYPES OF CORALS .
257 which gave LINDSTROM and others the idea of a prirnary septum being double , and jlcmked on both sides with lamellar " stereoplasrn .
" But confusion began when a " dark streak " in the middle of the septum was said to be the same as the apparently " double " primary septum .
Then the question came to be , was the primary septum single or double ?
And all kinds of appearances in fossil septa were raised to the dignity of " primary streaks .
" FRECH now professes to simplify the matter by calling KocH 's axial " dark band " the " primary septum , " and ffll the rest " stereoplasm .
" The actual differences of KocH 's middle and outer zone go for nothing .
This first misconception on the part of FRECH led to another some years later .
' : , , Whereas , in 1885 , FRECH used the various expressions spine , trabeculcl , " , isoli"rte Pfeiler " = the whole " septal spine " ( KocH ) of Plwliclophylforn , in 1890 , he used the same three terms in a new sense , namely , for the isolated " dark points " in the " Primarstreif " ( KocH ) of an Astrreid septum .
And the " Prirnarstreif " of KocH was called " prima , ry septum , " an entirely new use of KocH 's term , never intended by KocH himself , so far as he had given expression in published , vork .
Thus , only the " dark points " in transverse sections of the Astneid septum are recognized by FRECH as the " primary septum " ( single or double ) of previous authors , and the whole of the ccilcareoits fibrous deposit of the septum becomes , in his eyes , " stereoplasm .
" It stands to reason that LINDSTROM would never have thought of inventing tt new term , viz. , " stereoplasm , " for the whule calcareous structure which has been called a septum since the earlieRt days of Madreporarian literature .
Yet that is the end which in 1890-91 is arrived at by FRECH ( cf pp. 99 , 100 above ) .
If now , the " primary septum " of FRECH is analysed , we find throughout his work no indication of its structure .
The " dark points " a1\#183 ; e regarded as cross-sections of " trabeculre " or " spines , " but this use of " trabecula " is not that of PRATZ , , who was the first to describe the microscopy of trabecular stn1eture .
We learn from FRECH that each " primary spine " in the " primttry septum " may be seen microscopically as a number of points , or as shorter or longer rods , variously placed-again , that it may appear light or dark ; all of which are true observations , but give no ground whatever for transferring the term " septum " to this structureless , mysterious central portion , and leaving all the calcareous substance to be called " stereoplasm " !
Again , HEIDER describes these central portions as " sometimes filled , sometimes empty , " and HlNDE states that they are occupied as a ru1e by deposit which he 1/ 1 CompareA .
1885 , p. 929 , footnote 6 , and p. 931 , footnote 1 , whel'e the tel'rns " pl'imary spine " and " trabecula " arc used as the equivalent of both the " dark , central area " and the " light fibrous deposit " in KoCH 's original sense of the septal spine of Pholidophyllum .
D. " Trias Korallen , " Palooontograph .
, 1890-91 , p. 22 , where the same terms are used as foe equivalent of the dark central area only : " Der aus Primardornen bestebende Primiirstl'cif ' ( I. a in comparative columns a1ve ) .
258 DR MARIA M. OGILVIE ON 'rHE MlCROSCOPIC AND lielieves to ha , ve been introduced la.ter , and therefore , to afford little clue with regard to the origina.l condition .
I have already given my own explanation of the delrlc points , and spots , and lines , and rods , whether those of the n P !
'.'imarstreif " ( Koen ) or those in the fibrous deposit .
My conclusion is that they are due to the subsequent disintegration of organic cell-material , under conditions of preservation favourable for the slow conversion of any orgauic pa.rts of the calicoblasts left in the skeleton into graphite , or , generally speaking , coaly compound .
In rapidly growing corals it would seem that a considerable mass of organic cell-material remains uncalcified within the " pockets " which are invaginated ::1t the apical edge of the septum .
And these pockets seem to have been on the whole larger and more irregular in Palreozoic than in recent forms , enclosed uncalcified portions being in consequence greater .
Yet we have seen how in the recent genus , 11fadrepora , large " cores " of organic uncalcified substance are present .
The black appearance of these cores too , is often as bright and glancing as that described by Kocr-r for tho Palreozoic genus Pholiclophyllum .
Probably the exact nature of the subsequent changes undergone by enclosed organic matter is influenced by the rapidity or completeness with which the cellular fragments are endosed in the skeleton .
Whatever the state of preservation , or the exact nature of the chemical changes , the so-called " primary septum " ( if we could succeed in separating it out at the growing surface of a living coral ) would be , in its original condition , always organic and cellular .
In its carbonized condition we seem to observe dark , ramifying stems or detached rods in the middle of a Madreporarian septum .
These merely represent the original axes around which the calicoblasts became converted into calcareous skeleton .
The lines may branch from an " area of divergence " in some cases , in others they may be irregular , in others non-branching and parallel .
These axial lines are specially pronounced at the lowest part , which is likewise the youngest of most septa .
I have observed in the case of some Turbinolidre , that the \lt ; lark lines .
develop in i , he septum of the mature cornl almost parallel with one another ( inwards and all but horizontal ) , but in the embryonic calyx , they are arranged at first in a fan-like , divergent form , much as in Astn.eidre : : , It is to the mature coral that we must look for the characteristic setting of the trabecular axes in the septa .
The process of calcification would seem also to be more vigorous in the mature than in the quite young polyp .
\#183 ; * '1.'hi8 observation is important , as it indieates that probably in the embryonic stages and eai-ly development of the skeletal parts , the relations of the septal fol , ls arn much the same both in Ast.rooid oo and 'l'urbinolidoo , a peripheral " area of divergence " being present in both .
Possibly the thick wall and special wcill-fold of 'l'urbinolidre is only an adaptation to the free-living , simple habit of the typical 'l'nrbinolid polyp ; all tho more may we oxpeot a variation such flS " coonencbyme " whr.n roef-lifo is ospeeially :ci , ssumod by an fLllied genus .
SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
259 2 .
Simila1'ity in Microscopic St1\#183 ; 'ucture of the Septa in Palwozoic and Recent Mad1\#183 ; eporaria .
The Table of Classification ( Part I. , p. 252 ) \#183 ; which I have given above is , I believe , fail ' evidence of the value which microscopic structure can have when applied to systematic affinities .
And I am convinced that the application of the same main principles as I have there used , to Palreozoic genera , would introduce uniformity in the classification of Madreporaria rugosa ( E. nnd H. ) .
These main principles may be set down under the headings , I. Radial Structw\#183 ; es , II .
Theca or Wall , III .
1nterraclial Structures ( tabulre , dissepiments , synapticnlre ) .
Under the heading of Radial Structures , the chief sub-divisions are yielded by the direction in which the centres , or better , oxes of calcification ( = central portions of the " septal spines " in Kocn 's Pholiclophyllum ) are developed in the radial structure .
Every variety is present amongst recent corals , from a horizontal position to a vertical , and the same is true of Palreozoic corals .
I shail confirm this statement b_y mentioning a few figures ( of Palreozoic septa ) , easy of reference , and naturally I select figures of sections made longitudinal1y , and , \#183 ; where poesible , in radial direction through the septum .
Seldom , however , does any longitudinal section show more than a dark axis and a , lighter " structun~less " surrounding substance .
Several figures given by FRECH't of Devonian corals are important .
A longitudina section of Cyathophyllum heterophylloicles , FRECH ( Zoe .
cit. , Plate 1 , fig. 2c , p. 30 ) , shows dark lines curving inwards nnd upwards .
The lines are " mostly broken up into rows of connected points .
Each row of points is laterally flanked by a clearer stereoplasm zone .
" . . . . . .
Phillipsastrwa pentagona Golclf sp. var .
mic1\#183 ; ommata , F. RoEM , shows more marked dark lines ( Zoe .
cit. , Plate 3 , fig. 13a ) in correlation with the presence of prominent strire ( " Verticalleisten , " " Carinre " ) on the septal surface .
I-Iaplothecia filata Schloth , sp. , Plate 4 , fig. 7 A , is like Phillipsastrwa ; I-Iallia prolifera , A. RoEM , sp. ( Zoe .
c\#183 ; it .
, Plate 7 , fig. 5u ) again shows the same general direction of the dark axes of calcification , but this same figure has for me a particular interest , as the dark lines diverge at the base of the septum in fan-like shape just as in the septum of Galaxea .
Yet , later , all tho " lines " bend at a very low angle inwards , just as in Turbinolids .
And thus an example is afforded in the Palreozoic genera of the observation w\1ich I have stated above on p. 258 .
Moreover , the figure shows at the same part the separate short lines or " rods ; , into which the long , curving , dark line breaks up .
This is an evidence that the short , " dark line " was in Palreozoic as it is in recent genera , a separate axis of calcification corresponding to Cl single growth-part of any particular trabecula of the septum .
Still another section may be cited from the same author , Endophyllum torositm ( " Zeit.schr .
, " 188fl , Plate 41 , fig. 6 , Gel ) .
The interest of this section is in the entire * " Die Korallenfouna des Obordovons in Doutschland , " " Zeitschrift d. D.G.G. , " 9 , 1885 , p. 21 .
260 DR. MAR .
IA M. OGILVIE ON TRf\lt ; J :MICROSCOPIC AND individuality of the " dark lines , " representing , in my estimation , the axes of isolated trabecular parts .
Thus interpreted , the Pa.l::eozoic genus Endophyllum would present an antetype of the arrangement of trabecular growth-parts in the Enpsammidre .
It is , in fact , a type in which the septal spines tend to a serial arrangement , although they show in certain regions the irregularity of the Cystiphyllid septal spines .
All those \#163 ; gures given by Professor FRECH show no signs of structure other than the presence of darker rods in the midst of a lighter-toned subskrnce .
Calcareous fibres are no , where figured .
At the same time Professor FRECH is perfectly justified in drawing the following conclusion( " Zeitschr .
, " 1885 , p. 934 ) :-"In Phillipsastrma , Ilcplothecia , Cyathop liyllum , and many others , the septa appear to be built up essentially on the same plan as the living cora1s with , however , the variation , that origina1ly a greater number of septal spines ( = " dark rods , " in figures ) are independently laid down and afterwards bound together by differently-formed Stereoplasm .
" Dr. R. SCHAFER showed the minute fibroHs structu1\#183 ; e of Phillipsastrcea ( on Phillipsctstrcea , " Geo1 .
Mag. , " Sept. , 1889 , Pbte 12 , figs. 8 , 9 ) and the similarity of the arrnngement of the fibres seen in longitudimtl sections of the septa with that which P.RATZ had demonstrated in the Jurassic genera belonging to the Thamnastrreinre .
The same author has told me that his examination of the Palreozoic genus HelioJ\gt ; hy llum seemed fairly conclusive that the " carinre " were the homologues of the synapticular bars in the Mesozoic Fungid::e .
In Dr. ScHAFER. 's opinion , the genus Phillipsast1\#183 ; cet't showed a microscopic structure essentially like the Thamnastr::ea .
The hnbit of the corallum and the connection of the calicinal di8cs he regarded as an antetype of the " astrmiform " colony , common to the Astr::eids and Thamnast~\#183 ; reids as early as Triassic time .
Professor NICHOLSON'S figure ( " Mar ; rnal of Palreontology , p. 247 , fig. 127 ) may be again referred to , since it shows the fibrous structure in Zaphrentis and Streptelasma septa side by side wjth the septa of Caryophyllia , and still farther confirms KocH 's comparison of Caryophyllia , with Pal::eozoic types .
In those cases the dark lines seem , so far as the transverse sections allow a conclusion , to be very obliquely directed inwards .
It is very certain that in a large number of Palreozoic genera the spines are directed horizontally inwards as in Pholidophyllu1n , and thus the idea gained ground , rather awkwardly expressed sometimes , that the septum was a " two-leaved 'structure ' takin , g origin in a fold from the wall .
" .
These examples suffice to exemplify the structure of the Pal::eozoic " Rugosa .
" 'I'o judge from the few detailed figures of longitudinal thin sections and the fa.r greater number of good transverse sections , their structure was even i\#183 ; n microscopic features identical with that presented in recent Madreporaria .
In so far as septa are concerned , I recognized in the Table of Classification five leading types , viz. ( 1 ) Astr::eid ; ( 2 ) Eupsammid ; ( 3 ) Turbinolid ; ( 4 ) Madreporid ; ( 5 ) Poritid .
The facts which I gather from sections of Pa}a3ozoic " ' septa lead '~ In speaking of Palooozoic genera , I shall follow NICHOLSON in his distinction of Oyat-hophyllidre , Oysti phyllidre , and Zaphrentidre .
SYSTElV[ATIC STUDY OF THE MADREPORAR .
IAN TYPES OP CORALS .
261 me to recognize in the majority of species referred to the genus Cyathophyllum , tne antetypes of the Astrceidce .
Amongst allied Cyathophyllid genera , antetypes are yielded of the Mesozoic representatives of th~ Fnngidm , whilst the Zaphrentids seem to me more directly the structural forerunners of the Turbinolidrn , and the colonybuilding forms most nearly related to them , the Oculinidrn .
The Cystiphyllids may be looked upon as the group of PaJ::eo:wic genera most eBtranged from recent Madreporaria .
They show typically at each growth-period an utterly new set of inv_aginations , without serial connection with the foi\#183 ; egoing .
The whole tendency 1\#183 ; n the history of the Mad1\#183 ; eporarian septum is towm\#183 ; ds sym , metry and regularity in this respect .
Within Paheozoic time this change had made itself markedly felt , to the almost entire extinction of the Cystiphyllids .
Dming Mesozoic time certain Triassic genera , like StylophyllU1n , or Jurassic genera , like Hctplarcect , include species showing transitional stages from a loose , detached condition of the septal spines to the more serial and compact arrangement characterising the trabecula ( cf. below , p. 301 ) .
Lastly , in Eupsammids , we possess analogous cases at the present day of the more primitive disconnected mode of septum formation .
I am not of the opinion thA , t Eupsam,1nfrt should be for this reason derived from the Cystipbyllids , but rather believe the actual Palrnozoic ancestry of the Eupsarnmidre to have been more immediate allies of the Cyathophyllids than are the majority of Cystiphyllids ; whereas the genus Cystiphyllum , itself is no very remote ally of the Cyathophyllidre , the genus Calceolct and its allies seem to me entirely isolated in their character , even among Palreozoic corals .
As a distinct case of atavism in respect of skeletal radial structures , the camcnchyme of Turbinaria may be mentioned ; its likeness to the irregular succession of spines in the Cystiphyllids will be granted at once , although the calicinal features show clearly the inherent family differences .
From the above comparison it appears that the building-up of the trabecular growth-part by minute calcified calicoblasts , which I have described in the skeletal structures of recent l\fadreporaria , may be safely extended to the skeleton of Palreozoic types .
Conversely the " septal spine " of PholidophyllU ?
n presents no chance identity with that of the recent Madrepora .
It may be looked upon as an example of the ' ' Rugose " growth-part , similar in plan to the " growth-part " of the recent .1lfadrepo1\#183 ; n. At the same time , it a.ffords a striking testimony of the want of change in the history of the fundamental relations between the Madreporarian polyp and its skeleton .
3 .
THR " CCENENCHYME " OF '1 PERFORATES " ( see Note on p. 226 ) .
The figure3 , to which I have referred in discussing the septa of Palmozoic types , serve also to show that the theca is composed of thickened port.ions of the septa .
NICHOLSON points out , that the " theca , " in the case of Z:=.i.phrentids , is probably of epithecal nature .
Comparing now Cyathophyllum with a recent Astrreid , the homologies of parts at once allow the broad outer peripheral disc beyond the inner tabulate area of Cyathophyllu , m to be recognized as the represen~ative of the " costal " area beyond the socalled theca of Astrreids .
We are familiar with the fr , .
, ct that the aboral polypal bodywall extends over this costal area of an Astrreid and , doubling on itself , b8comes then the oval , or peristornal , body-wall , which is continued as the stomodreal wa11 .
The extracalycinal , or costal portion of the Astrreid skeleton , corresponds with the " Randplatte , " or edge-zone ( vide antea , fig. 2 ) , of the polypal flesh , and it will be remembered that the mesenteries , in Ast.rreids , are continued from the visceral chamber into the edge-zone .
Looking back in the history of Astrreid and Fungid evolution , it becomes all but self-apparent that the origin of the polyp 's " edge-zone " is to be found in that part of the primitive Cyathophyllid polyp which clothed the " peripheral disc , " i.e. , the dissepimental zone of the skeleton around the inner tabulate area .
Moreover , we find tlmt the epitheca , bears exHctly the same relation to the septa in both Astrreid and Cyathophyllid .
It is deposited in Astrreids usually , a little below the actual surface-level of the calyx ; in Cyathophyllids , it is practically at the surface of the calyx , on a level with the edges of the septa .
We know that in living Astrreids the position of the epitheca relatively to the soft parts of the polyp is always the same , viz. , at the farthest extremity of the extrathecal canals .
The polypal flesh tucks downwards over the skeletal rim and doubles again upwards , the epitheca being formed below the doubled angle of the tucked fold as it is carried upward during growth .
The relative position of the skeletal parts in Cyathophyllids would indicate likewise that the epitheca had been formed by the angle of the polypal fold , but that the said fold did not tuck downwards to any distance below the septal edges ; it apparently wrink1ed itself concentrically around them , and thus gave rise to a rough , irregular " rugose " epitheca .
The inner , or real " wall " lining of this rugose epitheca might be solid through and through , or there might remain free spaces amid the irregular corrugations .
The main consideration is that the epitheca , in all primitive Cyathophyllids , lcept pace in growth with the septa , and since the septal deposit was usually very thick at the periphery , there was a strong septal and epithecal bulwark around the calyx whose various skeletal elements wme all grouped together under the general , descriptive term of a " wall " or " theca .
" But , in some cases , the septa atrophied more or less towards the epitheca , and their thickest poi-tion was somewhere ir ; ithin the calyx , genera11y at the boundary between the tabulre and the dissepiments .
Here an " inner wall " was former ! .
F.'urther , in Phillipsastrcea , Acervularia , and several of the later Palreozoic species of C'yathophyllurn , this inner wall became more marked , and the outer epithecal wall dwindled , until eventually it became scarcely developed between adjacent polyps of a colony .
In other words , the transition to the typical Astneiform colony had already been marked out in many Palreozoic types of the C_yathophyllidm .
Calices adjoined SYSTEMATIC STUDY O1~ THU : MADREPORARJAN 'l.'YPES OJ !
' CORALS .
263 directly at the peripheries of their dissepiniental zones , the outer epithecate walls became abortive , except at the outside area of the whole colony , and the inner walls , composed mainly of septal thickening , became correlatively more important a , ncl elaborate .
On the other hand , Pal:::eozoic types , such as Zapll/ rentis , Amplexus , and allied genera , in which the dissepimental zone was either absent or extremely 11arrow , showed no tendency to the degradation of the primitive epithecate wa1l .
Even when the formation of a new inner wall became indicated in certain colonial forms , e.g. , the Amphiastr:::eid:::e , with few exceptions the septa still remained in connection with the outer wall , and epithecal rings were still formed round all the individual polyps .
Modification seemed rather to concentrate itself upon the basal skeletal tissue .
In certain lines of evolution the dissepiments became spongy in texture , or even porous , and were incorporated with the wall-lining of the epithecal rings to form a light coenenchymal framework .
This I take to be the origin of the particular form of colony characteristic of recent Madreporas , Tu1'binarias , the Ocidinids , etc. In all of these the epithecal rings develop apparently only where a polyp is exposed at the periphery of the irregular , or ramifying colony .
I may further point out how the te !
ldency for the epitheca to be less developed within the colony in these groups , \#183 ; while the coenenchymatous elements became more so , may be fitly compared with the specialization of the inner , or wall lining of the epitheca in Turbinolid:::e .
In these simple corals the inner folds of the epitheca continue to grow upwards at the same rate as the septa and cost:::e , while the outer folds become , at any rate , in many wellknown types , subject to irregular , or partial development .
These innermost folds give rise to the so-called " true theca , " which has , like the coenenchyme , special centres of calcification ; the outer folds deposit epithecal material , which varies in its structure , shov.-\#183 ; s no special centres of calcification , and is not always purely calcareous .
It also occurs very frequently in the Turbinolidre that the epitheca bears some kind of armature for the bette1\#183 ; protection of the solitary growing polyp-spines , knobs , ridges , crests , \amp ; c. ; not seldom , root-like processes are given off like those characteristic of the rugose epitheca of Palmozoic corals .
In the living Turbinolid , these " rootlets " are canals in open communication with the visceral chamber of the polyp .
This is an im1wrtant fact when we compare it with the coenenchymatous canals of the Madreporid : : : e , \amp ; c. In my opinion , the coenenchy.matous canals , and the root-like processes , are homologous structures originated by the inner wall lining of the epitheca , the ptirticular modification depending upon the habit of the polyp-solitary or colonial .
I shall now cite a few examples of the occurrence of root-like processes , taken from DUNCAN'S account of recent Turbinolidre .
" In Rhizot1'ochus a.ffinis the epitheca conies up to the very margin , is striated and coarse , yet is inseparable from the wall , and , indeed , not to be distinguished from it .
The radicles are large and are offshoots of the epitheca .
" " ( " Madreporaria , ' ' Zoe .
cit. , p. 14 .
) R. fragilis : " The corallum has an exquisitely fine pellicular epitheca ornamented with curves and vandykes , which is 264 DR. MARIA M. OG I.Lvrn ON THrn MICROSCOPIC AND not to be distinguished from a wall .
There is no theca , in the ordinary sense .
The rootlets are small , and their cavities are continuous with those of the interseptal loculi " ( loc. cit. , p. 15 ) .
Stenocyatlms , PouRTAIES .
" The costal tube1\#183 ; cles are hollow and communicate with the visce1\#183 ; al cavity .
There is some doubt aibout the epitheca , some specimens having it according to LINDSTROM " ( loc. cit. , p. 25 ) .
Pleurocyathus , MOSELEY .
Corallum , " entirely covered by a thin , plicated , coloured , bark-like epitheca , which rises higher them the margin of the calicle " ( loc. cit. , p. 27 ) .
Dcsmophyllum , EI-IRENE .
' ' Iu some instances the base extends as a film of hard matter on the supporting body and in others there are rootlets .
Costre visible near the calice , irregular , often as crests , nodules or ridges here and there on the wall " ( loc. cit. , p. 11 ) .
Also " MosELEY notices that his great Desrnophyllum ingens is covered with an abunda , nt dense epitheca , and some forms of Desrnophyllum crista gctlli Ehr .
sp. have it and others have not .
But the epitheca is not separable from the wall .
" Schizocyathus , PouRTALES .
In Schizocyathus fissilis , LINDS'l'ROM states , " that the w::1 , ll proper between the septa consists entirely of the same sort of thin epitheca whi'ch surrounds the whole outside of the coral , * or , in other words , there exists no wall as a separate formation distinct from the epitheca , " ( loc. cit. , p. 12 ) .
It is unnecessary to give more quotations .
It will already be sufficiently clear that the recent Turbinolidre have retained , in many cases unaltered , the epithecal wall of Zaphrentidre .
A deposit from the interseptal portion of the body-wall may thicken the epithecal deposit internally , but such thickening has no immediate relation to what takes place outside the polyp at the thick wrinkled rim of the calyx , Whereas in all Astrreids the body-wall of the polyp laps over the edges of wall and septa , and forms an edge-zone ( " Randplatte " ) , in many mature free Turbinolidre the body-wall is pushed out between the septa into the hollows of a wall , or int.o rootlets !
And in this habit , handed down from untold generations , we have a reason for the non-existence of an edge-zone of soft tissue in these types .
It is , besides , the morphological equivalent of what takes place in the formation of the so-called " porous wall " and " crenenchyme " of Madreporime and Turbinarinre .
The difference is , that in the solitary coral the out-pushings led , primarily , to the surface of fixture , while in the colonial coral the out-pushings of neighbouring coralliteR ( getting more or less involved with basal tissue ) formed a crnnosarc , or became directly attached to each other as confluent rootlets ( cf below , p. 277 ) .
While this is no evidence that the l\fadreporinre and Turbinarinre are directly descended from Zaphrentidre , it is in favom of theil ' having a very much closer relationship with Turbinolid.re and Zaphrentidre than has hitherto been imagined , and at least of having a collateral ancestral type .
Hence , arguing simply from the homologies of parts which I have traced in the " - ' The italics are mine in all cases .
SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
265 ancestral and recent types , I arrive at the conclusion : ( 1 ) that there never was an edge-zone ( in the present acceptance of the term " Randplatte " ) in a large number of Madreporarian corals , among others , Turbinaria , Alveopora , Montipora , 1lfadrepora , Pocillopora , Stylophora , Scriatopo1\#183 ; a , and many of the simple Turbinolidm ; ( 2 ) that the wall itself , i.e. , the " true theca , " is in the free living , simple Turbinolidm a modified form of the purely calcareous inner layers of " rugose epitheca " as seen in more primitive types .
And just as in Zaphrentids , these inner layers are sometimes incorporated with a narrow " dissepimental " and ' ' costal " area .
\Vhile , in the colonial types the " ccenenchyme " is either a modified form of the same primitive skeletal elements , or a particular development of the canaliculate or other outgrowths from the periphery of the polyp .
In the latter case , the homologies of parts still show the connection between the spined processes , outgrowths , and channelled rootlets of the Turbinolid wall and epitheca on the one hand , and the spined or tuberculate adornments and channelled passages in many forms of ccenenchymate colonies on the other .
I have refrained from giving expression to this view in the first part of my paper , since it may not be regarded by some as fully proved until we have more anatomical research on recent Turbinolidm .
' ' BouRNE and FowLER have suggested that the " edge-zone " in some of the perforate types may be in a retrograde condition .
rrhis may be the case in some Eupsammids , but fossil evidence is , it seems to me , in favour of the ccenenchyme in the MadreporiJm , \amp ; c. , being a distended form of wall from the first ; there is no fossil evidence to show that it may have developed as an extended visceral chamber .
It is worth repeating here the familiar fact that costm are present in these types which have no necessary relation to the septa of the peristome .
Moreover , the discontinuity of the septa and costm ( even when placed opposite each other ) is one of the most characteristic features of cmnenchyrnes in the " perforates .
" Generally speaking , it may be said that the distribution of costm on ccenenchyme depends on the particular position of the folds of ccenosarcal flesh around the polyps , and differs considerably in different members of one and the same colony .
Except at the innermost wall-rim these folds :=1 , re not influenced by the polypal system of septa and mesenterial loculi ; thus any multiple relation between the number of septa and costm is usually lost at no very great distance from the polypal rim .
It will be remembered that the costm marking the epitbeca of Madreporaria Rugosa are likewise independent of the septa in their position and number , and apparently depended on the particular position of the folds , varying at different periods of growth .
These relations are quite distinct from the strict symmetry of number and position which controls the septo-costm characteristic of typical " edge-zones .
" * The actual incorporation of dissepimental tissue in the wall has also been made out by Dr. RFJIS in several simple corals of Cretaceous and old-Tertiary age , although he did not apply his observation to any further cases of homology .
I may quote the genus Hydrophyllia .
( REIS , citat .
p. 293 .
) 2fi6 DR. MAR[A M. OGlLVIFJ ON THE MICROSCOPIC AND There are , however , other non-perforate ccenenchyrnatous types , such as Lophohelia , in which an edge-zone of varying length has been proved to be present .
( Vide FOWLER , " Ma\lt ; lrepornria , " 'Q.J.M.S .
, ' cix .
, pp. 4 and 7 .
) Observation of several allied fossil types shows that in most of the Oculinidre a septo-costate skeletal area , corresponding presumably to an edge-zone of polypal tissne , lengthened itse]f out in proportion as the septa , and the inner calyx generally , outran the periphery in rapidity of growth .
At the same time , the influence of the mesenteriaJ loculi ( and of course the septal symmetry ) woul\lt ; l become extended more and more to the folds of the body-wall enclosing theca and epitheca .
In these cases I consider the histor_v ' of the " edge-zone " not one of retrogression but of progression , signalizing the higher individuality and specialization of the polypal units within a colony which was originally solely coenenchymate .
The best develop\lt ; :id " edge-zones " are , however , beyond doubt found in the Astrminm and Fungidre , and are inherited from tho broad outer disc of Cyathophy llids with its septal continuations and dissepiments .
In the living Astrreicls and Fungids all the corallites of a colony are practically united by their calicinal , ~dge-zones .
To sum up the views expressed above , I give a general tabular form of the skeletal relations in some of the leading family types .
I place , as the heading of the three -paraJlel columns , terms which were first given by German zoologists to " plates " of the soft tissue clothing different parts of the periphery of the polyp , beside each Q. am I give the skeletal tissues , which , in my opinion , represent these particular " plates " of the soft body-wall distinguished in zoology .
SYSTEMATIC STUDY OF THE MAlJREPORARIAN TYPES OF CORALS .
267 / I " Innenplatte .
" I " Randplatte .
" I " Aussenplatte .
" [ _ ------\#183 ; -------\#183 ; -------1---------Pseudotheca .
Septo-costate area .
True thec:1 \#177 ; coenencbynrn ; epitheca .
- ------------------~-------------Cyathophy llid::e ( Pahcozoic ) Astrreinre ; :Fungidre ( pro parte ) ; recent types Zaphrentidre ( Palwozoic ) Turbinolidui parte ) ; types ( pro recent M. Pe1forata E.I-I .
excl .
Eupsammidm ( pro parte ) ; recent types Often present as an 'inner septal thickening at the periphery of the so-called " tabulate area " forming a pseudothcea within the calyx Alw:1ys present as a septal thickening and always within the calyx Often present as a septal thickening against the thecano inner pseudotheca Scptal thickenin ! !
occurs as pnrt ~\#163 ; the thecn , never within the calyx .
No rnner pseudotheca " " Edge-zone well represented 111 the fossil skeleton by the presence of n , bron , d oute1\#183 ; disc wit , h septa and dissepiments Edge-zone typically preseut , as an outer disc - represented in the skeleton by both radial and basa.1 elements Outer wall of sept::Ll n , nd thecal thickening present as tho lining of a concentric rugose usually well-developed epitheca Outer w:111 very seldom present ; the epithcc:1 retrngradc on individual polyps ; limited ns a rule to the periphery of colonies and often little developed \#183 ; beyond embryonic stages ---- 1-----------Edge-zone seldom indicated m the fossil skeleton-a narrow disc of dissepiments is present in some genera , but nearly , or wholly identified with the wall Edge-zone absent or very narrow the corresponding skeletal parts al'e embodied m wallthickening , , " Strong thecal lining and epitheca present ; may be drawn out as spines , muro-costm , rootlets , \amp ; c. Strong thec:1 alwn-ys present , with rnuro-costm in contact with the epitheca as far up as the latter dernlops , above that , free .
Thec:1 and epitheca may bear various protective outgrowths as m Zaphrentid:n .
In colonial allies ( Amphiastrrnidrn ) tho epitheca ascends round each corallite Theca al w:1ys present .
The coonenchyme is a grooved or channelled structure incorpomtiug mural n , nd muro-cost::Ll elements and clothed with epithecal tissue BOURNE ( Anat .
of Mussa a11d Euphyllia , loc. cit. , p. 48 ) has stated his agreement with KocH regarding the epithccal nature of the " wall " of Flctuellum from anatomical grounds .
I limit , however , the actual epitbecal tissne to the somewhat .
organic , external coat of coral .
The chief fact is that the wall of Flauellum comprising septal , n:iuro-costal , and mural calcification , is solid and is ere1\#183 ; ywhere in contact with this epithecal coat .
So far as I know it has not hitherto been suggested that a simple solid-walled Turbinolid , such as Flabellu m , without an edge-zone , gives the clue to the absence of the edge-zone in the JJ !
[ _ p e1forata ( excl .
Eups .
) .
The walls of the M. perforata , excl .
Eups .
, it is true are channelled , and the wall pores open into vesicular and muro-costal coenenchyme , but the structural homologies of the parts are perfect , and the essentin , I relations of the epitheca .
This would perhaps not have occurred to me if I had not been first struck by the identity , in groups apparently widely apart , of the microscopic structu1\#183 ; e of the : : ; epta and of the wall as well as the similarity in the internal plan of the calyx .
For example , the columella of Tu rbinaria mesenterince is entirely unparalleled in Astrreinre and Fungid:=e , yet it has plenty of likeness to that of various Turbinolid genera. .
Then take the basal structures , in several ccenenchymatous allies of the old " perforate " group , e.g. , Oculinids , Pocilloporids , etc.- they compare very closely with Turbinolid endothecal deposit .
The pre-eminence again of one long septum or the occurrence of a lamellar ( septal ) columelh is also common in all these groups .
4 .
THE TABULATE AREA OF p AL.iEOZOIC MADREPORARIA GAVE PLACE 'I'O 'l'HE CoLu.l\rnLLAR AREA oF NEozorc AND RECENT TYPES .
The most striking difference between Palreozoic and recent Madreporaria is associated neither with septa nor theca , but with the basal structures .
In Palreozoic forms , at the highest period of their development in Silurian and Devonian time , the ba : : , a , l structures were smooth throughout the calyx , in other words they were tabulate or dissepim1 , mtal : Yet a certain exception must be made in the case of many Cyathophylloid genera which showed a tendency to the formation of an irregular columellar structure in the central depression of the calyx .
The explanation of this tendency is very simple .
The typical Cyathophylloid calyx is shallow , and the septa mere narrow " ribs " marked out on the basal structures , very different from the prominent exsert septa of the typical Astrreid .
When we think how insignificant in height the septal invagination of the Cyathophylloid mnst have been and remained , we can readily understand that at its inner end it gave rise to no confused wrinklings , but rather smoothed itself out as it bent from the dissepimental into the fabulate area of the calyx .
In genera , however , where there was a relatively more rapid growth at the outer than at the inner end of the septum , the septal folds tended necessarily during each successive growth-period to be higher , that is deeper , at that outer part .
This was the first beginning of the changes which led to the exse1't Astrreid septum , with its well-marked " area of divergence " of septal and costal trabeculre , and its tendency to thicken laterally and form a pseudo-theca .
In accordance with the sloping downwards of the septal fold from the heightened septum , another change was induced in the Cyathophylloid form , namely , the gradual usurpation of the tabulate central area by the now deeper , longer , septal folds , and the SYSTEMATIC STUDY OF THE MADREPORARIAN TYPBS OF CORALS .
269 transformation of that area into a coZumellar area .
There is practicnJly no difference between the confused meshwork of septal ends in the centre of the calyx of certain Cyathophylloid genera and the so-called " spongiose columelh " of an Astrrnid genus such as Rhabdophyllia , .
Yet this change , which brought with it the disappearance of tabuloo , wac ; being accomplished by Cyathophylloids of Devonian age , and the Clisiophyllids of carboniferous time show marked change in the same direction .
The Zaphrentids , on the other hand , held more persistently to their tabulate base , as might have been expected from the narrowness or entire absence of a dissepimental zone , and the more turbinate shape of the calyx .
It is easy to trace how some of the modifications undergone by the " Zaphrentid " tabulrn went directly to the building up of a columella .
The convexly-arched shape of the tabulre in the genus Zaphrentis is indicative even in that genus of a tendency of the a , boral body-wall to be pouched upwards .
Still more positive evidence of the modification of the tabuloo to form a complicated columella is given by the Clisiophyllids .
The description given by WAAGEN of the columella in ' Lonsdaleia Indica ' is particularly suggestive .
He writes : " It is remarkable that in longitudinal sections there can barely be any tabulre obsel'ved outside of the colume11a .
" ( ' Salt Range Fossils , ' p. 898 , " Ccelenternta in Productus-Limestone .
" ) Not infrequently the tabulre seem to have been laid down so closely that they formed an almost solid calcareous deposit .
The Cyathaxonidre , E.H. , is a family including Palreozoic genera and one Triassic genus ; it is of special interest in the study of basal structures .
The essential feature is a prominent columella ; a second feature which may be regarded as correlated , is the scanty development or absence of basal structures .
Lindstromia , the oldest of the genera ( Ordov .
to Carb. ) , has well-developed septa reaching to the coZumella ; the septa are smooth on their surfaces and at their edges ; they have " pinnate " arrangement with regard to four principal septa , and a slight fossula is present .
The chief fact which at present calls our attention is that " stereoplasm " fills the lower part of the visceral chamber , while in the upper part dissepiments or tabulre are scantily developed ( cf NICHOLSON , ' Manual , ' Zoe .
cit. , p. 298 ) .
FRECH says the endothecal deposit is sometimes entirely stereoplasmic .
( ' Trias .
Korallen Paleont .
, ' Zoe .
cit. , p. 81 .
) In certain recent Oculinids in S eriatop ora , and in a few Turbinolids ( all of which I hold to be related to the ancient Zaphrentidre ) , the basal structures are at times individual laminre , i. e. , tabulre or thin dissepiments , and in the same coral may take the form of solid deposit , so-called " stereoplasm .
" We need not then place much importance on the difference in the basal structures of Lindstromia , and of the genus of the Cyathaxonidre next to it in age-Cyathaxonia .
The latter has similar septal and columellar features , but no tabulrn or dissepiments .
A Triassic genus , Gigantostylis , has been referred to this family by FRECH ( Zoe .
cit. , pp. 82 , 83 ) ; in it the septa are rudimentary ; a very slight deposit of stereoplasm round the base of the columella is present , as in Cyathaxonia , and the columella is extremely large .
Its i : ; tructure is finely laminated , the laminre becoming more and more convexly a.rch ed from below upwards .
This structure is worthy of note : since it marks a difference between these Cyathaxoni.dre and certain other genera , which appear in most respects allied .
Such a genus is the Silurian Duncanella ( cf NICHOLSON'S figure , Zoe .
cit. : p. 267 ) , which ROEMER had included in the Cyathaxonidre .
It has been described as l1aving no columella by some authors , by others as having a pseudocolumella .
To judge from NICHOLSON'S longitudinal sect.i011 , it has a distinct fasc icular columella made up of a number of pali .
The septa , besides , agree in stru cture with those of most Turbinolidre , but the same structural lines are not continued into the columella .
I take Duncanella to be an ancestral type of many genera included in the Ca.ryophyllacere , E.H. No fossu1a is present in it , and it has r adial symmetry .
The common feature in all these forms is the entire usurpation of the axial area , which is tabulate in the related family Zaphrentidre , by a large columella .
The P aheozoic forms have septa reaching the columella , the Triassic form has a circular trench around the columella not invaded by any skeletal structure .
At the same time it has retained only indications of the tetrameral symmetry possessed by L , indstromia and Cyathaxonia , and has no fossula .
I may here remark that the Jurassic genus Lingulosmilia ( KoBY ) , to which FRECH r efers , has less connection with this family than with some other Jurassic genera , ex. A mphiastrcect .
Opinions may differ regarding the best systematic treatment of such a family as the Cyathaxonidre .
FRECH has placed Gigantostylis as an independent sub-family of the Cyathaxonidre under the Madreporaria rugosa , E.H. ( = Pterocorallia , FRECH ) .
NICHOLSON , again , has removed Gyathaxonia from M. rugosa to the Turbinolidre , but has left Lindstromia , at any rate , provisionally , under M. rugosa , in the fami .
lie Zaphrentid.re .
He gives as his main reason the pinnate arrangement of the septa , in which it " agrees with the typical Rugosa .
" The tetramera1 symmetry of Cyathaxonia and the presence of a fossula are not regarded by NICHOLSON as a sufficient basis for separating it from the Turbinolidre .
Speaking generally , I incline to Professor NICHOLSON'S position in the matter .
As regards Linclstromia , however , I would make an exception , whern the characters of the corallum so entirely ::igree with later forms , and the pinnate arrangement is still recalled by tetrameral symmetry .
I would include it among the Turbinolidre .
It is otherwise with a genus like Zaphrentis , between which and any Mesozoic genus there are no links preserved to us , of si\#183 ; mple corals , showing intermediate stages between the loss of the pinnate development of the septa and the gradual advance towards radial arrangement within the polyp .
Such a genus must be held apart from Mesozoic types ( cf " Turbinolidre , " p. 332 ) .
The Cyathaxonidre therefore and the genus Duncanella give indications which cannot be mistaken of the signi\#163 ; cance of the elaborate pa1iform or fasciculate ColuSYSTEMATIC S'l'UDY OF 'l'Hl !
: MADREPOR .
ARIAN 'l'YPES OF CORALS .
27 L mella which is seen in so many recent Tnrbinolidre , and is correlated in that family with the almost entire absence of " endothecal " basal structures .
I do not mean by this that all Turbinolids with such columellas are descended from Cyathaxonidre ; simply that their ancestry may be directly traced to the Zaphrentoidea , and especially to types having a large axial area free from septal structures .
The flat central bodycavity , supported on the tabulate base of some primitive Zaphrentid genera , gave place to the much deeper body-cavity-suspended over a central pillar , and often twined round it-presented by more advanced descendants among the Turbinolidre .
I look upon the Tu1\#183 ; binolid columella then , as in many cases , a modified " basal " ( tabulate ) structure .
On the other hand , a number of Turbinolids have a laniellar columella .
Those find antetypes in such carboniferous Zaphrentoid genera as Lophophyllum , in which the counter-septum is prolonged far into the middle of the calyx .
Sometimes the principal septum was thus prolonged in Palreozoic genera .
In either case the fundamental point is , that the columella of certain recent Turbinolids can be traced with security to a radial , i.e. , septal structure modified from primitive Palreozoic types .
Amongst these Turbinolids I include now several genera previously classified by EDWARDS and HAIME amongst the Eusmilinre ( fam .
, Astrreidre , E.H. ) .
Where it was a septum which became marked out as a columellar support , the axial line of symmetry remains longer evident , and is indeed still present in many Turbinolid genera .
An appearance of bi-lateral symmetry is thereby frequently shown by the skeleton , whether the soft parf , s partake in it or not .
I may also mention the cristiform " main septum " in species of Enallohelia , Turbinaria , Madrepora , Pocillopora , etc. , which has assumed a collumellar function .
It may seem a small thing in itself to follow out the evolutionary history , so to speak , in the columella in Madreporaria .
And to bring together evidence of its having taken origin as a modification in one case of tabulre , in another of a principal septum , in another of a meshwork of septal ends and basal structures .
In all cases we read a lesson with which zoologisfa ; have already made us familiar in many other chapters of animal life , viz. , the economy practised in the workshop of evolution , how old structures become adapted to perform new functions .
Where no adaptation takes place the tendency is for these old structures to retrograde or vanish .
The fertility of resource in the adaptation is shown by the fact that both radial and bas:oil structures could be transformed into a central columella .
The new requirement on the part of the coral polyp was apparently always one and the same , and it was attained by the simple means of a central support in the visceral chamber .
Before we consider what the new requirement was , it is important to note that the different skeletal modifications to produce a columella must have been in every case associated with a change of shape in the skeletal support of the polyp 's visceral chamber .
The latter , from being flat , or saucer.shaped , assumed cup-like proportions , but was fitted ovm a ceutra , l stalk1 leaviug thus a deepened circular trench in the middle of the chamber .
The examination of living polyps shows well enough the use of this trench around the columella , and how well this particular shape suits the looping-up of the mesenteries and the protection of the gonads .
It is , therefore , no very far-fetched conclusion that the gradual appearance of the columella in fossil types of Madreporaria marked always a distinct step in the polyp 's evolution , a step which was associated with the support and efficiency of the mesenteries and their :filaments .
Passing now to Mesozoic and recent types , which have probable affinity with the Palreozoic Zaphrentoidea , I may mention certain " Eusmilinid " genera , more especially the Stylina group ( cf Stylina , Pleurostylina , Cyathophora , Coccophyllvm , etc. ) , and the Trochosrniliacere , E. H. ( cf Trochosmiiici , Ccelosmilia , Stephanosmilia ; Pleurosmilia , Axosmilia , Conosmilia , etc. ) .
In those examples some show a pillarlike columella indicating the invagination into an elongated pouch-like form of the aboral body-wrtll ; others have no invagination and form either dissepiments or tabulre ; others .
combine a pillar-like columella with basal dissepiments ; still others have a " lamellar " columella formed in a fold continuous with one , two , or more septal folds , or in the same radial direction , but deeply notched and giving rise to " pali " ; still others show a winding round of the invaginated pouch or .
pouches during their upward growth , giving rise to a columella like that of Conosrnilia ; dissepiments may be richly or sparsely developed .
Caryophyllia , Turbinolia , Flcibellum , and many genera belonging to the Turbinolidre show similar varieties of the columella ; the aboral body-wall is , however , in the Turbinolids almost entirely dependent on the columella as a basal support in the upper part of the calyx , having few or no dissepiments or tabulre .
The calyces of Madreporidre resemble Turbinolids ; on the other hand , the Pocilloporidre possess invariably tabulre as basal support .
Although those recent types show an extraordinary variety in the particular character of their basal skeletal deposit , they agree on the whole in keeping a central area mo1\#183 ; e or less distinct from the radiating septa , whether that area presents a lamellar , pillar-like , or coiling colurnella , or none at all .
In the latter case , other skeletal characters always suffice to point out the relationship of the genus with more typical columellate forms .
On the other hand the Astrreinre and Fungidre carry the radial structures into complete coalescence in the central area , and the columella has often been jn consequence distinguished in typical genera belonging to those groups as a " pseudocolumella .
" While I place no value on the presence or absence of a columella in nearly-related genera such as Stylina and Cryptocamia , there is a broader systematic aspect of the columella when it is studied in distinct generic types .
It is well known how much wider the generic title of Palreozoic forms is in its systematic significance than that of Neozoic and Recent .
If a classificatory system were to be uniformly carried out .
, nearly all the Jurassic genera at present classified under Trochosmiliacere , E.H. , SYSTJ\#163 ; M.A.1'IC STUDY OF THE MADREPORARIAN 'l'YPEB OF COB .
A.LS .
273 1night be included under a single generic title Trochosmili\#183 ; a , and the single genus would have about an equal rank with that accepted for any one Paheozoic genus such as Zaphrentis .
This one example is enough to show that only when we gather together the closely-rtllied N eozoic genera into broader groups can we hope to trace the main lines of descent among Madreporarian corals which may be founded on outstanding features of such group8 .
Looked at in this light , I think it of importance systematically to note the sometimes tabulate , sometimes dissepimental nature of the basal structures in the Mesozoic and Tertiary group of Stylinacere , the frequent sparseness of dissepiments in the Trochosmiliacere E.H. , also mainly a Mesozoic and Tertiary group , there being few Recent genera ; again , the independence and varied form of the columella in those two groups when it is present .
All those features show ( I ) that the Eusmilinre , E.H. , are , in the main , intermediate between Turbinolidre and Astrreinre ; ( 2 ) that their ancestry is to be sought , not in typical Cyathophylloidea , but rather in Paheozoic genera , intermediate between Cyathophylloidea and Zaphrentoidea .
NICHOLSON , while distinguishing those two Palreozoic types , carefully shows how they merge in one another ; and so do their descendcmt~ .
Summarizing ( cf also p. 318 ) , the above considerations , I may say that I look upon the columella and pali of Turbinolids as the homologue of the tabulate area of Zaphrentids , the septa representing the originally short septa of Zaphrentids in the limited dissepimental area .
Correlated with their shortness , is , I believe , the fact that they never developed in the mature condition a well-marked area of divergence , and that their axes of calcification are usually very obliquely placed , or point almost ho1\#183 ; izontally inwards to the central area .
Further , the " sunken calicles " charactel'istic of the ccenenchymate allies of the Turbinoli\lt ; lre are also correlated with the smaller prominence of the septa and the direction of the axes of calcification .
The primitive shape of the calyx of Zaphrentis and its Palreozoic relatives have been the main cause of the immense functional activity demanded of the mural and epithecal rim in their descendants ( vide antea , p. 263 ) .
And it will be at once seen that similarly the broad dissepimental disc of Cyathophyllum and its Palreozoic allies has been perpetuated in the endothecc1 ( whether dissepimental or synapticular ) of the Astrreimc and Fungidre , as well as in the importance of the edge-zone ( Randplatte ) in these same groups .
5 .
Precocity of the Radial Symniet1\#183 ; y in Cyathophylloids .
The broad dissepimental zone in the Cyathophylloid calyx indicates that the polyp had a broadly expanded peristomal and extrathecal disc .
This feature cannot have been other than favourable both for vegetative and sexual reproduction .
In the vegetative form of budding , great rapidity could be attained , and a bunch of peristomal and extratheca ; l buds developed simultaneously .
We may , indeed , trace back all the 274 DR. MARIA M. OGILVIB ON 'l'HE MICROSCOPIC AND wonderful variety in colonial " fissiparous " and " budding " Astneids to the primitive , expanded , oral surface of their Palreozoic ancestors .
Judging from the habits of many recent corals , the same surface probably sen-ed at certain seasons to give attachment to sexually-produceq .
polyps during the early stages of their development , and was thus functional as a nursing organ .
Both the ease with which vegetative budding was accomplished , and the protection afforded to young polyps must have encouraged the sessile , colonial habit , so characteristic of the descendants of Cyathophylloids .
Be this as it may , it will be readily granted that the broad disc of the Cyathophylloids was favourable for budding .
It gave also greater strength to the skeleton , since the septa were longer and could be more highly differentiated .
CorrelatiTely also the mesenteries may be supposed to have had fuller development .
An interesting observation was made by FowLER on the mesenteries in three living genera which depart little from the typical Palreozoic_ coral in their~ calycinal features : the genera Se-riatopora , Pocilloporct , and J.lfctdrepora .
\#183 ; He found that of the twelve mesenteries , six were reproductive or strongly developed in Madrepora , which were rudirnentary in Seriatopo1\#183 ; a. Of the remaining six in Seriatopora , , , _ two were reproductive-bearing , the others were well-developed but not reproductive .
In Pocilloporct , the same two as in Seriatopora were long and !
eproductive-bearing , in the other f\lt ; .mr a " mesenterial filament might also be detected " ( Fo~LER , " Madreporaria , " Q.J , M.S. , 1887 , p. 13 ) .
FOWLER remarks , in suggesting an explanation , " the tendericy observed in both Seriatopora and M. Duruillei towards . .
\lt ; the excl~sive assumption of function on the part of six mesenteries and towards a correlated retrogression ( ?
) on the part of the other six , has not attained to such a pitch in Pocillop .
( and M. aspera ) as in the other two forms .
" Exactly that point in this explanation which the author himself marks with interrogation , seems to me in the highest degree doubtful .
_ The skeleton of many Palreozoic forms shows 011ly one fossula , others have two fossu]re opposite one another , and the natural inference is that each of these fossulre were the pits for one large mesenterial filament .
Seriatopo1\#183 ; a bears out faithfully the primitive condition of an opposite pair of fossulre .
In each of the fossula one large mesenterial filament with gonad is present .
We have , therefore , good reason to believe that a similar pair of reproductive mesenteries were developed in bi-fossular ancestral types .
Wh'y then retrogression ?
Rather we may read the observation more simply and look upon Seriatopora as a more primitive genus than Madrepora and Pocillopo1\#183 ; a. When this primitive condition is departed from , it is only natural that considerable .
variation should be shown with regard to the particular mesenteries which became Rpecialized as reproductive or filament-bearing organs .
Pocillopora adheres so far to the tradition , Madrepora has become utterly inconstant .
At least one deduction may be drawn with fair safety , namely , that each of the twelve primitive mesenteries were potential gouad and filameut-bear'Brs .
SYSTEMATIC STUDY OF THE MADREPORARIAN 'l'YP.b:S OF CORALS .
27 5 If now we assume that any Pah-eozoic species got as far as Pocillopo1\#183 ; a in rendering a certain numbP.r of the mesenteries the bearers of small filaments , external conditions being favourable , they might act as valuable aids to reproduction and make themselves persistent in the type .
This accomplished , the change from bilateral to radial symmetry may be said to have been initiated in the species .
Inasmuch as the change helped towards perpetuation of the type , it may be looked upon as an evolutionary change re-acting on the whole organization of the polyp , and leading to better equipment of parts , muscular and digestive as well as reproductive .
For fuller development of the inner ends of the mesenteries demanded corresponding accommodation \#183 ; within the polyp and a certain modification of the form previously assumed by the aboral body-wall in the central part of the polyp .
This might take place in various ways , such as the formation of several instead of only one or two basal " pits " ( this I have observed in a Jurassic genus ) , the continuation of the septal invaginations into the centre of the calyx and a general deepening of the interseptal loculi ( Astneicls , Fung-ids , etc. ) , the formation of a styliform columella in the middle of the calycinal base , leaving a circular trench between it and the septa the said circulctr trench being homologous with a ring of pits 01\#183 ; fossulm ( Stylinidre , Cyatlwxonia , 1'urbinolia , etc. ) .
Whereas many Palreozoic coral species belonging to CyathophyllU1n , A cervu.laria , Phillipsastra:a , and other genera of the Cyathophylloidea were precocious in the matter of radial symmetry and its correlated features , there was \#183 ; considerable tardiness on the part of the Zaphrentoid genera .
Only this can explain the fact that in the Jurassic coral fauna of Eu1\#183 ; ope a 1.'ery large number of bilaterally symmet1\#183 ; ical genera are present whose calycinal structures have the characteristic features , macroscopic and microscopic , of Zaphrentis and allied genera .
The main advance shown by these Jurassic genera is the loss of pinnate development of the septa , although the septa strictly retain in most cases a definite arrangement with regard to " main " and " alar " septa .
A fossula is sometimes present , more often absent .
Again colonial habit has largely been gained , each corallite , however , having a distinct rugose epitheca of its own which invaginates inward and upwards to give rise to septal structures .
Enough has been said to show how essentially Palreozoic such genera seem , a number of them having been described by KoBY* and group _d as members of the Madrepo1\#183 ; aria rugosa .
I have had very interesting specimens amongst the Tithonian Stramberg fauna , and have had occasion to enter fully into \#183 ; morphological features in that paper .
I have grouped them and the allied genera belonging to other geological ages as a new family , the Amphiastrreidre .
In all the genera in question the septa have retained appa1\#183 ; ently unaltered the essential septal .
structure of ancestral types mnong Znphrentoid genera .
* KoBY , " Polypiers Jurassiqnes . .
de la Suisse , " 1880-89 , Geneva 'Mem .
de la Soc. Pal .
Suisse , ' vtils .
7- 16 , loc. cit. , pp. 419-437 and pp. 571- 2 .
6 .
The Change from Bilater\#183 ; al to Radial Symmetry was an Adaptation to the increase in numbe1\#183 ; of the " Reproductive " Mesente1\#183 ; ies .
In the foregoing discussion I have assumed that for any reason whatever an additional number of the mesenteries became filament-bearing in different Palreozoic species , and have indicaten how such a change would bring with it correlated changes in the invaginations of the aboral body-wall , and , therefore , in the subsequent symmetry shown by the calicinal skeleton .
The assumption is thoroughly justified by palreontology .
It is admitted that the so-called typical " Rugose " coral has one , two , three , or four fossulre , seldom so many as four , and usually one or two ; farther , that these foseulre represent specialized " reproductive " or gonad-bearing mesenteries . .
On the other hand , many of the Recent Astrreids have all their mesenteries thus specialized , the filaments and thickened reproductive part being suspended into the hollow of the visceral cavity .
This I hold to be the main direction of evolution in the group of Madreporaria ; all other changes-respecting bilateral or radial symmetry ; .
tabulre , dissepiments , synapticulre ; one wall or two walls ; camenchyme or no crenenchyme ; - I look upon as subsidiary .
Such modifications may or may not take place ; at the most they are correlative features accompanying a leading change in the increase in number and higher specialization of the reproductive mesenteries .
It is useful to compare in a general way the form of the skeleton of Rerent types with that of Palreozoic antetypes . .
The Recent Astrreid has a much deeper calyx than the Palreozoic Cyathophylloid ; its septa are stronger , higher , and usually more firmly built .
The ends are as a rule firmly interlocked in the axial columellar part of the calyx .
We are now perfectly aware of the pleated muscular nature of the mesenteries , and their government of the peristomal disc ; and we may take it that the greater atrength and prominence of the septa in Astrreids is evidence of higher differentiation of the mesenteries in respect of all their functions than was the case in the Cyathophy lloids .
Moreover the Astrreid septum is much more completely corrugated , more ridged and furrowed , and more granulate than the septum of a typical Cyathophylloid .
And one cannot doubt that the complication in the septal skeleton was called forth by a greater need for support in proportion as the mesenteries became more muscular and corrugated .
The interlocking of the septal ends centrally is correlated with the continuation of the mesenterial invaginations from periphery to centre .
The central columella of Astrreids afford the mesenteries a means of firm attachment , one might almost say of improved leverage .
Again , as all the mesenteries of the first , second , and usually third cycles are long , they split up the body-space into a great number of loculi , and these mesenterial chambers are at once seen to be of advantage to the polyp for purposes of digestion and nutrition .
SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
277 The Zaph1\#183 ; entoidea retained longer than the Cyathophylloidea their comparatively free axial area and shallow calyces ( cf antea , p. 271 ) , and we find that amongst the groups , which I regard as their descendants , various methods have been hit .
upon for increasing the digestive space.* The usual method was the adaptation by means of cmnosarc to a colony-building existence , and the correlated development of the cmnenchymal \#183 ; canals or grooves between polyps .
In the canals digestion and the circulation of the nutritive-fluid went on ; moreover , during the periods of vegetative increase , the buds sprang from these extra-polypal canals , seldom from the polyp .
The canaliculate cmnosarc of ccenenchymatous colonies is , properly speaking , a modification of the loose , often wrinkled part of the polypal wall at the rim of the oral disc .
The development of cmnenchyme seems to have been invariably correlated with a very small size of the calyces and fewness of the mesenterial chambers , for example : Seriatopora , J.Vladrepora , Turbinaria .
In all these , the digestion and nutrition of the colony is mostly ca , rried on outside the actual body-cavities of the polyps , in the canals of a spreading cmnosarc .
Now , I must make a companion statement to that made above , p. 27 5 , about the J urassie Arnphiastneidre .
The great maf ority of liviny caJnenchyrnatous Madreporaria have cctlyces whose skeletal featiires have retained the primitive \#183 ; structw\#183 ; al characters of Zaphrentids .
The only exception to this occurs amongst the Eupsarnmidre , a famiiy whose calicinal features indicate their ancestral line to be Cyathophylloid , although certain genera o ; cur which have ccenenchymatous colonies .
The camosarc in ccenenehyrnatous Eupsammids has the same significance for the colony as has just been indicated .
Taking Dendrophyllia as a ccenenchyrnatous Eupsammid type , an active nutritive and digestive canalicular system is there directed along the main branch of the corallum .
The canals run parallel with the upward growth of the corallum , and are connected by branch canals \-with the individual polyps .
The polyps carry on the protective and predatory duties of the colony as well as the elaboration of the reproductive cells .
While the cmnosarcal canals are a material aid to the rapid digestion and nutrition , and their ectoderm is actively engaged in depositing skeletal layers .
These canals are a higher form of stolonal growths .
One may almost distinguish various grades among Madreporarian colonies .
The possibility of developing rootlets may be accepted as a character of the class Madreporaria .
To this inherent feature the first or simplest form of colony-building is due , viz :- I.-The Stolonal , or lowest grade of a Madreporarian colony , in which stolons grow * One of the most remarkable is Euphyllia , a fissiparous colony in which the stomodreum is continued downwards into a number of convoluted canals lined , like the stomodreum , with ectoderm .
The ectoderm cells here perform the function 0\#163 ; digestion otherwise exerted by the endoderm 0\#163 ; the visceral cl1amber .
BOURNE , who describes this , attributes the modification chiefly to the unusual vegetable.living habit 0\#163 ; the polyp .
out from the body-wall of the polyp either near the base , as creeping stolons , or from the sides as lateral stolons .
This grade of colony occurs occasionally in all recent Madreporarian families , but was particularly common in the Palreozoic types- more especially among Zaphrentoids and so-called " Tabulates .
" IL-The Ccenosarcctl grade where " stolons " grow out from the body-wall of the polyp at the upper eclge or lip of the polypal cup .
( a. ) If the stolonal \#183 ; apertures are left open during farther upward growth , the crenosarcal stolons form a meshwork of canals between the polyps of a colony and the " crenenchyme " or skelekil deposit which they lay down is correspondingly porous .
( b. ) If the stolonal apertures are closeJ as successive periods of upward growth ensue ; the crenosarcal stolons run only on the surface of the colony and the " crenenchyme " which they deposit is practically cvrnpact .
Those two varieties are of equal dignity , they are the expression of different adaptation to different external circumstances .
This grade of colony is permanent in the families of the Madrepori\lt ; lre and Pocilloporidre ( cf below pp. 336-338 ) .
It is one which is correlated with sr ; nall , simply-built polyps whose inclivid1.tality is always retained b~tt is little advanced .
The bilateral symmetry of the simple Zaphrentid corallum , the smooth character of the septa , the horizontal septal spines , and broad ta.bulate base , are features of such constant recurrence among the genera of the families Madreporidre and Pocilloporidre that their close relationship with the Zaphrentids cannot be doubted .
At the same time the 8 , ncestral types of these two camenchymatous families may have branched off from the \#183 ; main stem of the Zaphrentoidea at a very early period in the phylogeny of Madreporaria .
The ccenenchymatous colony is almost ct rnirnicry of clirno1\#183 ; phisrn .
The tentacular mouths and polypal cavities have definite functions-to capture the booty and fertilize the reproductive cells , while the crenosarc is an extensive gastral organ .
Looked at in this light , it is no great step from a recent Madreporid to a Palreozoic Heliolites and other genera of the so-called Tabulates .
But I have not specially studied this aspect of the question .
Be that as it may , the mimic dimorphism which prevails in these families stamps them at once as lower in the scale of colony-building animals , than are the Madreporarian colonies , in which each polyp carries on all the essential vital functions of its existence .
III.-The " edge-zone " grade in which the original oral discs of the individual polyps are well developed and form an area of greater or less extent between the actual tentacular rings of the polypal mouths .
The remote angles of the edge-zones coalesce with one another .
The polyps in such a colony are bound to advance in evolutionary features more than in the case of the polyps of " crenosarcal " grade , since each inrlividual polyp is a complete organism carrying on all the vital functions within itself .
It is in such colonies that the number of mesenteries is increased , the muscles more highly differentiated and radia , l symmetry attained permanently .
SYSTEMATIC STUDY OF THE :MADREPORARIAN TYPES OF CORALS .
279 This grade was reached very early by Cyathopbylloids , more slowly by Zaphreutoids .
It is a grade which was inherited in time by nearly all descendants of PalreoJ:oic Cyathophylloids , the Astrreinre , and the Fungidre .
The early Eupsammidre were mostly simple corals , they became only very slowly adapted to colony huilding habits and show at present every variety in their modes of adaptation .
The direct Mesozoic colony-building descendants of the Zaphrentoids are Amphiastrreidre ( see p. 27 5 ) .
In Jurassic forms an edge-zone might be temporarily or permanently formed , but , as might be expected , it was very limited , and radial symmetry , vas not attained .
The same is in smaller measure true of the Oculinidre , a large family of undoubted Zaphrentoid descent , in which the polyps are either radially or bilaterally symmetrical , and the colonies waver between the c03nosarcal grade and the edge-zone _grade .
They usually present an intermediate condition .
There is in most Oculinids an overlapping edge-zone into which mesenteries are coutinued , but it passes into a common camosarc without mesenterial loculi .
It would demand too much space to enter into a full discussion of bilateral symmetry , as shown in recent Madreporaria , and I hope to give the subject special attention in a later paper .
I may , however , indicate some other considerations in addition to those already mentioned .
7 .
The old-Jctshioned Character of the Recent Genus , Seriatopora .
* The genits Seriatopora , which ha , s amongst r\#183 ; ecent Madreporaria the smallest known nitmber of reproductive mesenteries , agrees in skeletal respects with the bilclteral ancestral type .
Two reproductive mesenteries are placed in Seriatopora on opposite sides of the polyp , symmetrically with regard to the mid-line or long axis of the elliptical calyx ( see below , diagram E , p. 290 ) .
The two septa on either side of a reprnductive mesentery are rudimentary .
In all , twelve septa are present , and their relations at the porous theca are similar with those described under Turbinaria ( see above , p. 207 ) .
The basal skeletal tissue is usually a compact calcareous deposit like that common in Turbinolidre , but in the nearly related Pocilloporct the calyx is shallower , and the basal tissue is in the form of tabulce .
The latter genus has also twelve mesenteries ; two oppositely pJaced reproductive mesenteries are well developed , four other mesenteries give occasional indications of mesenterial :filaments .
The septa of Pocillopora though twelve in number are short spines jutting into the caJyx , only one being prominent and long and forming an axial line in the calyx .
Madrepora has usually six mesenterial filaments , and six non-reprorluctive mesentAries .
Septal spines are again present , with an axial line formed by one septum or the coalescence of two .
The spines are sometimes more nume.rous 1'11an twelve ; again * MOSELEY , " Notes on 8eriatoporci , Pocillopora , " etc. , 'Quart .
Journ. Mier .
Sc. , ' vol. 22 .
Fowum , " Madreporal'ia .
, ' ' 'Qual't .
Journ. Mier .
Sc. , ' vol. 27 , August 1887 .
280 DR. .
MARI .
A l\J .
OGILVIE ON THE MICROSCOPIC .AND in some species six of them are more prominent great irregularity prevails , but ~is is well known , bilateral symmetry of the calyx is fairly constant .
The identity in their minute structure with the septal spines of the Silurian genus Pholidophyllum has already been pointed out .
In all those cases the axis of calcification is directed vertically inwards or almost so , and wall and septa repeat entirely characteristic Palreozoic types .
The important feature in all these genera is that the bilateral symmetry is a distinct advantage still to the polyp , and on this account has apparently been retained .
If we look upon the opposite " fossulre " at the base of the body-chamber in Seriatopora as alar , we have in the case of the others still the same alar fossulre , although a larger number of mesenterial filaments are accommodated in them .
The midline of di vision is simply a strengthening calcareous rod sufficient for the small size of the calyx .
On the other hand the calyx of Turbinaria is much larger , the same median rod is present , but it is surrounded by pali , and is actually grooved so as to permit of an uninterrupted fossular ring .
Correlatively the number of reproductive mesenteries in Turbina1\#183 ; ia is much greater than in Madrepora , FOWLER gives it from 17 to 22 .
8 .
The Fossula and the Pinnate Arrangement of the Septa .
NICHOLSON points out ( Zoe .
cit. , p. 281 ) the frequency in certain Palreozoic genera of the elongation either of the " cardinal " ( Hauptseptum ) or " counter " septum ( Gegensepturn ) into the mid-line of the calyx in columellar fashion .
The elongation was correlated with the rudimentary condition of the opposite " fossular " septum ; in short the needed skeletal support was supplied by a septal or leaf-like invagination of the part of the aboral body-wall resting on the central tabulate area .
And the invagination ceased at the special " pit " or " fossula " of the aboral body-wall left for the reception of the reproductive part of the specialized mesentery .
The position of the fossula , if single , might be in the dorsal or ventral portion of the dorso-ventral axis , or at one side in the transverse axis of the bisymmetric Paheozoic calyx .
If paired , the pairs corresponded with one or both axes , the dorso-ventral axis ( = " sagittal axis " of HERTWIG ) and the axis at right angles to it .
Iftwo pairs of fossulre occurred , it can easily be realized that the " pits " in the skeletal floor of the calyx might become practically coalescent and form a hollow ring or trench .
A case is described of a half-moon trench where two " alar " pits and one " cardinal " pit coalesced ( " Menophyllmn , " vide N1cHOLSON loc. cit. , p. 295 ) .
The carboniferous genus Menophyllum may be compared , in the peculiar shape thus attained , with one or two genera which I have had occasion to examine in the Up .
Jurassic , Tithonian fauna of Stramberg .
The .Jurassic genera ( ex. Amphiastrma ) show frequently a well-marked cardinal fossula with short septum .
Young stages of the calyx show four septa , precisely arranged as ill Tetnworalla .
Then a pair SYSTEMA'J.110 S'l'UDY OF THE MADREPORARIAN 'l'YPES OF CORALS .
281 of long septa develop on either side of the short main septum , and together with the long " alar " septa and somewhat shorter " counter " septum , make five long septa and one rudimentary septum in that stage .
In slightly more developed calyces another pair of long septa appear between the " alar " septa and those on eith er side of the cardinal septum-eight septa are thus present and remain the typical number of the first cycle in those genera , the whole calyx being entirely of Paheozoic character ( cf. diagram , p. 201 , " paired insertion of septa in Jurassic types .
" ) The carboniferous genus Pentaphyll'Um has five long septa and one short cardinal septum .
And it seems to me the general similarity which is shown with the carboniferous types by those of Jurassic age is no case of atavism , neither is it necessarily one of generic relationship .
We probably see in all of them indications of transitional stages which various genera may have passed through at different periods during increase of their fossular area .
The changes at the same time led to the Joss of marked bilaterality , since the caxdinal and counter septa owed their individuality to the ancestral single or paired reproductive mesenteries .
In other Jurassic genera ( Thecidiosmilia , KoBY , Lingulosmilia , KoBY , Opistophyllitm , aut .
) , and indeed in other species of Amphiastrma , the septa develop in pairs , but the main septum is long and all the septa between it and the two alar septa are longer than those on the anterior or counter septum side of the elliptical calyx .
The number of septa may be equal in the posterior and anterior parts of the calyx , but frequently there are more septa between the two alar septa anteriorly than posteriorly .
Another peculiarity of some of the species of these Jurassic genera is the crowding together of the inner ends of the septa on the posterior side , and in such a manner that these septa arrcinge themselr , es in a definite pinnate order of development with regevrd to the main .
septum .
In several Stramberg genera I found by polishing the posterior epithecal surface , that the pinnate paired arrangement of the septa could be demonstrated in the better preserved specimens .
The same young specimens of Amphiastrma , in which transverse sections proved the slow appearance of the septa in pairs according to KuNTH 's law , showed u.lso the pinnate order superficially .
But while full-grown specimens showed marked bilateral symmetry , I could :find no sign of pinnate arrangement externally .
The permanent number of primary septa in a calyx varies in different genera , 6 , 8 , 10 , 12 are all common .
All those genera of which I now speak may be demonstrated by reason of their septal and " thecal " structure to be allied with the Zaphrentoids of Palreozoic time .
In other words , Mesozoic corals , probably of Zaphrentoid descent , show pinnate arrangement of the septa in young stages of growth but not in mature calyces .
It is well known that among Palreozoic corals pinnate arrangement of the septa is not universally present in mature calyces .
KuNTH , FRECH , and other authors have been able to show that the pinnate arrangement is present in embryonic stages of certain species of Cyathophyllitm although absent in the mature coral .
Again , the pinnate arrangement is not necessarily associated with bilaterality._ The septa may be practically radial in the mature calyx , and yet pinnate insertion takes place ( ex. Palceocyclu.s , Streptelasm a , etc. ) .
It is , however , necessarily associated with a slow cleveloprnental histvry of the polyp , where tangential expansion of the calyx takes place mainly at certain active areas the posterior and two lateral areas .
The pinnate arrangement of the septa within the calyx is a feature dependent on the developmental relations between the radial lines of growth and the periphery of the calyx .
The primitive developmental history may be thus expressed .
New mesenteries and new septa appeared in paired fashion , the " pafr " being placed on either side of the main septum or on the abaxial sides of the two alar septa .
Contemporaneously with each new invagination , the body-wall distended peripherally , and thus effected a shifting round of the previously-formed invagination , so as to make room for the new invagination .
If the septal invaginations were very short , a small amount of distension at the periphery of the body-wall would be sufficient to free the successive invaginations entirely from communication with one another , and the calyx would ultimately show radial structures free from one another at their inner ends ( ex , Omphyma ) .
Even if the invaginations were long and extended into the middle of the visceral cavity , the peripheral distension might go on rapidly enough to render the inner ends free .
Very frequently , however , the im\#183 ; Rginations were long , and the distension of the body-wall at the periphery went on too slowly to allow the successive invaginations to become free throughout their entire length .
Then , inside the calyx , the septa might be seen diverging pinnately from main and alar septa .
Again , if distension were uniform in the axial-abaxial ( postero-anterior ) and in the lateral axes of the calyx , the latter might ultimately assume radial symmetry of its parts ( Streptelasrna corniculum ) .
On the other hand , if peripheral distension were not uniform-say , it was more rapid at the " alar " regions of insertion than at the " main " region then the periphery of the calyx necessarily tended to become elliptical , the long axis coinciding , in the case assumed , with an axial-abaxial line of bisymmetry in the calyx .
In a large number of the primitive Madreporaria , new invaginations were interpolated before complete development of the previous , and distension did not sufficiently free the successively-formed invaginations throughout their entire length .
Hence the pinnate position of the septa in relation to the " main " and " alar " septa was retained during farther growth , and gave rise to their " divergence " within the calyx , in accordance with KuNTH 's laws .
Already , in early Palreozoic time , the pinnate insertion of septa at fixed regions of the polyp ( calyx ) had been often departed from .
Cyathophyllwn heterophylloides , FRECH ( Devonian ) shows feebly-marked pinnate insertion of the septa at the periphery in the embryonal end of the calyx , and irregular radial position of the septa in the upward continuation of the calyx .
It is , moreover , only one of several examples which could be quoted from Palreozoic genera .
And such examples prove that SYSTEMATIC S'rUDY OF THE MADREPORARIAN TYPES OF CORALS .
283 already , in Palreozoic time , a tendency displayed itself towards the peripheral expansion of the calyx at early stages of development round the whole periphery , instead of at the primitive fixed localities , With peripheral expansion of the embryonal bodywall at any pla , ce is associated the insertion of new radial septa ] invaginations of the body- wall at that part .
The same tendency \.which appeared in mid-Palreozoic time in certain corals seems to have made itself felt throughout different groups at different time-periods , since the embryonal ends of many Jurassic corals ( Amphiastrreidre ) show still the pinnate insertion of septa otherwise characteristic of Palreozoic types .
Meantime , those families ( Astrreince , PseudoastnBime ) which are probably directly descencle\lt ; l from Cyathophylloiclea , show , in Jurassic time , no sign of pinnate insertion of sApta even in the embryonic calyx .
And in many of them a distinct cyclical arrangement of septa indicates a simultaneous development of the primary septa at a number of localities in the periphery .
A glance at the " Jurassic " circle in the " Phylogenetic Table " below ( Part II .
, p. 331 ) , will show that the genera belonging to the sub-families Astrreinre , Thamnastrreinre , and Lophoserinre were at that period in reality cyclical , or at least radial , types ; whereas those belonging to the Oculinidre and Amphiastrreidre were , in the majority of species , altogether backward in the attainment of radial symmetry .
In the Amphiastrreidre several cases of embryonic pinnate arrangement of septa have been demonstrated .
It is at least striking that the families-backward in this respect were also backward in another , namely , the rapidity in colonial types of vegetative budding within the " theca .
" Each bud was still completely developed as an independent member of the colony , and had its own theca and epitheca ( often erroneously called " theca " ) .
In addition to the epitheca of the individual , there was the general epitheca of the colony .
It seldom happened that a series of buds were surrounded by a common epitheca and themselves had no individual epitheca .
In these more primitive families the aboral body-wall of the polyp went through , at the giving-off of each bud , the complete process of peripheral expansion , accompanied by the appearance of new invaginations in a radial direction , whereas in the case of the Astrreinre and Lophoserinre , etc. , the aboral body-wall of the polyp curtailed the process .
Radial invaginations formed very rapidly , and grew np independently of any wall-ring of tissue .
The proliferation of polypal " mouths " went on with much greater rapidity , and massive astrreiform colonies were built up of fissiparous and meandroid habit .
In these colonies no epithecal rings or " true wall " surrounded the individual calyces ; all calyces were enclosed within a common basal epitheca .
In branching colonies , natural1y , an epitheca surrounded each branch .
I conclude , therefore , that the pinnate development of the septa was , generally speaking , thrust backwards in the life-history of the Madreporarian polyp , sooner in some groups than in others , sooner in some genera and in some species than others , 2 o 2 until what was a typical feature in the majority of Palreozoic corals became an extremely exceptional feature as time went on , was limited , in fact , to embryonic stages and then seemed to vanish .
Yet not quite .
For the primary mesenteries develop in the recent Madreporaria.n polyp in pairs according to an order of development exactly corresponding to that observed in the radial skeletal parts of Palreozoic Madreporaria .
Comparative series are given in Diagrams A , B , and F , G , showing the stages in the development of the mesenteries in the :Recent Astroides , and their correspondence with the primitive order of insertion of septa , still shown in certain Jurassic genera .
A similar general result has been deduced by ORTMANN ( " Beobachtungen an Steinkorallen von der Sti.dkLi.ste Ceylons , " SPENGEL , ' Zoolog .
J ahrbiicher , ' vol. 4 , p. 569 .
Jena , 1889 ) .
ORTMANN draws attention to the gradual incoming of the change from bilateral to radial symmetry among Madreporaria .
He points out that the tendency of corals to build colonies reacted against the upholding of the primitive bilateral symmetry .
BOURNE also emphasises this factor in bringing about the change from bilateral to radial symmetry .
The main arguments considered by ORTMANN ' are thus stated :-( 1 ) " Between the Palreo zoic , Tetracoralla , and the Hexacoralla of Secondary , Tertiary , and Recent time there is no essential difference .
" ( 2 ) " Bilateral corals are chiefly such as are of simple-living habit " ( i.e. , not colony-building ) .
( 3 ) " From Palreozoic time , when the bilateral corals occur in the greatest number , until Recent time , bilaterally symmetrical corals have become less and less frequent .
" ( 4 ) " The bilaterality of the Hexacoralla has been thrown back to the earliest embryonic stages .
" ORTMANN very justly places great importance on this last consideration , and draws from it and from the others the only possible conclusion : " We must , therefore , take it for granted that the so-called Hexacoralla are directly descended from the Tetracoralla .
" It is impossible to base a classification of Madreporaria on septal symmetry .
Will one include amongst M. rugosa , E. H. ( = Tetracoralla , , HAWKEL ) , the genus Cyathophylhtrn ; it will be also necessary to include a nurn ber of Jurassic genera , in which young calices show paired insertion of septa with reference to posterior and lateral parts of the calyx !
Between these gener\#183 ; a and many others with marked bilateral symmetry uf their septa , there are all grades of transition .
Again , certain genera belonging to the same family may have bilateral symmetry , others radial symmetry .
Even within the same recent genus , e.g. , Madrepora , certain species might , with a considerable show of right , be referred to the Tetracoralla , and others to the Hexaco1 \#183 ; alla .
There are , therefore , abundant cases in Mesozoic , Tertiary , and recent Madreporaria of marked bilateral symmetry of septa retained in the mature ca1yces .
In point of this feature , such types stand nearer a primitive " Rugose " type , in which pinnate insertion and bilateral symmetry of the septa were associated , than do certain Paheozoic types , in which the symmetry of the septa is radial in the mature calyx , and no pinnate iw ; ;ertion is observed .
SYSTEMATIC S'fUDY OF THE MADREPORARIAN TYPES 01 !
' CORALS .
285 In fact , if the feature of pinnate insertion were to be strictly adhered to as a great classificatory principle , many genera , Palreozoic equally with Mesozoic genera , would have to be split up in an utterly artificial fashion , certain species referred to the one group , and closely-related species to the other .
I therefore entirely disagree with the main orders of Madreporaria erected by EDWARDS and HAIME-M .
rugosa , lvl .
aporosa , JI .
perforata .
The " Rugose " Epitheca is present in M. aporosa in all simple forms , always as a common base around complete colonies , and even in certain colonies around each individual corallite .
The tabulre of M. 1\#183 ; ugosa are not peculiar to that order ; they are still frequent in Mesozoic genera , classified in various families , and can be demonstrated in a large number of cases to have gradually given place to conical structures , viz. , pali and columella .
The latter change had taken place in Palreozoic time in some cases , in others it was of later date .
~ilateral symmetry is also not peculiar to Palreozoic genera , which , indeed , are often radially symmetrical .
Thus there remains only the pinnate insertion of the septa , and this is not a constant generic feature in certain Rugose genera .
The conservatism in favour of upholding an Order , M. rugosa , is a palreontologist 's prejudice .
He wants , for convenience , a general title under which the Pt !
, lreozoie genera can be classified , and so long as it does not make much disturbance in the con ceptions of Palreozoic genera , he is not unwilling to admit a few Mesozoic genera under the same title .
But he finds it impossible to recognize systematically that the paired development of the mesenteries in the embryo of recent types , and their pinnate insertion with respect to the embryonic " main " and " alar " invaginations of the body-wall , is exactly the same feature as was shown by Palreozoic septa .
Nevertheless the fact is there , and the embryonic development of Astroides repeats in rapid stages the phylogenetic history of its race .
Few more beautiful examples can be afforded of this general evolutiouary truth .
Pinnate insertions of the radial invaginations in which septa were deposited may therefore be regarded as a primitive character in the Madreporaria , which , as time went on , gave place to cyclical insertion of the septal invaginations .
It was a change attained at quite different periods in nearly related genera .
Systematically , I would erect no order , no family , sub-family , or genus on this ground alone .
But when pinnate inser\#183 ; tion in the mature calyces of certain Palreozoic genera is correlated with a number of other systematic features , which give a general character to the corallum differing from that of any subsequent form , there is reason to erect a distinct genus , or sub-family of Palreozic age , as the case may be .
The main sub-divisions must not rest on the geological age , neither should they rest on features which can be demonstrated to be highly subject to phylogenetic changes of the most varied kind .
They must , if possible , rest on features which have remained constant for certain types throughout the history of the group .
Such features are presented by the fine structure of the calcareous skeleton .
It is well known that AGASSIZ thought the " Rugosa " were probably not Madreporarian corals at all .
Most palroontologists of the present \lt ; lay show in their work not only a recognition of the right of the " Rugosa " to be called " Madreporaria , " but also have placed a goodly number of Paheozoic genera not among " Rugosa " but among I-Iexctcorallct ( NICHOLSON , ZITTEL , and others ) .
A backward step has , on the other hand , been made by FRECH , * who classifies the Triassic corals under the Orders " Pterocomllia , " FRECH ( = Rugosa , E.H. , = 1'etracoralla , HAWKEL ) , and " Mcidreporaria : , ( = Hexacorallct , HJECKEL ) .
He has lately been followed by V OLz , i- who similarly draws a sharp line of division between " Rugosa " and ' ' Madreporaria .
" 9 .
Correlation in the position of " Directive Mesenteries " with that of the fossul(J3 ( axial , abaxial , ala1\#183 ; ) in Pal(J3ozoic Madreporaria .
LACAZE DuTHIERs ' account of the developmental stages of certain recent " Hexacoralla " types has been siuce frequently confirmed .
From it the number twelve would seem to have become fixed as the fundamental cyclical number for mesenteries and septa .
It is shown in diagrams A and B how the various stages proceed ; pairs of mesenteries appear in a definite position around the oral region , and , when six pairs have formed , the first cycle of twelve septa begins to take origin in the entoccelic and ectoccelic chambers .
The axial and abaxial pairs of mesenteries are called " directive " mesenteries , because in Actinians and in most Madreporaria their pleats of musculature are placed on the faces of the mesenteries away from each other , while in all the other pairs of mesenteries they are placed on adjacent faces ( see diagrams C and D ) .
Thus a middle line , the " sagittal " line of HERTWIG , can be drawn through the polyp between these axial and abaxial pairs , and the arrangement of the two halves is the same on either side of this mid-line .
Both in Actinians and in Madreporaria the presence of a sagittal line demonstrates the bilateral symmetry of the polyp .
In several colonial forms of Madreporaria no mesenteries are marked out as directive , the polyps of the colony being then radially symmetrical .
This change is a retrograde one , correlated with rapid vegetative increase in the colony and the loss of individuality on the part of the single polyps .
LACAZE DuTHIERS described the appearance of the early pairs of mesenteries as taking place at intervals of unequal duration .
The first two pairs appear soon after one another , then there is a longer pause before the third pair appears .
The ' " FRECH , " Die Korallenfauna der Trias , " ' Palreontog .
, ' 1890 , p. 80 .
t W. VoLz , " Die Korallenfauna der St. Cassianer Schichten , Vorliiufige Mittheilung , " 1894 .
SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
287 fourtl1 succeeds the third rapidly , then a , second longer pause ensues , Ll-nd the fifth and sixth then appe:1r almost Bimultaneously .
The general facts which LACAZE DuTHIERS observed in the development of the mesenteries are , as has already been Diagram A. Diagram B. 0 ) ?
l e ( ' 7n / Jl e a 0 C .
Diagram A.-Inscrtion of a pair 0\#163 ; septa after each pair 0\#163 ; mesenteries ( hypothetic ) wou1d result in the ectocwlic and entocO ' ; ;lic position of septa ( primary cycle 12 ) .
Diagram B.-Insertion of a pair 0\#163 ; septa after every two pairs of mesenteries ( " pause " of LACAZE DuTHTERS ) would result in the entoccdic position only of the sepia ( primary cycle 6 ) .
m. main septum ; c. count , erseptum ; a. alar b:epta ; ( mesenteries after LAcAzi ; : DU'l'RIERs ) .
288 DR. MARlA M. OGILVrn ON 'l'HE MICROSCOPIC AND frequently pointed out , in complete agreement with the fundamental facts in the pinnate insertion of septa in primitive Madreporaria .
The three " pauses " in embryonic development , occurring after the appearance of each two pairs of the first twelve mesenteries , have been presumed by one or two authors to represent the interval during which septal invaginations formed .
I have indicated in diagram A above , that there may have been , primitively , types in which septa appeared after each sing-le pair of mesenteries .
This would give a raison d'etre for the occurrence of septa in every ectocrnle and entocrele of a polyp .
" \Vhereas the other possibility , --insertion of one pair of septa after two pairs of mesenteries-would tally with cases of entoccelic septa only in a polyp .
Of the Diagram C. Diagram D. T'J TT -~~ ~ Diagram C shows 12 septa occupying the 6 ectocooles and t , he 6 entocooles formed by the mesenteries ( cf. diagram A ) .
Diagram D shows 6 septa occupying the 6 entocooles of the mesenteries ( cf. diagram B ) .
Note also that the developmental " pairs " of mesenteries indicated by the numerals coincide only in the cases of the 2nd and 3rd " pairs " ( the " directive mesenteries " ) with the niusciilar " pairs .
" two , there seems most ground for accepting the types with both ectocrnlic and entocrnlic septa as the more primitive .
On such a basis the doclecameral cyclical symmetry would be explained as in reality a companion primitive form with tetrameral symmetry ; while hexameral septal symmetry would seem to be associated already with signs of " hastening " of developmental stages .
It is certainly the case that the incoming of radial septal symmetry in Madreporaria was more often marked by the number 12 in the primary cycle than by the number 6 .
In using the word " pair " either of the mesenteries or of the septa , I refer to " pairs " as they develop .
There is , however , a possibility of confusing such a " developmental pair " with a secondary use of the word which has become fairly general , namely1 the pairs of mesenteries whose muscles work in connection with one another .
The accompanying figures illustrate the double use of the word " pair " as applied to the mesenteries .
LACAZE DuTHIERS ' use of the word includes not only the sense of place in the polyp , but also the sense of contemporaneity of development , each " pair of mesenteries " being simultaneously invaginated .
In speaking of the " pairs SYSTEMATIC STUDY OF THE MADREPORARIA .
N TYPES OF CORALS .
289 of mesenteries " in a mature polyp , only the sense of " place " is implied , and a " pair " may be composed of two mesenteries which were not simultaneously invaginated , but which are afterwards brought into connection by their musculature ( cf diagrams C and D ) .
For example , FowLER defined the " entocCBle " as the cCBlenteric space between the adjacent faces of two mesenteries belonging to one " pair of mesenteries .
" The term " pair " is applied here only in the sense of place and musculature ( ex. pairs 1 and 6 , 4 and 5 , in diagrams O and D ) .
By comparing the " entocCBles " in a mature polyp with the arrangement of the embyronic mesenterial pairs , it becomes evident that only the axial and cibaxial entocceles are contained within a " pair " of mesenteries in the sense of LACAZE DuTHIERS .
In other words , the " di1\#183 ; ective " mesenterics enclosing the axial and abaxial entocceles are paired in development , ctll other entocccles are enclosed by mesenteries belonging to different developmental " pairs " in the embryo .
I have shown this in diagran1s A and B , founded on LACAZE DuTHIERS ' results .
The numerals give the order in the appearance of the mesenteries in the embryo , and it will be seen that only the second and third " pairs of mesenteries " in the embryo remain next one another so as to form a " pair " in the mature polyp .
Leaving out of consideration the pairs of mesenteries connected by their musculature in the mature polyp , it will be seen in diagrams O and D that the muscles are placed on opposite faces of the mesenteries numbered " 2 " and " 2 , " " 3 " and " 3 , " in the diagram , and also on opposite faces of the mesenteries " l " and " 5 , " on either side of the median lateral line .
These two mesenteries are originally in the position of the transverse line of symmetry in the polyp .
It is true that " 5 " comes into muscular connection afterwards with " 4 , " and " 1 " with " 6 , " and th:us it appears ultimately as if only the axial and abaxial " pairs " of mesenteries had the muscles on opposite faces .
Embryonic stages , however , show that " directive " mesenteries were primitively present in both the original lines of bisymmetry in the polyp : axial and abaxial , i.e. , dorso-ventral , and also the line at right angles to it .
In the latter line lay the " alar " fossulre and " alar " septa of Palreozoic corals .
If we were to transfer this diagram of the principal twelve mesenteries in the recent coral to any typical calyx , introducing the fundamental fea.tures in respect of septa and fossulre of a Palreozoic coral ( say Oinphyma ) , these very simple considerations become clear , and , with them , the sense and bearing of " directive mesenteries " in Actinian and Madreporarian polyps .
The mesenteries with the muscles on opposite faces were- to judge from their position relatively to the fossulce-destined to be the specially modified gonad-bearing mesenteries ; u\#183 ; ere , in fact , functionally active in reproduction , in addition to the usual function of the mesenteries in supporting the body-wall .
The \#183 ; presence of " directive mesenteries " is therefore hereditary , and has its origin in a time when one , two , or at most four mesenteries or pairs of mesenteries were specialized for reproduction .
Now these mesenteries cannot be of any very special use , and we need not feel surprise to find they have become inconstant in Madreporaria .
In most Recent Madrepornria the greater number or all of the mesenteries are reproductive ; the primitive significance of the aJar pair has long been lost , and even that of the axial and abaxial pairs is no longer always retained ( ex. Lophohelia , ~Mussa , EupAyllia ) .
In many Jurassic genera there is evidence that the alar regions were specially modified , and not the axial and abaxial .
In the Recent genus Seriatopora ( diagram E ) it is distinctly the alar pair which are reproductive-bearing ; that Diagram E. Plan of the septa and mesenterial " pits " in S eriatopora ( aft.er MOSELEY ) .
The pair of mesenteries numbered 1 , are specialized for reproduction , and occupy the lateral ( alar ) " pits " or " fossulai " ( cf. pp. 279 , 280 ) .
is , the first-developed pair of mesenteries in the embryo are apparently those which are specially _modified as gonad-bearers in Seriatopora .
Doubtless it was indifferent whether the alar or dorso-ventral line of symmetry was retained in a calyx , and either or both were marked out for a longer period in some genera than in others .
The elliptical shape of many Madreporarian calyces still indicates one such line of symmetry .
HEIDER 's description of the mesenteries in J.11.adracis pharensis is especially worthy of attention , both for the question of bilateral symmetry and that of the musculature .
I quote it in full :The septa are exclusively entocrnlic , as FOWLER has also given for liladracis asperula .
The directive pair is plainly marked out i.n most sections , and the ( dorsoventral ) plane is directed in individual polpys of the same colony more or less towards the middle point of the colony .
The polyps seem by this means to be oriented to a certain degree according to a main axis .
This is plainly seen in the nodular , more acrogPnous colonies , whereas it disappears altogether in flat colonies .
I found , moreover , very often that the musculature unif ; ing the mesenteries in paired fashion is developed one-sidedly only there , where the mesenteries include septa ; on the other hand , the contractile threads are developed equally on both sides of the mesenteries where no septa occur , i.e. , close below the oral disc of the polyp .
This gives , therefore , the impression that the one-sided development of the musculature is in a certain connection with the septa ( " Korallen Studien , " IL , Madracis pharensis , 'Arbeiten aus dem Zoo ! .
Inst. zu Graz , ' 1891 , p .
317 ) .
I su~join diagrams F and G , illustrating interesting cases of pinnate insertion of SYSTEMATIC STUDY OF THE :MAOREPORARTAN TYPES OF CORALS .
201 the septa in one or two Jurassic genera , founded on persornil observations .
KoBY observed some of these stages in the corresponding types of the Swiss Jurassic fauna .
These cases occur within a group of genera whose other features declare them to be allies of Palmozoic Zaphrentids .
I have erected for these Jurassic types and their later representatives a new family of Madreporaria , the Amphiastrreidre ( seep .
275 ) .
Diagram F. Diagram G. \#183 ; m , m , CD CD C C 111 , m \#169 ; \#174 ; C Observed cases of the development of septa in pait\#183 ; s. The successive stages may be compared with .
those in the development of mesenteries ( diagrams A and B ) m = main septum , c = counter septum , a = alar septa , A comparison of the septal pairs with the " embryonic pairs " of mesenteries drawn in diagrams A and B shows very clearly that there were types ( and they were very numerously represented ) still existing in Jurassic seas , in which the slow , paired mode of septal development was adhered to ( cf pp. 280- 283 ) .
The same types counted usually eight .
, ten , or twelve septa in their so-called primary cycle ; if eight , the arrangement was bilateral ; if ten or twelve , it might still be bilateral , but showed a tendency to be practically radial .
From these facts , together with the known facts of embryology , it will be a\lt ; lmitted to be pretty well proved that the sub-orders Tetracoralla and Hexacoralla are based on a feature which became gradually in the course of time removed backward in the life-history of the individual until now it is presented only in embryonic stages .
The diagram of the recent polyp Seriatopora shows twelve septa bilaterally arranged , correlated with the retention of the primitive feature of a pair of fossuhe .
I have entered the numerals against the mesenteries in diagram E according to the same system followed in diagrams A to D , for the purpose of demonstrating the upholding of the significance of the " embryoni0 pairs " in this Recent coral .
10 .
BRIEF REVIEW OF MADREPORARIAN EVOLUTION , It is a familiar fact that many " radially-symmetrical " Madreporarian corals are still distinctly bilaterally symmetrical in young stages .
Fungia is a well-known example .
The wide-spread occurrence of bilateral symmetry in the polyps of camenchymatous colonies has been frequently drawn attention to in this paper ( cf p. 271 ) .
These are cases of bilateral symmetry among recent corals .
As soon as fossil corals are studied , the number of bilateral types is seen to be very great in the geological ages between Palreozoic and Recent time .
Bilateral symmetry is much more widely spread amongst Trias~ic and Jurassic cora1s than it has been the custom to admit .
It is a constant feature in many genera , and in others all stages of transition are seen between bilaterally symmetrical calyces , even in some cases with a fossular depression , and radially symmetrical calyces .
On the other hand , radial symmetry in Triassic and Jurassic corals had no greater connection with the numbers 6 and 12 than with 8 , 10 , 14 , and in fact all multiples of 2 .
Six is the first number at which radial symmetry is attained in the rnesenteries of the embyro , and as such , it seems to have become early fixed for the primary cycle in many cases , more especially where the calyces were small ( Oculinidre ) .
Moreover , if the septa are entoccelic only , the first cycle of septa can only have six and not twelve ( cf diagram B ) .
Even then , the number eight seems to have been very often hit upon ( Stcphanocamia , Astrocronia , Stylina ) .
The broadest internal feature of difference between Palreozoic and later corals is , perhaps , best expressed in saying that a single well-developed reproductive mesentery , or one pair , seems to have characterized the majority of the former ; while , with a few SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
293 exceptions possibly in the Triassic and Jurassic periods , reproductive filaments occur in later corals , at least in one pair , and usually in a great number of pairs of mesenteries .
The correlated features have already been noted , e.g. , the extension of the original fossula or a deepening of the calyx ; a columella for the better support of the mesenteries , and the adaptation for this purpose of the Palreozoic tabulre ; " hastening " in the processes of budding and in the development of the embryo .
It was not until Cretaceous time that the condition of radial symmetry became general .
It may then be said to have made itself a sine qud non in all colonybuilding types of Madreporaria except the ccenenchymatous families Pocilloporidre and Madreporidre .
With this exception , there has been universal radial symmetry in t~e Madreporarian corals of Cretaceous , Tertiary , and Recent seas .
Nothing more than the mere tradition of their ancestry affects the corals of these ages , displaying itself now and then in the form of a columellar mid-line , or in the appearance of bilaterality where , in reality , the parts are not bi-symmetric , and at the same time not radio-symmetric .
In short , a lamellar columella is probably a carefully-retained inheritance in some cases and in others a freak of atavism .
Where the strength of the skeleton can do without it , it becomes obsolete .
The same may he said of the pillar-like colurnella .
In the dwindling of this columella we are probably now seeing the end of the basal tabul(JJ , once common to all Madreporaria .
One great innovation , which has reacted against the development of a columella , was the synapticula and in general a lighter skeletal structure .
The synapticulate Thamnastrreinre branched from Cyathophyllids and appeared with Astrreime in Triassic and Jurassic time .
They were at once highly successful .
The Cretaceous period brought new forms of Astrreinre and the genera Cyclolites and Thamnastr\lt ; 1Ja chiefly represented the Thamnastrreinai , but Early-Tertiary':- time saw the complete ascendency of the synapticulate Lophoserinre , Funginre , and Eupsammidre over the Thamnastrreinre .
The Turbinolidre , E. H. , attained in the Cretaceous and EarlyTertiary periods to full development ; since then they seem to have steadily continued .
It is also important to observe that contemporaneously with a rich Turbinolid fauna , the " recent " Perforate types appeared : Madreporidre and also Poritinre .
The LaterTertiary period continued directly the Cretaceous and Early-Tertiary fauna ; at the same time , the form of the coral skeleton tended to become ever lighter , and fragile ccenenchymatous genera had special advantages over the more compact in reef-building .
The general character of the post-Jurassic fauna bears a distinct impress when compared with the Triassic and Jurassic fauna .
Amongst other features , the numbers six and twelve then came more and more to the front in radial symmetry , although fissiparous and multi-septate Astrreids continued the irregularity which had been natural to them from the first .
The number was wholly arbitrary , as is evidenced by * Early-Tertiary corals have been made the subject of a recent monograph by REIS:- " Die Korallen der Reiter Schichten , " 'Geogn .
Jahreshefte Bay .
Landesuntersuchung , ' II .
Jahrg .
, 1889 .
294 DR. MARIA M. OGJLVlE ON THE MICROSCOPIC AND the fact that it became often fixed in genera remote from one another , while nearly related genera , or even species , differed .
The broad conclusions whi_ch I draw from such a review of the Madreporaria are:- .
A. Pal(JJozoic.-A change , which was advantageous , showed itself in Cyathophy1lids and in Cyathaxonidre , as early as Silurian and Devonian times ; other types , which did not participate in the change , or could not rapidly enough adapt their disconnected habit of growth to it ( e.g. , the Cystiphy1lids ) , became extinct ; before the close of the Palreozoic era , rather abnormal " colume1lar " genera were fashioned ( e.g. , the Clisiophyllids ) , but were apparently unsuccessful ; at the same time a large group of genera ( Zaphrentids ) seemed to make no sign of change .
B. \#183 ; Triassic and Jurassic.-The Madreporarian fauna was a mixed one , including many unaltered , or little altered , Zaphrentids , a few true Cyathophyllids , a great number of altered Cyathophyllids ( Astrreidre , Thamnastrreime , \amp ; c. ) , and a smaller number of altered Zaphrentids ( Turbinolidre , Eusmilinre pro part ) .
The varied aspect of Astrreid colonies , and the graceful branching corallum of such members as the Oculinidre , are characteristic for the period .
The Triassic fauna is remarkable for its small size , so far as we yet know it .
In Jurassic seas , however , large simple corals were common ( Montlivaltia , Plesiosmilia , etc. ) .
The young forms show paired development of septa in certain examined cases , but cyclical arrangement ( primary cycle of uncertain number ) of the septa is the rule in the mature calyx .
C. Cretaceous , Tertiary , and Recent.-Lightness of the skeleton is the most marked feature , and the widespread occurrence of synapticulate groups gives quite another impress to the fauna .
The simple corals \#183 ; decrease in size , while compound corals , colonial and camenchymate types , show constant tendency toward farther modification .
Cyclical development of the septa is present in the embryo , and the number of septa in the primary cycle is very generally twelve .
Several large families of recent Madreporaria , Fungidre , Eupsammidre , Madreporidre , Poritidre , although their relationship to Triassic and Jurassic ancestors may be determined , present a Madreporarian fauna quite as distinct in its character from the fauna of those two periods , as the Early-Mesozoic fauna was , in its time , different from the Palreozoic fauna .
Without the Triassic and Jurassic fauna , it might be possible to speak of Tetracoralla and Hexacoralla ; with this intermediate fauna , however , the 1.l1.adreporaria afford 011e of the most brilliant examples of a single evolutionary MOTIVE reacting on the whole region of the body i'Y/ , the neighbourhood of the main part a.ffected , and inducing an important series of cor1\#183 ; elated changes .
SYSTEMATIC STUDY O.F THE MADREPORARIAN 'rYPES OF CORALS .
295 11 .
Recapitulation of Phylogenetic Changes .
It will be useful to recapitulate briefly here the direction in which evolutionary changes have acted on the two great groups of Palreozoic Madreporaria-Zaphrentidre and Cyathophyllidre .
I need not repeat the character of the changes and their various modes of correlation so as to produce the wonderful variety of our living Madreporarian fauna .
Recapitulating merely the actual changes , these were :1 .
An increase in the number of mesenteries , and more especially in the number of gonad-bearing mesenteries .
Higher differentiation of the musculature of the mesenteries and elongation , inwards and outwards , of the mesenterial loculi .
2 .
Greater firmness and lightness of the septa ; the grouping of the calcareous fibres became more and more complex , and gave rise in many families to a complicated system of ridges and granulations on the septal surfaces .
3 .
Instead of one , two , or four fossular pits for the corresponding number of reproductive mesenteries in the base of a calyx , the whole base became deepened , an axial style or columella projecting into the hollow of the deepened calyx .
A very great number of mesenterial filaments could thus be accommodated .
This change is the basis of the gradual incoming of radial symmetry in place of bilateral symmetry in the Madreporarian calyx .
4 .
The central tabulre were modified as columella and pseudoco1umella , and a large number of septa usually connected up at their inner ends with the columellar structure .
The architectural form of the calyx was thus considerably improved in respect of its skeletal strength associated with economy of space .
5 .
The " outer dissepimental zone " was also modified in a large number of families , and in a manner much like the modification of the tabuire .
By becoming round or erect , instead of being flat or vesicular , they gave place to synapticulce .
This change allowed the soft parts to remain very deeply sunk in the skeleton , even while upward growth continued at the lip of the calyx , and thus afforded a much larger area of endodermic surface to perform the gastral and other functions of the polyp .
The change gave increased 1Jital activity assoc\#183 ; iated with economy of space .
6 .
Introduction of a definite " wall " at some part of the calyx .
In Cyathophyllidre the wall was formed in the calyx , either near the centre at the limit of the dissepimentR.l outer zone with the central tabulate area , or near the periphery .
In either case it was formed by septal thickening , and is called a " pseudotheca .
" In Zaphrelltidre the wall was formed at the periphery as a modification of the innermost part of the epitheca .
The wall , in this case , was a structure independent of any septal thickening ( the latter might or might not be present ) ; it shows , when studied microscopically , lines of calcification set at right angles or obliqueiy to these in the septa , and is called a " true theca , " or simply " theca .
" In a great number of simple corals , or branches of branching colonies , no wall was introduced .
7 .
Retrogression of the primitive basal and peripheral " epitheca " in many genera .
This change largely depended on the introduction of a wall within the calyx in Cyathophyllidre .
Amongst the descendants of the Zaphrentidre , the epitheca is still well developed in most of the simple corals , is retrograde in the colonial types .
I append the following plan of recent families descended from these two chief Palreozoic mother-stocks of Madreporarian corals .
a , mmidae Proto -Zaphrentid l 2 .
Influence of Geographical Changes on the Palceozoic Polyps .
Every geologist knows that in Central and Northern Europe the period between the deposition of Carboniferous limestones and the advance of the Rhretic and Jurassic Ocean was a long geologic age of disturbance and for the most part upheaval .
The Permian and Triassic periods saw Europe undergoing the fluctuations of Archipelago conditions .
Shallow , muddy basins , broad deHas , salt lakes , and low coasts occupied the North-Western areas .
Central Europe saw in Middle-Triassic time one longer period of marine interruption , represented by the deposit of Muschelkalk .
Even in this limestone deposit corals are very rarely found , but the intervening ages of Permian sandstones and conglomerates , as well as the Bunter pebble and sandstone beds , are naturally devoid of coral remains .
Tho Madreporarian fauna of Central and Northern Europe found itself subjected in Late-Carboniferous and Early-Permian time to highly unfavourable conditions .
Pressed in by land to the West and North-West , an eastwards or southwards wandering was the only possible escape , and we may presume it was effected by the hardiest genera .
SYSTEMA'l'IC STUDY OF THE .MADREPORARIAN TYPES OF CORALS .
297 The Russo-Asiatic Ocean became the general harbourage for the descendants of the European Madreporaria , until in Middle-Triassic time bays and fiords in the Muschelkalk sea of Central Europe again offered the necessary conditions for coral existence .
At the same time a series of physico-geographical changes were taking place in the Eastern Alps , accompanied by submarine volcanic activity along the Southern zone .
And we find that local coral faunas abounded in the form of numberless reefs and banks .
The Central Europe to which the Mid-Triassic Madreporaria returned , showed very different contours from those of the Early-Carboniferous era .
Climatic conditions and the physical environment were still variable , and as those are known to have marked effect on the particular character assumed by a coral fauna , we need not feel surprise when we find that the Triassic representatives in Europe were limited to reef-faunas , bearing strong local impress in the different regions .
It would therefore be who1ly unfair to draw general conclusions from a comparison between the freely-developed pelagic and reef-faunas of the Palrnozoic era in Europe and the limited reef-faunas of a few localities in Mid-Triassic time .
It is otherwise , however , if we compare Palrnozoic with Jurassic Madreporaria .
The assemblage of coral forms in the Jurassic ocean is quite as representative as that in Mid-Palreozoic seas .
At the same time , the interval between the two geologic eras is immense ; and the vicissitudes which the corals must have undergone , in their constant wanderings to and fro , cannot have been without their effect .
For I am far from the opinion that it was a typical " Hexacorallian " fauna which the Jurassic sea brought into Europe , or a suddenly changed one .
I have in the foregoing chapters given too many sound proofs of the agreement in structure and homologies of parts which may be traced throughout the main families in both eras , and of the large number of transit.ional types .
It seems to me that the relatively frequent and widespread changes in the geography of the Carboniferous , Permian , and Triassic eras more than sufficiently account for the ' ' evolutionary motive " which I have above advanced in explanation of the changes in Madreporaria .
Many analogies might be found in other marine creatures .
The ins6nct of preservation of type is as undeniable as the instinct of self-preservation .
And surely the difficulties in the support of existence can never have been greater or more protracted for Madreporaria than those which drove them out of our regions in Fermo-Triassic time .
The effort to overcome those difficulties had the result of higher differentiation of the mesenteries and alteration of form in all the hardiest , most numerously represented , Palreozoic families .
One leading cef/ ltse-physical changes of .climate and environment , inducing a fierce1\#183 ; struggle for existence ; one continiwus counterbalancing eff ort-self-preservation and preservation of type-worked out a correlated series of changes in the animal kingdom of Madrepo1\#183 ; aria during the long interval of time between the JJ1idPalceozoic and Late-1.Wesozoic genlogic eras .
298 DR. MARIA M. OGILVIE ON 'l'HE MICROSCOPIC AND CHAPTER , X.--TRANSITIONAL NATURE Oli ' 'l'RIASSIC AND JURASSIC GENERA .
Seeing that many of the homologies which I have indicated between Palreozoic and Recent Mac.lreporaria are , so far as I am aware , expressed here for the first , time , I think it desirable to treat c~rtain intermediate types rather more in detail .
This is moreover necessary , since the general conclusion to which I have arrived differs very considerably from the opinions of the only author* who has up to this time made a special study of Triassic corals .
The interest of the Triassic Alpine fauna is many sided ; at present I confine myself to the evolutionary features .
Professor FRECH wrote , in 1885 , at the close of his paper on the Upper Devonian Corals of Germany ( Zoe , cit. , p. 942 ) : " On zoological grounds we can scarcely refuse to accept that the Tetracoralla were the ancestors of the Hexacoralla .
The geological occurrence speaks also for this hypothesis .
The change( " Umpragung " ) seems to have taken place during the period of deposition of the Bunter Sandstones , since the few Permian forms belong still without doubt to the Tetracoralla , whilst the known genera from the Muschelkalk and Alpine Trias are connected with the living .
" .
The standpoint taken here rests on a general hypothesis , and assumes a change of : ; t_ sweeping nature to have taken place during one geologic era ; no corals in that era being forthcoming in evidence of the change .
The same standpoint was retained by Professor FRECH after detailed examination of the Triassic fauna of various Alpine localities .
We read again ( ' Palreontographica , ' Zoe .
cit. , vol. 37 , p. 113 ) : " The great change of the Anthozoan fauna had been carried out before the Muschelkalk and after the Permo-Carboniferous .
" My own examination of the same sp~cimens enable me to criticize the grounds on which Professor FRECH 's statement is based .
According to that author , the coral fauna includes :I .
Maclreporaria .
1 .
Fam .
Astrceidce .
Gen. ThecosmiZia , lsastrcea , Phylloccenia , StyZina , Astroccenia , Stephanoccenia , MontlivaZtia .
Sub-fam .
StyZophyllince , FRECH .
Gen. StyZophyllopsis , Stylophyllum .
2 .
Fam .
Thamnastrceidce .
Sub-fam .
Thamnastrceince , FRECH ( vide antea , p. 189 ) .
Gen. Thamnastrcea .
Sub-fam .
Astrceomorphince , FRECH .
Gen. P 1rocycl0Zites , Astrceomorpha .
3 .
Fam .
Spongiomorphidce , FRECH .
Gen. HeptastyZis , Spongiomorpha , Stromatornorpha .
* F1uwn , " Die Korallenfauna der Trias , " 'Palreontograpbica , ' vol. 37 , 1890 .
SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
299 II .
Pterocorall\#163 ; a , FRECH ( = Rugosa , M. Eow .
= Tetracorallct , HJECK .
) .
1 .
Fam .
Cyathaxonidm , E.H. Sub-fam .
Gigantostylinm , FRECH .
Gen. Gigantostylis .
2 .
Fam .
Zaphrnntidm .
Gen. Pinacop !
tyllum , Coccophyllwm .
III .
Alcyonaria .
IV .
Tabidata .
( The latter sub-divisions I leave out of consideratioi1 , as my paper concerns the first two only .
) It is clear from the names given to the subdivisions I. and II .
that Professor FRECH draws a sharp limit between certain families and genera which he exclusively regards as Mndreporaria and others which he regards as " the last remnants ( ' Superstiten ' ) of the Palreozoic Pterocorallia .
" The impossibility of defining such a limit is proved , in my opinion , by the presence even of a single genus such as Coccophylluni .
The overlapping of the Palreozoic families upon Mesozoic families has been admitted by several authorities .
Take , for example , Professor NICHOLSON'S treatment of the Turbinol-iclm ( ' Manual , ' loc. cit. , p. 266 ) :- ' ' The Recent genera Guynia and I-Ictplopht1llia , again , are remarkable in the fact that the symmetry of the corallun'l .
is tetramera1 , and they thus serve to lead us to such ancient types as the Cyathaxonidre of the Carboniferous rocks ... .
The Silitrian corals which have been included in the genus Cyathaxonia belong , however , to the genus Linclstrcemia , which may also be referable to this division ( Turbinolidre ) of the Madreporaria , though it will be here provisionally placed among the Rugosa .
" The Silurian genus Duncanella , as well as the genera Petraia ( Silurian to Carboniferous ) and Polyccelict ( Silur .
to Permian ) are farther treated by Professor NICHOLSON as probably members of the Turbinolidre ( cf also antea , pp. 280-286 ) .
I need not , therefore , go farther in showing how prejudiced a picture Professor FRECH 's sub-divisions are apt to give the student .
More general inferences with regard to the systematic significance of the same Triassic genera may be made on the basis of research in the present paper .
I shall apply the nomenclature of families and sub-families proposed at the end of this paper , and re-arrange the Triassic genera in FRECH 's list according to what I regard as their phylogenetic significance .
800 DL{ .
MARIA M. OGILVIE ON THE MICROSCOPIC ANO MA .
DREPORARIA .
A.-Oj Cyathophylloicl Ancestry .
Fam .
Astrwidm Sub-fam .
Astrwinw .
Gen. Thecosmilia ... ... Montlivaltia .
( As above , excluding Stylina , Stylophyllopsis , Stylophyllum .
) Fam .
Fungidw .
Sub-fam .
Thamnastrroinm .
Gen. Thamnast1\#183 ; wa , LE SAUV .
AGE , em. PRATZ .
Sub-fam .
Fun_qinw .
Gen. Procyclolites , FRECH .
17ie Sub-family Astrmomorphinw , FRECH ( = Pseudoagaricime , PRA'l'Z ) , 1s fully discussed below .
Fam .
Eupsammidw .
pro { Gen. Stylophyllopsis , FRECH .
parte Gen. Stylophylluni , REUSS , em. FRECH .
Fam .
Turbinoliclro .
B.-Of Zaplwentoid Ance : : ; try .
Sub-fam .
1'iirbinolinw .
Gen. Gigantostylis , FRECH .
.Fam Arnphias1.rwidw ( seep .
328 ) .
Gen. Pinacophyllum and Coccophyllum .
Two species represent the genus Pinacophyllurn , FRECH ; I would refer the typical species P. parallel11 , m , FRECH , undoubtedly to the Amphiastneidre , a family numerously represented in Jurassic strata .
The second species is referred by FRECH , with mark of interrogation , to the genus Pinacophyllurn .
It had been determined by REUSS as Fletcheria .
FRECH referred both species originally , in 1889 , to the genus Amplexiis , removed them in 1890 to the new genus Pinacophyllurn .
Fam .
Styliniclw .
Gen. Stylina .
O.-The Family SpongiomoqJhidw , FRECH , is exclusively Triassic .
FRECH has compared it with the Stromatoporoids on the one hand , and the Turbinaridre on the other .
Its near affinity to the genus Astrwomorpha is also pointed out by Professor FRECH .
Probably the Spongiomorphi.ds are most closely allied with the Poritidre ; they might almost rank as a sub-family .
SYS'rE : MA.TIO STUDY OF THE :MADREPORA .
RIA .
N TYPES OF CORA .
LS .
301 In the above list I have endeavoured to indicate the main lines of systematic treatment relative to pre-Triassic and post-Triassic forms .
One of the most striking features is the great richness of the Triassic fauna in members of the Eupsammidre , .Astrreidre , Fungidre , and its poverty in those of the Amphiastrreidre , Turbinolidte , and Stylinidre .
This observation goes far to con6rm some of my argnments in the preceding chapter .
The " precocity " of the change from bilateral to radial symmetry in the Palreozoic Cyathophylloids finds expression in the rich variety of form attained by their descendants as early as Triassic time .
On the other hand , the tardiness of the change in the Zaphrentoids is consistent with the fact that the Turbinolidre had not attained a steady type even in Jurassic time , and their full development was postponed until the Middle-Cretaceous age and its favourable oceanic conditions .
The general resemblance between the recent representatives of Astrreidre and Amphiastrreidre ( cf diagram , p. 331 ) which caused them to be placed in one family ( Astrreidre ) , becomes all the more readily understood when the near relationship of their early ancestors is held in view .
I have already referred to Professor NICHOLSON'S indicati011s of genera intermediate between the three sub-divisions which he made ( 1 Manual , ' p. 284 ) of Palreozoic Madreporaria : Cyathophylloidea , Zaphrentoidrn , Cystiphylloidea .
Professor FRECH ( loc. cit. , p. 83 ) draws attention to the same feature : " It has been frequently pointed out of recent years ( amongst others , also by NEUMAYR ) , that the various classificatory features ( ' verwandten Merkmale ' ) are not sufficiently significant for the separation of the Cyathophylloids and Zaphrentoids .
" Those ar~ side-lights which give clearness not to one family or one era of Madreporaria , but to the whole group from the earliest time to our own .
The Triassic Sub-family Stylophyllince , FRECH .
The interest of this sub-family , erected by Professor FRECH , rests in its decided Palreozoic character .
The habit of the corallum is typically simple ; when colonies form , they do so by " endothecal " or " tabulate " budding , and look .
like colonies of Cyathophyllum .
The septa are made up of short , thick spines ; often the spines are irregularly set and never properly combine as a septum .
The base of the polyp is either tabulate or dissepimental ; the theca thick and rugose .
Bilateral symmetry is frequently shown ; radial symmetry is chiefly present in larger well-developed specimens and then is never cyclical , the septa being indifferently long and short .
In spite of all those features of identity with Palreozoic forms , Professor FRECH has given reasons , at considerable length , for his opinion that the genera of this subfamily are probably sprung from Montlivaltia .
Had the genera occurred in Palreozoic strata , I feel almost sure a place would , by most authors , have been found for them possibly between CyathophylJoids and Cystiphylloids , and I think , therefore , I am justified in presenting another view of the case to that already published ( ' Triaskorallen , ' loc. cit. , pp. 42-47 ' ) .
Professor FRECH 's diagnosis of the sub-family may be given in full : " The two ( or three- ) ' ) genera " of the Stylophyllinre " form a peculiar side-branch of the Astrreids , which is almost limited to Trias , and has not until now been more sharply defined .
The forms belonging to it have , as a distinguishing feature , strong septal spines remaining in part or wholly free ( as in the Thamnastrreidre ) .
The septal spines are , however , not built up by nodules .
Moreo\--er , synapticulre and horizontal bars ( pseudosynapticulre ) are absent .
The dissepiments are strong and are developed sometimes as horizontal lamellre or tabulre .
" The discussion " On the apparent relationship of Stylophyllum with Palreozoic corals ( Calostylis and Cystiphyllnm ) " leads to the statement at the end that , " a group ( Stylophyllinre ) of Triassic and Liassic corals , belonging with certainty to the Astrreidre , is , in many points about their inner structure , reminiscent of old Pa.lreozoic forms like Cystiphylluni and Calostylis .
" This reminiscence is said to be comprehensible in any one of three ways :(1 .
) " Stylophyllum descends from one of the Palreozoic genera .
" ( 2 .
) " Stylophyllum is to be looked upon as an atavistic form .
" ( 3 .
) " It is a question of one of the frequently occurring cases among fossil corals , that 'hard parts very much like one another occur in very different sub-divisions ' ( NEUMAYR ) .
" The last of the three explanations is the one which Professor FRECH accepts , his conclusion being : " According to the morphological investigations , it may be presumed that Stylophyllopsis branched off from Montlivalt\#183 ; i'.a .
" Setting aside , for the present , the above statement of FRECH 's opinions , I shall enumerate a few of the observed facts .
The species Stylophyllum paradoxum , FRECH , is one under which an unusual variety of forms bas been included .
In fact , one of the most striking features in almost any species of Stylophyllum is that few specimens can be said to be really alike ; and that , although the differences may be slight from one specimen to the next , there are extremes at the two ends , so to speak , of a specific series , which seem to have little in common .
This I take to be thoroughly natural , if .
we regard the sub-family as a Triassic assemblage of nearlyrelated species undergoing differentiation in one di1\#183 ; ection and anothe1\#183 ; , and fairly rapidly .
Advance and retrogression would act hand in hand in such an assemblage .
A specimen of S. paradoxum , FRECH , figured by him in Plate 14 , figs. 25 , 25A , may be t aken as an average in the matter of calycinal size and character of the septa .
There the septal spines are seen to Le scarcely arranged in regular longitudinal series , but rather each growth-period was practically marked by an .
independent set of small " spiniform " invaginations of the aboral body-walls .
With this the * The two my n gen era are Stylophyllum and S tylophyllop sis .
A thi1\#183 ; d genus of latimeandroid habit ; is called Meandrostylis , and treat ed as a sub-genus of Stylophyllurn .
SYSTEMATIC STUDY OF THE M:ADREPORARIAN TYPES OF CORALS .
303 presence of broad tabulre or large dissepiments is correlated .
This specimen was originally described by ScHAFHAUTL ( ' Neues Jahrb .
d. Mineral .
, ' 1851 , p. 413 ) , as Cyathophyllum ceratites ( non Gold\#163 ; ) , and one can scarcely feel surprise that the generic title should be a Pa]reozoic one .
Still more Palreozoic in appearance are the very small specimens ( Zoe .
cit. , Plate 14 , fig. 9 ) , in which few spines are present , and these are grouped in threes and fours here and there .
The species Stylophyllum pygm(Eum , FRECH ( Zoe .
dt .
, p. 56 ) , shows remarkable .
agreement in habit and genera ] features , to judge from the illustrations on p. 56 , with the Jurassic form " Lithodendrom mitrat\#183 ; um , " QuENST .
( cf. ' N atth .
Kor .
Pa]reontographica , ' vol. 21 , 187 5-76 , Plate 51 , fig. 9 ) .
I would be inclined to place this species in the family of Amphiastrreidre , in the neighbourhood of the genus Coccophyllum .
On the other hand , the large specimens ( Plate 14 , fig. 12 ) bear , what might be called in contrast , a Mesozoic impress , i.e. , the spines are definitely arranged in radial lines , and are sufficiently compacted to form septa extending from the rim of the calyx to the centre .
At the same time , the " endotheca " varies , being sometimes more dissepimental , at other times more tabulate in character .
A Triassic species like this Stylophyllum paradox.:um or pygm(Eurn , showing such variety and with an outspoken Pa1reozoic appearance , is , in my estimation , most simply explained as an evidence of structural transition .
Further , Professor FRECH is incorrect in stating that the sub-family which he erected for StyZophyllum and Stylophyllopsis dies out in Liassic time ( Zoe .
cit. , p. 47 ) .
The Jurassic " Nattheim " coral fauna* presents a genus certainly of more :fixed character than that accorded to those of the Stylophyllinre , but including species of almost identical characters with species of the said Triassic and Liassic genera .
I refer to the genus Epistreptophyllwrn , MILASCH .
, which was figured and described fourteen years before Professor FRECH 's work on " Triaskorallen .
" Having compared the N attheim originals with the Triassic originals of Professor FRECH , and also with certain species occurring in the " Stramberg " Jurassic fauna , I am in a position to state that in some cases it is impossible to find generic distinction between the Triassic and Upper Jurassic forms .
Therefore this " peculiar side-branch of the Astrreids " was not " almost limited to Trias , " as Professor FRECH states .
The Triassic specimens of StyZophyllum and Stylophyllopsis which come into question in this comparison with the Upper Jurassic genus Epistreptophyllum are always the largest and best developed in any particular species , vide Stylophyllv..m paradoxum , FRECH ( Plate 14 , fig. 12 ) ; Stylophyllum tenuispinum , FRECH ( Plate 14 , figs. 28 , 28A ) ; StyZophyllopsis poZyactis , FRECH ( Plate 15 , fig. 17 ) ; StyZophyllopsis Lindstromi , FRECH ( Plate 12 , fig. 2 ) .
Nor is Epistreptophyllum an isolated Jurassic genus-some poor remnant of a long * BE0KER and MILASCHEWITZ , " Die Korallen der Nattheimcr Schichtcn , " 'Palroontographica , ' vol. 21 , 1875-76 , p. 210 .
past !
On the contrary , it is one of several genera characterized by the fact that the " spines " ( growth-segments ) of the septa are more or leRs unattached both in a longitudinal an\lt ; l in a horizontal direction .
A bout the middle of the septum the tendency is for the spines to be compacted , but centrally and often peripherally the tendency is for them to be freer .
Lateral coalescence of the septal surfaces occurs frequently .
Besides , these genera show frequent signs of bilaterality .
Such genera .
are Dermosmilia , KoBY ; * Baryphyllia , KoBY ; J--Iaplarcea , M1LASCHEWITZ .
A moment 's consideration will suffice to show that this Triassic and Jurassic series of genera leads .to the family of Eupsammidre , destined to play a very important role in the next following Cretaceous and later time .
The Table of Classification ( antea , Part I. ) points out that the unfilled apertures left in the septa of the Eupsammidre , and very often also in the Lophoserinre , are due to incoherence of the " septal spines " referred to in my table on p. 252 as the " successive growth-parts " of septal trabeculre .
The individuality of the " septal spines " is one of the ancestral features in the group of Madreporaria ( cf antea , pp. 253-261 ) , and it is a feature which has in the course of time become more and more seldom .
Amongst Recent types it is found in very few families .
It is , however , only natural that in the Triassic age , a period rich in atavistic and transitional types of Madreporaria , there should be such septa more or less loosely built up of septal spines in genera possibly remote from one another .
The systematic position of the genus must be determined in accordance , as far as possible , with other features , such as the relations of septum , wall , and basal structures , the main direction of the spines and their method of coalescence if any .
Yet Professor FRECH emphasizes this featitre as the distinguishing feahtre of the sub-family Stylophyllince .
The other features which he gives--absence of synapticulre and variable development of dissepiments , sometimes iu vesicular , sometimes in tabulate form-are not in themselves of sufficient weight for the erection of a new sub-family .
The result is that a number of specimens have been grouped together , and form an unnatural sub-family , said to be of exclusively Triassic and Liassic age .
Amongst these specimens one here an\lt ; l there may be picked out and shown to have quite different systematic affinities from its neighbours .
If I were to accept this sub-family as Professor FRECH bas described and classified it , I should , in my opinion , overlook some of the important Triassic connecting links between Palceozoic and Jurassic genera .
The majority of the specimens included in this sub-family are ancestral forms of Eupsammidre , others belong to the Arnphiastrreidre , and still othns show no essential difference from Pa1reozoic forms of Cyathophyllidre and CyRtiphy1lidre .
I find it impossible to agree with Professor FRECH in his treatment of all the specimens as a narrowly-limited group for .
themselves which had branched off from the Astrreidre !
* Ko HY , " Polypiers J urassiqucs do la Suisse , " Gc.nova , 1880-89 , 'Mem .
do la Soc. Pal .
Suisse , ' vols .
7-16 , pp. 194 , 546 .
SYSTEMATIC STUDY OF THJi .
J MADREPORA .
RIAN TYPFJS OF CORALS .
305 The so-called Stylophyllinre , FRECH , may be descended from side-branches of Cystiphylloids , Cyathophylloids , and Zaphrentoids , but only a certain number of the specimens represent a fixed , definite form differing from ancestral types .
Wherever such a form exists among the so-called Stylophyllinre , it can readily be compared with certain Jurassic genera .
The systematic position of the Jurassic genera in question is in the Eupsammidre , yet of a character very nearly approaching many early types of Fungidre .
The Triassic Sub-family Astrceomorphince , FRECH ( = Pseudo-Agaricince , P.RATz).* \#183 ; The Fungidre may be looked upon as struggling towards definite form and structure in Triassic and Jurassic time .
The Thamnastrreinre attained full development rapidly .
In Triassic time only one genus is described by FRECH-Thamnastrcea .
At the same time , in the Triassic fauna of St. Cassian , several others occur with the septal structure characteristic of the Microsolena members of the Thamnastrreinre in Upper Jurassic time .
The Triassic genus Astrceomorpha was looked upon by PRATZ as an ally of the Thamnastrreinre ( = Pseudoastrreinre , PRATZ , pro pa1\#183 ; te ) , and was grouped by him more especially with the genus SideJ'astrcea .
I have discussed the relations and characters of this genus in my monograph of the Stramberg fauna , in connection with several Upper Jurassic genera .
It is , however , necessary here to point out that the Triassic genus Astrceomorpha cannot be associated in one and the same subfamily with the Tertiary and Recent genus Siderast1\#183 ; cea .
The " Siderastrcea " sections given in Part I. , pp. 178-181 , of this paper show the essential agreement in trabecular structure and in the calycinal features between Siderastrcea and the various members of the Fungidre .
At the same time , a comparison of the same sections with those given by PRATZ for Astrceomorpha show important differences .
PRATZ was not quite right in his description of the septum of Astrceomorpha : " Ein jedes Septum macht vielmehr den Eindruck eines selbstaudigen unregelmassig gestalteten Trabeke]s .
" Some well-preserved specimens in the Munich Museum show on weathered septal surfaces that the septum is in reality composed of a number of upwardly-directed trabeculre , which are placed almost perpendicularly to the upper edge of the septa .
The trabecu]re are extremely fine , and are disguised b_y the strongly-built horizontal ridges which project into the interseptal loculus .
The ridges are often developed as knotted thickenings united by thinner parts .
So that , in tangential sections through adjacent septa , the general impression given is strikingly similar to that of recent Poritidre .
FRECH has pointed\#183 ; out the probability of relationship between the genns AstrCBo , morpha and the curious Triassic family Spongiomorphidre , and the resemblance between the latter family and " Perforates , " more especially the Cretaceous genus * Of .
antea , pp. 182-187 .
\#165 ; PCCCXCVI , -B , 2 ~ DR. MARIA ii .
OGILVIE ON THE MICROSCOPIC AND ( Actinacis ) . . .
I am inclined to think the recent genus P01\#183 ; ites is the \#183 ; genus amongst the " Perforates " which is most closely assodated with these Spongiomorphid ').'riassic antetypes .
Turbinaria seems to me to have in its irregu]arly-built ccen- : chime a character which renders it distinct from Porites on the one hand , and fro)11 Astr03omorpha and the Spongiomorphidre on_ the other .
The general habit of Astr03omorpha , its closely-set calyces , indistinct wall , net-like structure of all the skeletal structures , the small number of septa , styliform columella , richly-distributed dissepiments are a combination of features found , so -far as I know , among recent , genera only in the Poritidre .
But they are features which at once distinguish the genera in question from the Fungidre .
And\#183 ; I would not agree with NEUMAYR and FRECH in saying that " Microsolena ( a Thamnastrreid ) builds a bridge to the Poritidre " ( antea , chap .
v ) .
On the contrary the other indication given by FRECH of a possible connection between .
certain Perforates , Astr03omorpha , and the Spongiomorphidre , seems to me preferable .
I give a microscopic section of Heptastylis ( fig. 75 ) for comparison of its Fig. 75 .
Heptastylis .
Longitudinal section , showing the erect trabeculre and horizontal synapticulre .
Fascicles of fibres are seen here and there in the section .
structure with P01\#183 ; ites .
It is .
one of the Spongiomorphids which shows distinct calyces , and which can with ample justice be compared with younger types among the Poritidre .
One fact still deserves to be mentioned about Astr03omorpha , since I have not yet seen it in published descriptions .
It is that in many calyces there are traces of bilateral symmetry in the arrangement ?
f the septa .
Sometimes a main septum i$ \#183 ; SYSTEM .
A.TIC STUDY OF THE M.A.OREPORARi .
A.N TYPES OF CORALS .
307 at once distinguishable .
Again , other calyces in the same colony are as distinctly radial .
in their symmetry .
The above considerations lead me to entirely split up the sub-family Pseudoagaricinre , PRATZ = Astrreomorphinre , FRECH .
The Recent genus SiderastrC13a , placed in it by PRATZ , is a Fungid .
The Triassic genus Procyclolites , placed in it by FRECH , has , I believe , a much greater resemblance to simple Thamnastrreinre and Funginre than to AstrC13omorpha .
The latter genus may alone be left as a certain representative of the sub-family Astrreomorphinre , FRECH .
This sub-family is probably best included under Poritidre , at any rate , I shall provisionally place it here .
The Triassic Family 8pongiomorphidC13 , FRECH .
I would be much more inclined to look upon the Spongiomorphidre than on the StylophyHinre as a peculiar Triassic group whirh soon became extinct .
The resemblance to the Stromatoporoids is pointed out by FRECH , and is the most that can be said of the possible ancestry of the Triassic fi.tmily .
FRECH sub-divides the Stromatomorphidre into cyclic and acyclic forms , according to the arrangement of the trabeculre in definite calyces , or the absence of any such arrangement .
The Genera Astroccenia and Stephanoccenia .
These two genera appeared for the first time in the Triassic period , had a very wide distribution in J U:rassic time , and , since then , have been on the wane .
It i8 doubtful if they are represented in recent seas .
HELLER* described a coral from the Adriatic Sea , which he called Astrocronia pharensis , but HEIDER , t after a detailed examination of the coral , referred it t\lt ; ?
the genus Madracis , one of the Oculinidre .
The Triassic species are distinctly in a state of transition , and have still much in common with Palreozoic corals .
Their septa are extremely short , and sharply_ dis- ' tinguished from the large central area representing the tabulate area in ancestral types .
In this area the spines are arranged without a.ny definite radial or vertical system , repeating the structure seen in Cystiphylloids , or , in smaller measure , in the central p\#183 ; art of Endophyllum .
In Jurassic species the central ( columellar ) area is apparently diminished , \#183 ; the pali of Stephctnocania having a definite order relative to the septa , and surrounding a central columellar palus . .
It may be said , moreover , that no two of the Triassic species have t_he same number 9\#163 ; septa in the primary cycle ; \#183 ; Stephanocronia alpina _has septa in two .
cycles of \#183 ; 10 ; \#183 ; s. juvavica has a \#183 ; verf irregular arrangemen\#183 ; t , scarcely cyclical ; the number varies from 24 to 30 ; . . .
HELLER , " Zoophyten und Echinodermen des .A.driat .
Meeres , " 1868 , p. 27 .
t HEIDER , " Korallenstudien II .
, Madracis pharensis , HELLER , sp. " '.A.rbeiten aus dem Zool. Inst. zu Graz , ' vol. 4 , No. 2 , 1891 .
\#183 ; 308 DR. MA .
RIA M. OGILVIE ON THE MICROSCOPIC AND S. Schafhciutli is quite regular , with 3 cycles ( 6 + 6 + 12 ) ; Astroccenia Waltheri has two cycles of 10 ; A. Ohmanni two of 8 ; and A. hexactis two of 6 .
The var.iations in the number of septa are seen , therefore , to br~ like those so characteristic of Stylina and its allies .
Probably this is one of the reasons why Astroccenia and Stephanoccenia were for so long placed systematically un9-er the Eusmilinre , E. H. , in the neighbourhood of Stylina .
They were then removed to the Astrreinre on account of their toothed septal edges .
A prominently-toothed septal edge rarely occurs in any other family than AstrreiJre and Fungidre , but it has now become familiar that very often the septa of the main cyclei : ; may have toothed edges , and these of the subordinate cyc1es in the same calyx have edges almost smooth .
The microscopic structure is , however , the same in both cases , and is the feature which must be considered in determining the systematic position of a genus .
Sections which I have made of Jurassic specimens of Astroccenia and Stephanoccenia show ~greement in the septal microscopic structure with the ccenenchymate type , Stylophora ( ante , pp. 83 , 94 ) , in which the septa are neither deeply toothed nor prominently granulate .
I am inclined to look upon both genera as closely allied with the family of Pocilloporidre .
Not only the septal structure , but also the general habit of the corallum , the well-developed epitheca , the irregular secondary infilling in the angles between adjacent calyces are features presenting strong similarity .
FRECH has given longitudinal sections , illustrating the typical characters of the Triassic species of Astroccenia and Stephanoccenia ( Zoe .
cit. , p. 33 , figs. A and B ) .
He draws attention to the difference between the loose , spinate columella in Astroccenia , which is connected with the septal ends and the distinct pali in Stephanoccenia .
Secondary compacting of the loose spines in .
the columellar area would produce naturally the effect of a longer septum , with irregular-sized teeth towards the centre .
I have observed such appearances in many Jurassic specime~s both of Astroccenia and Stephanoccenia .
This kind of septum I may call heterornorphous , where central spines once free from the septum become compacted with it so habitually as to be almost specific .
It was already foreshadowed in Palreozoic corals and is of common occurrence in Poritidre , Trochocyathinre , and in some recent Astrreidre .
One important difference may be observed bet.ween the typical Jurassic species of Astroccenia and Stephanoccenia and the representatives in Triassic age .
It is that in the latter tabulre occur centrally , while in the Jurassic species there are no tabulre .
A second interesting point is the entire uncertainty in the number and arrangement of the septa in the Triassic species , eight occurring now and then as the number in the primary cycle .
While Jurassic species show regular cyclical arrangement of septa , and usually in hexameral or dodecameral type .
Systematic Significance of Triassic Corals .
Sum1narizing the above observations which I have myself made on Triassic and SYSTEMATIC STUDY OF THl !
l :M:ADREPORARIAN TYPES OF CORALS .
309 Jurassic corals , I find the utmost corroboration of the slowness of the so-called " change " in the aspect of the Palreozoic fauna .
To refer again to my list of Triassic corals given on p. 300 , for comparison with Professor FRECH 's list , p. 298 , the result , as I draw it , is as follows : Division C is almost exclusively of Pcilmvzoic character , very few species finding direct comparison with post-Triassic forms .
Divisfon B includes genera which ( with only one exception , Stylina ) were actually placed by Professor FRECH under Palmozoic families .
They indicate the strong Palreozoic impress still retained by numerous Triaesic species .
Division A includes( ! .
) Stylophyllinm , FRECH , a sub-family newly erected by Professor FRECH , showing , however , remarkable evidence of having been in active process of change , associated on the one side with Palreozoic forms , on the other identical with the early types of Eupsammidre .
( 2 .
) Procyclolites , which ought to have a quite different systematic position from Astrmomorpha , with which it had been classified by Professor FRECH .
Its microscopic structure resembles Palreozoic carinate Cyathophylloids , and it is possibly best looked upon as a curious antetype of the synapticulate Funginre , resembling certain Cretaceous types even more than the known Jurassic relatives .
( 3 ) .
Thamnastrminm represented by the typical genus Thamnastrma of distinct Mesozoic character , and introducing the great Jurassic concourse of this extinct subfamily of the Fungidre .
( 4 .
) AstrCl3idm .
This large family , although so profusely represented in the Triassic fauna , is of distinct post-Palreozoic character .
Indeed , its prepunderance in individuals , together with the striking radiate character , have done much to encourage the idea of a palreontological break between the Palreozoic and Triassic-Recent coral faunas .
Yet the curious fact is that everyone admits , in a general way , the descent of this family from the Palreozoic Cyathophylloids .
The latter may be said then to have most completely undergone a change in form during the interval that elapsed between Carboniferous and Triassic time .
At the same time , true Cyathophyllids are present in the St. Cassian Triassic strata , * which is a somewhat older deposit than the Zlambach strata above discussed .
According to those results , I find it utterly impossible to accept Professor FRECH 's dictum that " The great change of the Anthozoa had been carried ou , t before the Muschelkalk and after the Perrno-Carboniferous period .
" I have indicated , in direct contradiction of this , the connection aiforded by the Jurassic corals with the main groups of Triassic corals , and have brought forward evidence of a continued series of * WILHELM VOLZ , " Vorliiufige :M:ittheilung , " " Ueber die Korallenfauna der St. Oassianer Scbichtcn , " 'Jahresb .
Schles .
Ges .
Vat .
Cultur , Naturw .
Sect. , ' June , 1894 .
There , under the sub-division Bugosa ( in contradistinction to Madreporaria ) , the families Oyathophyllidw ( gen .
Ocelocmnia ) and Zaphrentidw ( gen .
Pinacophyllum ) are represented-each by a single genus .
310 DR. MAR !
A. M. OGILVIE ON 'l'HE MICROSCOPIC AND changes going on during the Triassic and Jurassic .
ages , changes unmistakably transitional between the Palreozoic and Recent Madreporarian fauna .
CHAPTER XI .
--SUB-ORDERS OF MADREPORARIA .
EDw ARDS AND HAnrn 's SuB-ORDERS .
EDWARDS and HAIME grouped the types and families which have been treated in this paper , under the sub-orders MadreporarZ:a Ru:gosa , Aporosa , Perfo-rata .
My reasons for not accepting the sub-divisions M. aporosa and M. perforata , follow from the detailed study of skeletal feat , ures carried out in Part I. Reference may be made to chapter viii .
above .
The sub-order M. ru , gosa was , in the first instance , erected for Palreozoic .
corals ( see " Brit. Foss .
Corals , " E. and H. , " Introduction , ' ' p. 64 ) .
These are usually enveloped in a rugose epitheca , and were said to have a " septal apparatus , never :r forming six distinct systems as in all the preceding Zoantharia ( Apo , \#183 ; osa , Perfo1\#183 ; ata , Tabulata ) , but appearing to be derived from four primary elements .
" The terms ( 1 ) Tetracoralla , and ( 2 ) Hexacoralla , were afterwards given by HJECKEL in the same sense as EDWARDS and HAIME 's sub-orders , ( 1 ) M. rugosa , and ( 2 ) M. apo1\#183 ; osa + M. perforata , respectively .
The two foregoing chapters in the present paper sufficiently point out the .
conventionality of these two sub-orders .
The feature of " tetrameral " septal symmetry is one which was subject to slow phylogenetic change , taking place sometimes completely , sometimes less completely in genera descended from one and the same ::i , ncestral type , and in no measure limited to a particular period of geological time .
THE TERMS EuTHEC .
ALIA AND PsEUDOTHECALIA .
Relation of the Edge-Zone ( Randplatte ) to the Theca .
_ Of recent years HEIDER suggested the possibility of subdividing M adreporaria into Euthecalia and Pseudothecalia .
Zoologists have also beeri inclined to .
lay some weight on the presence or absence of an edge-zone ( " Randplatte " ) surrounding the main visceral chamber of the polyp .
HEIDER studied the " edge-zone " in Cladocora and Caryophyllia , and described it as a smooth lappet clothing the outer surface of the calicinal wall .
The crelenteron was typically continued into it , divided up into mesenterial loculi continuous with those of the main crelenteric cavity .
From what is already known about the edge-zone in recent types , it may be demonstrated to be a polypal feature correlated with any one of two possible sk~let.al_ features:- ( 1 .
) The presence of a pseudothecal thickening _ at some part of the calyx SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
3il ( .A:strreidre ) , or of a thickened synapticular development , i.e\#183 ; .
, a modified form of pf !
eudotheca ( Fungidre , Eupsammidre ) .
( 2 .
) The occasional retardation in growth of the theca relatively to the septa ( Turbinolidre , .Oculinidre ) .
This relationship observed between " edge-zone " and " skeleton , " in living genera , affords a standpoint from whi.ch I have studied fossil genera , and I arrived at certain general principles explaining the probable history of the edge-zone in pseudothecate types .
It seems to have been in the first place well developed in some of the colonial Oyathophyllid types , such as Acervularia , Phillipsastr(J3a , and probably had -primitively the function of a nurse-organ , allowing rapid development of buds .
This is very certainly the part which it played in many Jurassic genera ( ex. Amphiastr(J3a , J)iploc(J3nia , Placophyllia ) .
Frequently the outer area , corresponding to " edge-zone , " is beset with a whorl of buds .
Without doubt the " narrowing of a calyx , " so common i , n Palreozoic genera , is a similar morphological process and led , in many cases , to the formation of a temporary edge-zone . .
The genus Cyathophyllwrri ' , with its tabulate , well-marked , central area , and .broad , vesicular , outer zone , may be regarded as the antetype of most forms whose polyps have edge-zones , and this because the relations of septa , pseudotheca , and epitheca , are essentially unchanged in the simple Cyathophyllurn , and the simple Astrreid , Fungid , or Eupsammid .
Yet one cannot generalize from the skeletal features since , in certain types , a pseudotheca does not occur , and there is no marked difference in the basal depo~it , central , and peripheral areas .
Neither can one generalize very safely from the polypal parts , since the edge-zone may be well developed or rudimentary in apparently nearly-related genera .
For example , Dencll'ophylli\lt ; !
' is an Eupsamn1id in which the edge-zone has distinct mesenterial loculi ; whereas in Astroides , another Eupsammid , the edge-zone is a simple layer , without ccelenteric loculi .
HEIDER/ who carefully pointed out the difference in the character of the edge- .
zone , was inclined to ascribe great importance to it , and suggested its influence on the formation of the theca .
According to HEIDER , the types with a well-developed locular edge-zone were those in which a pseudotheca was present , while the types in which the body-wall was a mere plate of tissue were those in which an eutheca , or true theca , was developed .
I am not of opinion that any such distinction can be carried out .
Galaxea , with a true theca , has a , locula , r eJge-zone .
HEIDER suggested that all the Madreporaria might possibly be , on this basis , arranged in two main suborders : " Euthecalia " ( ex. Astroicles , and perhaps Flabellum ) , t * HEJDF .
R , 'Korallenstudien , ' Zoe .
cit. , 1886 , p. 178 .
t HEIDER 's definition of the two subdivisions is made on the basis that all skeletal structures are laid down by the mesoderm ( cf. pp. 9-11 ) .
He defines the Pse11dqthecalia as those in which the body .
wall , with its three layers of endoderm , mesoderm ( = mes9glcea , BOURNE ) , and ectoderm , remains \#183 ; unchanged at the peripheral portion of the calyx .
It is the same in polyp and edge-zone , and lays down no skeletal wall-lamella ( Mauerblatt ) .
The.\#183 ; septa are secondarily united by their lateral surfaces , and " Pseudothecalia " ( ex. Cladoco1\#183 ; a , Dendrophyllia , Rhodopsammia , Balanophyllia , Car,11ophyllia , etc. ) .
It is true that Caryophyllia appears in the published sections to have a pseudotheca , and to be in this respect more like Astrreidre than like the majority of Turbinolidre .
On the other hand , KocH 's description of the development of Caryophyllia mentions the appearance of wall or ring-folds between the septa .
HEIDER has thrown some doubt on these observations .
Yet Cm\#183 ; yophyllia is one of the best-known genera , and precisely this uncertainty shows how impracticable it would be at present to carry out a subdivision of Madreporaria into these two suborders .
Already one may safely say most Astrreidre which have a wall are " pseudothecal .
" But in the case of all other families there are species and genera which require careful examination before general results can be drawn up .
ORTMANN has suggested three orders ofMadreporaria , based on the thecal characters : ( 1 ) Euthecalia , ( 2 ) Pseudolhecalia , ( 3 ) Athecalia.* His observations were directed largely to the appearance of the centres of calcification in the skeletal structures .
He showed by transverse sections that the theca possessed tangentially-placed centres of calcification in the following :-Flabellum , Euphyllia , Galaxea , Desmophllum , Pocilluporct , Amphihelia , Lophohelia , Acrohelia .
Transverse sections of most of these genera may be seen also in the papers of FowLER and BOURNE , frequently referred to .
ORTMANN 's arrangement of families isOrder Euthecalia .
Sub-order Pocilloporacea , fam .
Pocilloporidre ; Pocillopora the type examined .
Sub-order Stylinacea , fa.m .
Stylinidre ; type examined , Galaxea .
Rub-order Eusmiliacere , fam .
Euphyllidre ; type examined , Euphyllia , , Under these sub-orders naturally more families are to be included , but I cannot find any complete list given by ORTMANN .
and this forms the " pseudotheca .
" In the Euthecalia the peripheral part of the aboral body-wall is said by HEIDER to lay down a calcareous wall-lamella within its mesoderm layer , an " eutheca , " with which the septa later unite .
Fowurn , in the course of anatomical studies , has frequently referred to the difference in the behaviour of the peripheral region of polyps with regard to an edge-zone .
He , however , always places Astroides in the same category as Dendrophyllia , possessing an edge-zone intimately associated with the polypal cavities .
This is unquestionably the correct view .
Still , even if HEIDER cites one example incorrectly , the other , viz. , Flabellum , gives very clear expression to his meaning .
The difference of the wall in Flabellum and in an Astrreid is a skeletal feature of the highest systematic importance , especially when the relation of the " eutheca " to septa in si1nple f01\#183 ; ms and to C03nenchyme in colonial .forms is prope1\#183 ; lie appreciated .
For example , the insertion of new costre alternately to the septa is a feature only present in an eutheca , never in pseudotheca .
Again , in ccenenchymatous forms new costre may be introduced in such numbers , or be so near each other in the wall of small calices ( ex. Porites ) , that little room remains for " wall-lamellre , " and the calyx is then often described as without a_ wall , immediately surrounded by coonenchyme .
At the same time the mere presence and relative position of the costre shows the homology of the extra-calicinal portions of the colony , i.e. , the coonenchyme with tho firmlybuilt tbeca , thecal costre , and epithecal thickening in simple forms .
* Athecalia ( ORTMA .
NN ) , ref .
antea , " Note on the Pqrous Theca , " P\#183 ; 249 , SYS'l'EMA'rIC STUDY OF THE MADREPORARTAN TYP .
ES OF CORALS .
313 The Turbinolidre and Oculinidre are also placed under the Euthecalia .
One name is unsatisfactory in the above list , viz. , Stylinidre .
The typical Mesozoic StylinidR have a pseudotheca , and not an eutheca .
Galaxect differs from them in this respect .
Otherwise , my results agree with those of ORTMANN , in so far as all the above families have thecce .
Several of the Turbinolidre and Trochosmiliacere , KH .
, are at present being examined in this respect in the Munich Museum .
Order Pseudothecalia .
The Astrreidre .
Order A thecalici .
In this order ORTMANN places all the remaining Madreporarian families , including such types as Fungict , Me1'Ulina , Lophoseris , Coscinarcea , Siderastrcea , Turbinaria , Montipora , Alveopora , Porites , Madrepora , Dendrophyllia , and traces them back phylogenetically to the Triassic Thamnastrreidre ( = Pseudoastrreinre , PRATZ ) .
My results disagree with ORTMANN in this respect .
If I were to accept main Orders based upon the characters of the Theca so far as yet known , I , vould arrange the recent Madreporian families as follows :Order Euthecalia .
Turbinolidre , Oculinidre , Amphiastrreidre ( type Euphyllia ) , Pocilloporidre , Madreporidre , Poritidre .
Order Pseudothecalia , Astrreidre ( excl .
Eusmilinre , E.H. ) , Fungidre ( incl .
Pseudoastrreirne , PRATZ ) , Lophoseridre , Eu psammi\lt ; lre .
The Order A thecalia , ORTMANN , is , in my op1mon , unnecessary , and has been based on the principle of the absence of , or porosity of : the wall .
ORTMANN 's diagnosis relative t.o the wall in Athecalia is " a true wall ( theca ) is absent ; the septa are united with one another by synapticulce .
These are either distributed upon the septal surfaces or combine so as to form a porous , loosely-made ccenenchyme or a porous false wall .
Septa of the neighbouring calyces confluent or breaking up into the ccenenchyme or the false wall , " Zoe .
cit. , p. 495 .
I have referred to the various points in this diagnosis in a previous chapter ( antea , pp. 200 and 261-268 ) .
It remains to say that the _ccenenchyme of Turbinarict and Madrepora is of an entirely irregular character .
Synapticular rods are combined with flat dissepimental plates , and are placed now vertically , now obliquely , now horizontally .
They cannot be compared morphologically with the regular trabeculate outgrowths in Fnngidre and Lophoseridre .
Besides , they are distinctly outside the calyx .
The sections which I have given of the calyces jn those types show that at the peripheral limit of the calyx a thecal ring of calcification exists , which is m greater part independent of the septa .
These are skeletal features .
The actual anatomical features of the polyp are m favour of my view .
FowLER 's work has shown very clearly the near affinity which exists between Pocillopo\#183 ; ra , Seriatopora , an\lt ; l Madrepora .
The whole tendency of my wor k has been to show that the thick , peripheral deposit around the calices of the simple Turbinolidre proves the calicoblastic activity of the same region of the polypal wall which is supported upon ccenenchyme in these colonial thecate types .
Whether the camenchyme is porous or compact I regard as a minor feature .
Moreover , in types which are said to have compact crenenchyme , occasional spaces are left unfilled at parts of rapid growth , ex. Stylophora .
In spite of a certain attractiveness in the subdivision of the Madreporaria into two ma.in Orders , Eitthecalia and Pseudothecalia , as above arranged , I do not feel that it is fully justified in the present position of research .
The very fact that the Eupsammidre at present include one or two forms which zoologists have shown to have either no edge-zone , or a very rudimentary one , warns against premature acceptance of these sub-divisions .
The most probable explanation of this is a certain retrogression in the edge-zone , and correlative increased active development of the remote neighbourhood of the polypal wall .
The point is however not clear .
Again , the Stylinidre show in their other skeletal features relationship with the Oculinidre , yet they have a pseudotheca .
Further , both pseudotheca and theca may be totally absent , and yet the coral show by its septal and basal structures clear points of affinity with genera in which a wall is developed .
It would be thoroughly artificial to gather such types together and remove them from what has hitherto seemed a natural position .
The farther we go back in time the greater is the number of genera which seem to have only epithe0a and septa , and no tangential thecal structure .
I regard this condition as the primitive condition among Madreporaria , and one which has been retained to the present day by a few genera , while others have developed various waJl structures ( of pp. 316,317 ) . .
Embryology gives also some insight into this question , since we find in all thoroughly examined cases , that the embryo has no theca , but only a basal skeletal disc of attachment representing the embryonic epitheca .
The primary skeletal invaginations are purely radial , and are pouched up vertica1ly from the basal body wall resting on the skeletal disc .
It seems fairly natural to suppose that this embryological stage presents a typical picture of some primitive Madreporarian .
I picture the ancestral Madreporarian as a small disc or ring ( cf Duncanella ) of living tissue attached by its whole basal surface .
Tentacular feelers were sent out .
on the upper surfa .
, ce of thE ; i cl.isc1 and o:q the lower , very often stolonal or irregular SYS'rEMATld STUDY OF THE :M:ADREPORARIAN TYPES OF CORALS .
315 ccelenteric expansions of the disc were rendered permanent as a nieans of better attachment .
The base of the disc and the rootlets or rugre became strengthened by calcareous cell deposit , and formed a part of the primitive " Rugose epitheca .
" A few phases of evolution having been successfully passed through , the enterprising coral may be pictured as taller and stronger with a whorl of tentacles and a number of rootlets .
Its basal body-wall was pulled up and tucked in at the sides , but curled loosely round the whorl of tentacles , and so far hooded the polyp when the tentacles were retracted .
In every tuck calcareous cell-deposit gathered here and there , just as it did in the basal disc .
The primitive " septal trabeculre " and " spines " thus formed , were therefore continuous with the epitheca , and , in fact , between adjacent " septal " tucks , the ectoderm continued actively depositing epitheca .
If the curls and wrinkles of the peripheral ring of the body-wall were particularly loose : it might easily happen , with further growth , that only the innermost wrinkle kept up the original connection with the septa , while the outer wrinkles became more variable .
ThuR the inner part would be differentiated as a theca , and would be protected by the less regular deposit of the farther wall-wrinkles .
The so-called theca and epitheca are inherent parts of one and the same tangential structure .
Should calicoblastic activity in the peripheral area prove , on the contrary , to be relatively small , and the main area of deposit be in the radial folds , the tendency towards higher complic1:1 , tion of the septa !
structures would be given .
Rapid upward growth of the septa demands some mode of support , and correlatively the pseudotheca would be developed .
The most usual position of the pseudotheca in Palreozoic genera is at the limit of the dissepimental and the tabulate zone .
The steeper inclination of the dissepiments and the depression of the polypal body-wall towards the centre of the calyx were probably the chief factors in determining the secondary appearance of a pseudotheca at that part .
At the same time , it is highly probable that the first cause of a pseudothecal development was the narrowing of the calyx so characteristic of Palreo.zoic corals , and connected with the process of budding .
In the course of time , the increasing strengthening of the septa and the pseudotheca would lead naturally to retrogression in the development of the epitheca .
More especially would such retrogression make itself felt in colonies where vegetative budding within the calyx went on to any large extent .
Thus the Astrreiform colony became a fixed type .
One other modification of the primitive Madreporarian ma.y be mentioned .
Where a brood of young corals grew up together , any wall-outgrowths , e.g. , rootlets ; readj}y caught each other for support ; and since the coelenteron of each polyp was continued into the ramifications , the fusion of these meant an open connection between adjacent polypal cavities .
Probably this was the most primitive form of a Madrep9rarian colony .
It was at any rate very early attained in Madreporarian evolution 2 s 2 31G DR. MARIA M. OGII , VIE ON TI-Im MICROSCOPIC AND ex. Ericlophyll-um , JJ ; Iichelinia , etc. Old-fashioned Palreozoic colonies such as Ericlophyllum , are admitted to have a " Rugose epitheca , " an\lt ; l " irregular continuations " ; newer-fashioned colonies , of Turbinm\#183 ; ia for example , are said to have a " theca " and a " camenchyme " ; yet I see no real morphological difference between the old and the new fashion .
The costate wall has been developed as a wonderfully complete cana1icular system , but the crenenchyme in the case of Madreporidre , Pocilloporidre , and Poritinre is nothing more than a ramifying modification_ of the primitive tangential structure ( cf antea , pp. 261-268 ) .
I have thus traced the history both of the solid Turbinolid " wall " and the " Perforate wall " of the allied reef-builders back to the basal skeletal disc of tho primitive Madreporarian .
The theca with epithecal surface is probably more ancient than the free " inner wall " or " pseudotheca " of Cyathophylloid and Astrreid types and many other facts corroborate this conclusion .
The bilateral symmetry of the polyps of Seriatopora and Madrepora is in keeping with the primitive nature of their crenenchyme .
With regard to the phylogeny of the theca , pseudotheca , and ccenenchyme in Madreporaria we may distinguish the following evolutionary stages :I .
The ancestral stage frequent enough in Palreozoic corals-a " Rugose epitheca ; " without any centres of calcification in connection with the septal centres .
IT.-(a ) .
The.ca in solitary corals.-The differentiation of an inner part of the epitheca as a theca possessing centres of calcification connected with the septa !
centres .
In this case there is never more than one wall , composed of theca and epitheca together .
II.-(b ) .
Pseuclotheca in solitary corals.-The epitheca remains as it was , but a ring of septal thickening is marked out within the calyx at a greater or less distance from the epitheca .
It appears as an " inner wall , " the epitheca as an " outer wall .
" II.-(c ) .
Camenchymate , colonial types.-A permanent connection is formed between the polyps of a brood by means either of a light skeletal framework in the place of thick walls , i.e. , by a porous wall-growth on the " thecate " type ex. Madreporidre , or on the " pseudothecate " type ex. Eupsammidre , or by means of s11perficial compact or sub-compact deposit laid down in the area between adjacent thecre , themselves compact : ex. Pocilloporidre .
The skeletal structures laid down between polyps have no fixed type ; the l'im round the growing horizon of the brood is marked by an epithecal pellicle .
This epithecal pellicle is only the outer specialized zone of the so-called ccenenchyme between the polyps .
It is more strongly and firmly developed and has a protective function , while the ccenenchymal part has only a circulatory function .
III.-Pseuclotheccite , colonial types .
-Two distinct stages are presented , ( 1 ) the fascicular 0r branching colony where each member of the colony may have its pseudotheca and its epitheca well developed and appear two-walled .
This stage is the original form of Cyathophylloid and , generally , of Astrreiform colonies .
( 2 ) .
The SYSTEMATIC S'l'UDY OF THE MADREPORARTAN TYPES OF CORALS .
317 " fissiparous " and " meandroid " stage , where the inner wall in the middle , or near the periphery of the calyx , is the only apparent wall , and may by consta11t selfdivision of a polyp come to include several calyces .
The epithecal deposit may be almost abandoned or retained strictly as a means of attachment for the colony .
Meandroid colonies springing from quite different genera behave similarly in this respect , and the stage is highly characteristic of the most advanced colonial types of Astrreidre and Lophoserinre .
The mere fact that the primitive basal and tangential structure went through these evolutionary stages , makes it impossible to use the morphology of the theca as a basis for the main subdivisions of Madreporaria .
Clearly an Astrreid which was directly descended from a Cyathophyllid might differ in this respect and have to be placed in an entirely different subdivision from its ancestral types .
Again , species of Montlivcdtia with an epitheca , and neither theca or pseudotheca wou]d possibly have to be removed from pseudothecate Astrreids : similarly M\#183 ; ic1'0solena from the Fungidre and placed alongside of certaiu species of Zaphrentis and other Pitlmozoic corals which also had not advanced beyond the primitive stage .
GROUPING OF FAMILIES IN THE MADREPORARIA .
No feature has remained so steady during the long geologic ages as the fine microscopic structure of the " trabecnla , " or are known in its primitive form , the " septal spine .
" Indeed the whole microscopic structure of the septum has undergone little change .
In very many cases absolute identity may be demonstrated in the fine structure and growth relations of the septa a , like of the oldest and the youngest Madreporarian genera .
At the same time evolution reacted also on the septa , and has induced in advanced types far greater complication than existed so far as is known in any of the primitive genera .
No complication however introduces any new microscopic elements of structure .
The same fundamental lines of structitre established in the septa of primitive .Madreporaria are followed in the most highly differentiated recent types ; elaboration has strictly followed the first principles of septal structure .
Further , di'mensions and form may vary in closely relate\lt ; l. genera , but never the actual microscopic building-up of the Septal t'l'Ctbecula .
It is , therefore , to the microscopic structure of the trabecula as well as to the fundamental lines of structure followed by trabeculce in the septum that I have looked as the basis of any grouping of families amongst the Madreporaria .
Subordinate to that , the general habit of families helps to indicate their affinities .
The following five groups of coral families are not intended to be part of a classificatory system .
They will merely serve to summarize and accentuate some of the main aspects of this paper .
I take for granted all the definitions and explanations of terms which I gave in the last chapter of Part I. of this work .
I refer for instance to such terms as " simple " and " compound " trabecul::e ( antea , p. 141 ) , " area of divergence " ( p. 140 ) , " trabecular 318 DR. MARIA .M. .
OGlLVtE ON T1LE1 MICROSCOPIC AND part " ( p. 235 ) , " growth-segment " of the septum ( pp. 234 , 236 , 237 ) , " fascicle " of fibres ( p. 118 ) , " centre of calcification " ( pp. 125-128 ) , etc. , etc. It will also be remembered that the true original use of the term " spine " was ( 1 ) for spiny-looking short projections fi\#183 ; o1n the inside of the calicinal wall projections which did not , however , unite together in vertical series so as to form a septum ; ( 2 ) for spiny-looking projections up fi\#183 ; om the base of a calyx , also not uniting to form a septum .
These are the only legitimate uses of the term " spine .
" A trabecula on the other hand must have some method of attachment with its neighbours to give rise to a septum .
It understands also a certain length , a sequence of growth .
MUROCORALLIA .
The Palreozoic corals had septa either composed of trabeculre , or represented by a series of spines , and were in this sense as in others the strict forbears of all younger forms of Madreporaria .
The broadest distinction which can be drawn for Palreozoic types refers to the direction followed by the trabeculre in the septum .
In Rome genera ( ex. Pholidophyllum , Acanthocyclus ) the trabeculce are directed almost horizontally inwards from the epithecate wall , and keep the same horizon of height as the wall .
Naturally , the inference is that the fleshy body-wall of the polyp was pouched inwards from the periphery , and folded radially along the septal lines just at the same level as it lay around t.be wall .
Thus the radial and the tangential skeleton kept pace side by .
side in growth .
A still further point is afforded by the microscopic structure , which shows that the main body of the calcareous fibres were deposited along the sides of the septal flaps while very little or none at all formed along the middle , i.e. , the apical angle of the fold .
The laterally-deposited fibres are not in union along the middle of the fold , whose place is signified in the skeleton by a band varying in breadth and appearance according to the original amount of deposit laid down there , if any , and the particular nature of the subsequent changes undergone by it .
This middle band , which I now distinguish as the representative partly of the apical or upper deposit of the septal fold : partly of the original unfilled space between the two flaps of the septal fold , has been called the " primary septum , " " dark line , " etc. , by former authors , while the part which I have called the lateral deposit has usually been spoken of as the " stereoplasm " or " secondary thickening " of the septum .
The application of these terms looks fairly easy in the case which I am just discussing of septa with horizontally placed trabeculre , or the trabeculre making only a very low angle with the , val , but they are terms which only cause mistake and difficulty when one tries to apply them generally throughout Madreporaria , and I therefore deprecate the use of them at all ( antea , pp. 89 , 125-128 , 198 , 256-259 ) .
Suffice it here to note that the septa which have their trabeculre making a low angle with the wall , or quite horizontal , show on microscopic examination of a transSYSTEMATIC STUDY OB ' THE MADREPORARIAN TYPES OF CORALS .
319 verse section a fairly distinct and continuous median band dividing the two side wings of the septum .
The horizontal or low-angled position of the trabecula is characteristic ( among the Paheozoic Madreporaria ) of the Zaphrentid forms , in short , of these types which have a broad tabulate area in the middle and a very narrow dissepimental zone at the periphery practically embodied in the strongly built epithecate wall .
The septa in these types are as a rule short , extending little into the lumen of the calyx .
It is chiefly among these Palreozoic types , too , that the septa are so often vicariously represented by rows of short spines projecting inwards from the wall .
A general glance through the Palreozoic genera possessing these correlative features points to the predorninance of the tangential elements of the skeleton .
In the epithecate wall lay apparently the strength and power of resistan~e of the coral , the soft polypal parts being rigidly protected within the calyx , sunk , so to speak , below the hard out .r rim of wall and epitheca .
The septa , spines !
and tabula::i formed merely an internal framework , which never elaborated.itself to any very high degree of differentiation , The battle with the waves and with natural enemies concentrated itself clearly on the wall .
There we find , in allies of these corals , all sorts of spines and tubercles , quite echini\lt ; l sometimes in type , no doubt serving for the better protection , or possibly sometimes balancing of the coral .
Besides the Pala::iozoic Zaphrentida::i , two other families , the T1.trbinolidce and the Amphiastrceidce , whose ages range from Pala::iozoic to Recent , have their septa built up on the same principle and show precisely the same relations of growth between the wall and the septa .
A still further structural point which these three families have in common is the possession of a well-built theca , whose fibrous elements are set in a direction at right angles to those of the septa .
I have , therefore , grouped these three families together under one main subdivision of the Madreporaria .
And I have chosen a name to indicate the pre-eminence of the wall , viz. , MUROCORALLIA .
The actual size of the trabecula::i will be found to vary greatly within this group .
Sometimes they are extremely narrow , even thread-like to the naked eye , sometimes they are of quite appreciable width .
But these of any one coral vary remarkably little in size .
Their fibrous structure is always the same-simple , with a single " axis " or " centre " of calcification ( whose shape is lens-like or rod-like when cut crossways ) , and the calcareous fibres are placed rectangularly outwards on either eide .
Owing to the low-angled position of the trabecula::i in the septum , transverse sections usually cut for a long distance through one trabecula , or pass very obliquely from one trabecula to its neighbour ; hence the uncalcified axial line of one trabecnla or another is always visible in the thin section and gives that appearance of a " dark band " or : , light band , " as the case may be , between the lateral deposit of fibres .
Very often the " dark " or " light " appearance penetrates to some distance on either side of the actual median line , marking off , as it were , an inner set of the calcareous fibres from the outer .
This may , indeed , possibly have helped at the first to make she term ' ' primary septum " be accepted , But the actual fibrous deposit is oor:i .
tinuous throughout ; it is only that the crystalline deposit laid down at or near the edge of the septal fold is sometimes set at a different angle from the later deposit , sometimes also it contains a larger quantity of uncalcified cell-material and undergoes in consequence slightly different chemical changes .
It remains to mention the structure of the " wall " in these types .
The " true theca " ( eutheca , p. 310 , seq. ) has its own centres of calcification , and the term was originally applied to the particular kind of theca in which microscopic transverse sections display a " dark line ' with independent centres of calcification in the wall just like the one seen along the mid-line of the septum .
But , from what I have already said about the septal " dark line , " it will be easily understood that the presence of a similar line in the wall simply indicates that the polypal flesh took the form of a fold over the wall , the two flaps hanging down inside and outside the wall .
Suppose , now , that no part of the polypal flesh hangs down outside the wall , there would be no " dark line " to mark an apical edge .
Thickening of the wall would go on onesidedly , namely , on the inside .
This is the usual case among Palrnozoic Zaphrentids .
Suppose , again , that only a very small lip of the polypal flesh hung down outside the wall and the interior of the calyx was fairly deep , then the thickness of wall deposit would be greater on the inside and the dark line indicating the upper edge of the fold would be nearer the outside than the inside of the skeletal wall .
Conversely , if the outer lip was much deeper than the inner wall lining , the position of the " dark line " would be nearer the inside .
I have already mentioned the observation in Galaxcea , and have referred to Dr. FowLER 's similar observations on Lophohelia , where the position of the " dark line " in the wall varies according to the age of the polyp .
Now , this is extremely interesting , and shows how the so-called " true " theca is but a modification of the primitive epithecate wall of corals ( antea , p. 316 ) .
For , as any polyp forms the habit of having a lip of flesh around its wall ( in other words , an edge-zone or " Randplatte " ) ( 1 ) , the epitheca is necessarily la.id down below instead of at the rim of the calyx ; ( 2 ) a fold of flesh encloses the upper part of the rim or wall , with the result that a " dark line " marks the apical edge , and a two-sided calcareous deposit corresponds to the double flap of polypal flesh .
But one has only to study living Turbinolids to convince himself of the worthlessness of fixing any great c)assificatory value on the state of development of the epitheca or the depth of the edge-zone of polypal flesh outside the wall .
I look upon the primitive theca such as we see it in the majority of Zaphrentids as a true theca , inasmuch as the laminrn composing it are placed tangentially , and are not lateral thickenings of septal deposit which have met across an interseptal loculus , as in the case of a pseudotheca .
CCENENOHYMATA .
Exactly the same " s\#163 ; mple " structure of the sept.al trabeculrn , and the same lowangled or horizontal direction of the trabeculrn in the 1 : , eptum , characterise several SYSTEMATIC STUDY OB~ THE MADREPORARTAN TYPES OF CORALS . .
321 colony-building families of Madreporaria the M adreporidre , Pocilloporidre , Oculinidre , and Stylinidre .
Moreover , the relations of growth between wall and septa are virtually the same ; But an important difforence of habit has led me to place these families in a separate group .
The Zciphrent\#163 ; dm are exclusively solitary corals ; the Titrbinolidm may be also said to be solitary corals , for only one or two isolated cases of a branching colony have been cited ; the Amphiastrmidm include both solitary and colonial corab , but the corallites of a colony always retain a high degree of individuality , since each member of a colony is surrounded by its o\ un distinct epithecate wall .
On the other hand , these families- the Madreporidre , Pocilloporidre , Oculinidre , and Stylinidre-only exist as colony-builders .
The colonial aspect also is the most marked feature about them , it is in the living state that of innumerable tiny polyps embedded in a common camosarc .
The polyps carry on the active vital functions , the camosarc is circulatory and nutritive , very often also it carries on a sexual budding .
In the dry skeleton one sees the corallites surrounded by skeletal deposit , which varies a good deal in its actual form , but goes generally under the name of " ccenenchyme .
" It is in referenee to this striking and invariable feature in these families that I have chosen the name of cwnenchymata for the sub-division .
At the same time , not all ccenenchymatous corals are included here , only those which possess the characteristic trabecular and septal features .
The trabeculre then are all simple , they are almost uniform in size in any one coral , they are either perfectly smooth externally or are marked by regularly-spaced small granulations , their apices end smoothly at the septal edge , or give\#183 ; rise to very fine serrre of equal size , like the trabeculre themselves .
The position of the trabeculre in the septum is like that of the MU ?
\#183 ; ocorallia , whenever , as in them , the septa are not prominent , but keep the same horizon of height as the wall or are even sunk below it .
It is true that amongst this group there is a tendency for the septa to become rather more prominent arid correlatively for the trabeculre to take up a higher angle with regard to the wall .
Similar advance is being made indeed among the Turbinolids .
Still ; these external appearances to which I have just been referring , are good practical guides in classificu.tion .
Of course , the sure guide is alwa , is the microscopic , the simple character of the trabecul re , the lens-shaped axes of calcification , and bi-symmetric arrangement of fascicles .
Whilst these are al1 rese mblances in the structural plan of the Murocorallia and the Cwnenchymata , a difference at once arises in consequence of the ccenenchyrnatous develop1m1nt .
Costal radial structures are present in the ccenenchyme , and their trabeculre are placed in a direction exactly opposite to that followed by the trabeculre in the septa .
The costal radial structures may be continuous with the septa through the wall of the corallite , or they may be introduced at the wall nlternately bet.ween septa .
The reverse position of the true .beculre in septa and costre may be seen in the illustrations of crenencbymatous types above ( figs. 2 , 35 , 5G ) .
32 : ?
DR. MARIA M. OGILVIE ON THE MICROSCOPIC AND A true wall ( cf pp. 158 , 159 ) was primitively present in all these families , and has been retained in the gl'"at majority of genera .
But especially among the Stylinids , more rarely among the other three families , the true wall became obsolete , and the inner limit between tabulre and dissepiments became fixed as the wall of the corallite .
This only helped to enlarge the area occupied by cCBnenchyme in these types , a , lthough the original limit is , in many cases , perhaps still indicated in living polyps by the fact that a fairly wide lip or " edge-7.one " with mesenteries extends beyond the wall of the corallite , and continues the mesenterial canals beyond the visceral cavity of the polyp over the coskc and cCBnenchyme .
This edge-zone passes then imperceptibly into true cCBnosarcal flesh , clothing these parts of the cCBnenchyme which are more remote from the comllites , and as a rule less regularly built .
Septocorallia .
I have already described under Murocorallia the one type of trabecular structure and direction in the septum , which characterised a very large number of the earliest known corals .
The other type whicl1 stood in strong contradistinction to it was shown by the ancient Cyathophyllid.s .
The septa of the Cyathophyllids were as a rule of considerable breadth , that is , they projected a long distance into the lumen of the calyx .
Correlated with this , is the fact that the central tabulate area was relatively small , while the broad peripheral area had a rich , basal deposit of dissepiments in the interseptal loculi .
The septa had a greater extent of surface , speaking generally , in Cyatbophyllids than in Zaphrentids , and that fact in itself sufficiently explains the greftter elaboration of their trabecular structure .
Strength and firmness seem to have been the law governing the septal skeleton , and clearly the attainment of this might call for considerable deviation of plan in the arrangement of trabecuhe according to the dimensions or shape of the septum .
In the Cyathophyllids the ground-plan was to have all the trabeculm pointing to the upper edge of the septum .
It may be observed in living corals that the shape of a septum frequently goes on changrng during the growth of a young polyp , and the first-formed trabeculre cannot , in that case , point to the ultimate upper edge of the septum in its mature shape and position .
But the actually growing trabecuhe do so all round the septal edge .
The shaping of themselves , so to speak , according to the shape of the septal edge at any particular period , causes the trabeculre to be very different in their size .
Some are mnch broader and , indeed , thicker than others .
The smaller are always distributed near the inner edge of the septum , the larger occur at the more exposed parts of the septum , i.e. , usually where the curve of the upper edge is greatest .
The actual arrangement of the trabeculre in the mature Cyathophyllid septa .
I type , is that the trabeculre are ( l ) either all directed inwards and upwards from the epithecate wall towards the centre of the calyx , or ( 2 ) the trabeculre fan out , SYSTI~MATIC STUDY 0111 'l'HI\#163 ; l'YlADRBPOR .
AlU .
AN 'l'YPES OF CORALS .
323 right and left in the septal plane , passiug on the one hand inwa1\#183 ; ds and upwai-ds , on the other band outwards and upwards to the septal edge .
The latter varie~y is by far the most general , the first variety being practically limited to some early types .
In both varieties the trabeculm meet the septal edge at high angles ; their apicAs indeed form practically right angles with it in the case of curving septa .
Curv"'ld septa are however rather exceptional among Palmozoic genera .
'l'bey are present notably in one or two genera , whose affinities have usually been considered doubtful , I mean the low discoid genera , like Palmocyclus and Hadrophyllum , where the septa although low have .
a well-curved edge and trabeculm , diverging in fan-like form right and left .
It will at once be seen that this arrangement of trabecu]m is bound to give a different aspect to the septal edge .
Whereas in the Murocorallian type the trabeculm lie along the septa , making a very small angle with its upper edge , they seem here to stand up in the septum and present their apices freely at the growing upper edge .
A transverse section is bound to show some difference in appearance , and the difference will be greater in proportion as the trabeculm in the latter type happen to be big and strong-in other words in proportion as their axes are at greater distances from one another , .
for each trabecular axis shows a certain uncalcified part in section , while the fibrous part cf the trabecula is disposed more or less radially around the axis .
The trabecnlar axis instead of being lens-shaped in transverse section , is really only a point or group of points in well-preserved specimens , and the term " centre of calcification " was originally given to describe exactly such a trabecular axis seen in transvrrse section .
In the first part of this paper I have described in full the sections made through several leading types of Astrmidm and Fungidm ( cf pp. 127-189 ) .
It is therefore unnecessary for me to repeat here the details of the microscopic structure characteristic of these two large families of Madreporaria .
The important fact is that these younger families agree with the Palmozoic Cyathophyllids in the essential features of trabecular structure , and in the position of the apices of trabecu]m at the upper edge of the septum .
Some of these younger types have precisely the same distribution of trabeculm in the septum as the older types .
Others again have a more elaborate system .
Instead of having only one area of divergence in the septum from which trabecuhe fan out to the right and left ::and vertically upwards , these elaborate forms have several such areas ( cf figs. 14 , lfi , 37 , 38 ) .
These variations are always associated with a specially large size of septum , or with rich toothing and granulation of the septum .
As I have already said , however , no degree of elaboration alters the fundamental law that the free apices of the trabeculm are at all periods of growth directed towards the upper edge of the septum for that period .
The differences in the septal structure of the Palmozoie Zaphrentids and Cyathophyllids would seem to indicate a simpler development of the mesenterial muscles in the former .
Certainly these differences have not become less in the younger 324 DR. MAB , lA 1\1 .
OGILVIE ON THE MICROSCOPIC AND families descended from these , and the a1mtomy of the \#183 ; typical Turbinolid and a typical Astrreid gives full evidence of the fact that the simpler septa } structure answers to a Jess pleated , less highly differentiated condition of the muscular thickenina of the mesenteries .
r. , It is in the same three families , viz. , the Cyathophyllidm , Astrreidre , and Fungidre , that the compound trauecula reaches its highest and most varied development , and that the granulations of the septum in consequence are most elaborate .
Many of the forms have been described above ( figs. 68 , 69 ) .
The apparently radial symmetry of the fibres around an axis is in reality the result of the small bunches of fibres which I have called " fascicles " being arranged around an ideal axis .
The appearance of several centres in transverse sections of a compound trabecu]a is produced by the section meeting the origin of the several fascicles .
Now some of the genera belonging to these families are entirely composed of compound trabeculre , others again have compound trabeculre compactly cemented together by simple trabeculre , in still others the trabeculre are all simple , but they differ greatly in size and are grouped together in such a way as to make the effect rather that of flat , spread-out compound trabeculre ( cf Mussa p. 133 , fig. 18 , and p. 123 , fig. 14 ) .
A true wall exists in very few of these types .
It may be said generally that the primitive members possesserl an epithecate wall , but that this wall bad already become retrograde in the great.er number of Mesozoic types .
On the other hand a pseudotbeca had made its appearance in a fair number of Palreozoic types and became almost general in the younger families .
The retrogression of the theca and the incoming of the pseudotheca is a skeletal feature correlative with the whole tendency displayed by these three families the tendency to increase the importance of the septa .
The septa and not the wall have practically the whole weight of the polyp to support , and hence arose their spined or dentiform edges and their thickly granulated sides .
The predominating part played by the septa in the skeleton of these types has induced me to choose Septocorallia as the name of this group of three families the Cyathophyllidre , Astneidre , and Fungidm .
The uame is also designed in contrast to the gronp of MurocMallia .
Spinocorallia .
Although by far the larger number of the Palreozoic Madreporaria belong to the Murocorallict and Septocorallia , there still remain a few types , Cystiphyllwn and its allies , with a quite peculiar character of septa .
Some indeed had no septum properly speaking , but only consecutive rows , regular or irregular , of loose spines placed upon the basal skeletal elements .
These spines frequently grew together and gave rise to a patchy , interrupted kind of septum .
The Palreozoic types of this character are usually classified as Cystiphyllidre .
The Mesozoic period saw a farther development of such genera .
The spines frequently attached themselves both lengthways and sideways so as to produce septal trabeculre .
Still , however , many spines remained SYSTEMATIC STUDY OlP THE l\IADREPORARIAN 'l'YPES 0~ ' CORALS .
325 free and prevented the same regularity of plan in the arrangement of trahecnhe which marks the septa of other types .
Even among the recent representatives of these forms , the septa seldom show any very well-defined law determining the direction of their trabeculre , and the looseness of many of the " trabecular parts " still indicates their past history as loose spines set in the calyx .
I have classified these Mesozoic and younger forms i.n the family of Eupsammidre : \Vherever a definite arraugement of trabeculre is presented by an Eupsammid genus it approaches most nearly to that of the Septocorallia .
It shows either highlyinclined parallel trabecuhe or irregular fan-shaped groups of trabeculre .
The trabecul::e are short , and successive trabecul::e ran , lie :fit on exactly end to end , with the result that irregular gaps are left in the septum .
It is just this 1rregularity which essentially distinguishes all Eupsammids from Astrrnids and Fungids .
The microscopic strncture of trabeculrn , which unite as a septum , likewise agrees with that of the Septocorallia in the main , but the axes are lenR-shaped , the groups of calcification centres elongate : and the arrangement of the fibres assumes in consequence more the appearance of bi.symmetry than of radial symmetry in transverse sections .
Nevertheless , the trabeculrn are really grouped , and the fhscicles have a very distinct character , the primary part of the deposit near the origin of the fascicle showing crystalline :fibres set rectangularly to the axis , while as the deposit is farther remote from the origin the crystalline fibres become more and more oblique in their setting .
And it is this vny principle , together \#183 ; with the lens-like form of the axes which produces in transverse section of an Eupsammid septum the appea , ance of an inner and outer septal deposit on both sides of the mid-line .
No doubt this kind of appearance gave rise to the idea of a fuur-leaved septum held by some palrnontologists .
But these are appearances shown only by particular sections , and owe their explanation to the facts I have just indicated .
If I were to go on appearances , I might say all septa in which the trabecular axes are lens-like in shape and close together , appear to be at least two-leaved in transverse sections , in some cases four-leaved .
Whereas all septa in which the trabecular axes are thread-like or rocl-like in shape , and are not very close , appear t.o be a single-leaf in transverse section , in some cases th1\#183 ; ee-leaved ( i.e. , where the deposit near the origin of fascicles differs in its crystctlline setting from the deposit more remote ) .
But these appearances cannot lead with any security to a systematic analysis of corals .
Thus , one must not be deceived by the appearances of Eupsammid transverse sections .
The affinity of this family is undoubtedly rather with the Astrrnidrn and Fungid:::e than with the Madreporidrn in spite of Enw ARDS and HAnrn 's classification of it along with the last-named family among the " Perforata " ( antea , p. 201 ) .
And this seems all the more reasonable in the light of the parallel fact that its Palrnozoic antetype , the Cystiphyllidrn , has marked affinities with the Palrnozoic antetype of the Astrrnidro and Fungidrn , namely , with the Cyathophyllidrn .
32G rm , .
l\[AH , IA l\J .
OGILVIE ON 'l'I-rn MICROSCOPIC AND The s,1 , me fuwfomcntal feature which distinguishes the septa of Eupsammids gives also the pec uliar clmracter to its wall , coBtal , an\lt ; l cmnenchyma ] elements .
The free spine is everywhere the skeletal basis , and l1as given rise in a nnmber of genera to an irregular synapticnlar network around the periphery , where costal and basal elements combine to build up a porous wall structure .
The same elements extend as a porous ccenenchyme in colonial types .
The true wall present in primitive types became then greatly modified in the course of time .
The predominance of the spine in all primitive types of this group , and its influence in the p~rticular line of evolution followed by all the skeletal parts m younger types , have suggested to me the name of Spinoco1\#183 ; allia for the group .
I provisionally include the ancient family of A rehceocyathidm within this group in addition to the Cystiphyllidm and Eupsammidm .
I do so with some hesitation , since I have not had an opportunity of personally examining the Archmoeyathidm , and can only judge of the affinities of the family from published descriptions .
Porosa .
The family of Poritidm is practically specialized as a reef-building type .
Considerable areas are covered by the spreading growths of genera belonging to this family .
Their most striking structural features seem to me to be the net-like meshes formed by their trabeculm in all the radial and tangential skeletal parts , and the rapid succession of fascicles in individual trabeculm .
These features are undoubtedly correlated with their life-habits , since rapidity of growth and lightness of skeletal framework are both ad vautageous to an extensive reef-growth of any one coral .
There is no great elaboration of cal yces , sirnpl y a few septa , of quite uncertain number , are marked out here and there in a circular group .
These circular groups an~ closely packed together or are far apart from one another , quite according to adventitious circumstances .
In the main mass of any reef-growth the trabeculre stand up erect in the septa of these calyces just as they do in the intervening skeletal parts .
But often some particular position of the calyx has the effect of producing deviation from this rule , and causing the trabeculm to grow obliquely .
In small colonies it is frequently the case that the trabeculm make an angle of about 70 ' or 80 ' with the calycinal rim and diverge inwards to the calyx , and outwards away from the calyx .
This becomes even more marked in the calyces near the periphery of a small colony .
As may be imagined , the divergence of trabeculm at once causes a resemblance between such colunies and an Astrmid or Fungid colony .
But we have here only an external resemblance .
The relation of the septa both to the tangential and to the basal structures , affords important distinctions between the Poritidm and both of those families .
Microscopic examination , moreover , reveals fundamental differences in the fibrous composition of the skeleton .
Besides , I hold that as the Poritidm are reef-builders on the grandest scale known SYSTEMATIC STUDY OF THE MADREPORAIUAN TYPJi ; s 0.1 !
' CORALS .
327 in the Madreporaria , it is the reef-type which should be taken as the standard for the family , and not the small astrreiform colonies formed by some species .
Septal and mural trabeculre in Poritidre are only distingni.shed from the trabeculre in the rest of the skeletal p9xts by being , as a rule , more closely mRtted together by interstitial 8econdary deposit .
The microscopic structure in all trabeculre of Poritidre is always grouped , with the fascicles of any particular " trabecular part " arranged circularly at successive planes of origin .
The branching-out of the fibres of the fascicles produces a corresponding bulging-out all round the trabecula at the emergence of these fascicles .
Thus a succession of trabecular parts is like a nodose stem where the thick parts correspond to the emergence of numerous fascicles belonging to each nodular portion , but do not form the cross-bars ( synapticulre ) .
If one seeks to compare this with some of the Fungids which have fenestrated septa , it will be found that the fascicles have their points of origin at or near the base of each trabecular part in the latter , and the bulging-out practically forms the pseudosynapticulre .
The pores characteristic of the skeleton of Poritidre are between the small parts of the adjacent trabeculre .
Another set of pores , however , arise in consequence of the " true synapticuhe " ( antea , p. 222 ) , which cross the interseptal loculi , and whose size and structure repeats that of an individual trabecular part .
Thus the whole framework has the appearance of similar rods set crossways .
In most cases , however , a fine vesicular tissue of dissepiments is also deposited in the midst of the rodsc : : .
i.ffolding .
The Triassic Spongiomorphidre described by Professor FRECH ( cf. antea , p. 307 ) show the same essential features as the reef Poritidre and may , I think , be safely included with them as an independent group , to which I give the name of Porosa .
There are , too , certa , in ancient forms which from the descriptions seem to have much in common with Poritidre .
I refer to the genus Thec1:a and its allies .
The ancestry of the Spongiomorphids and Poritids will have still to be thoroughly investigated .
So far as my reading goes , the Theciidre nppear to be at least close allies .
It comes to be a question how far the recent Poritidre are related \#183 ; with the recent Madreporids .
The external signs on which EDw .
ARDS and HAIME chiefly founded their group of M. perforata , including in it Eupsammids , Madreporirls , and Poritids , carry no longer any real weight .
Porosity of the septum may be seen to a certain extent in a few Astrreids , most Fungids and Eupsammids , and in all Poritids , whereas the septum of Madreporids is , as a matter of fact absolutely compact .
Such a feature only leads away from the fundamental laws of coral structure .
A sure indication of relationship between the Poritidre and the ccenenchymatous group generally , is afforded hy the peripheral arrangement of the skeletal elements .
The introduction of new costal trabeculre in the wall , the multiplication of costm as the distance from the wall increases , the variation in the area covered by the ccenenchymate skeleton , and even in its precise form , are all features which show 328 DR. MARIA M. OGILVIE ON 'l'HE MTCROSCOPJC AND strong resemblance of habit between the QCBuenchymata and the Porosa .
At the sttme time , the calyces and the trabecuhe of the latter have too many distinctive features to permit actually uniting the two groups in the present light of our knowledge .
It rna.y be that zoological investigations will advance this matter , which is certainly one of the mosL interesting fields of research in the whole of the lVfadreporaria .
Phylogenetic Scheme of Macfreporaria .
" ' ' I now subjoin a general phylogenetic scheme designed for the purpose of showing at a glance the main lines of generic descent followed during the working-out of Madreporarian evolution .
I have entered the most familiar Madreporarian genera upon particular curves corresponding to the geologic periods which seemed to mark the most distinct phases paHsed through by corals .
These periods are--(1 ) Cambrian to Devonian , ( 2 ) Carboniferous and Permian , ( 3 ) Triassic , ( 4 ) Jurassic , ( 5 ) Cretaceous and Old-Tertiary , ( 6 ) Young-Tertiary :md Recent .
Each of these six periods of time forms what I might call a natural moment or phase in the history of corals .
Each successive age saw the retrogression or extinction of many of the types which had flourished in the preceding age , while some newer form of coral life reached its high-level of adaptation and success .
I might institute six sub-orders of Madreporaria for these six geologic periods with quite as much right as underlies the sub-orders of Rugosa ( Tetracoralla ) and Elexacoralla I The result would , at least , be less misleading , since it would profess to be nothing but a convenient historical subdivision of the Madreporaria !
And this is indeed a most useful method of studying any order of animal life-to trace the rise and fall of each successive wave of change that swept over it , to gather the poor broken types and watch the farther course of those that floated on , to scan the rocks of old ocean beds and reason out the why and wherefore of the main tumults of change and migration which shook the animal world .
These old sub-divisions of Rugosa , Tetracoralla and Hexacoralla , have therefore done their service to the knowledge of Madreporaria .
They gave expression to an idea that the corals of the Palreozoic world ceased to exist and a new class of -* ' A \#163 ; ow genera in the scheme ( p. 331 ) marked with an asterisk have been erected lately by myself ; the descriptions of them and illustrative figures will be found in my " Monograph of the Jurassic Corals from Stram berg " ( ' Palaontologische Mittheilungen , ' Koen , Stuttgart , 1896 ) .
SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
329 corals began to people the seas in Triassic time .
As the science of Palreontology advanced , this idea was felt by many to be out of touch with the march of evolution which could be traced satisfactorily in so many other orders of animal life .
Research concentrated itself more or less on the question .
QUELCH , ORTMANN , and others showed the weakness of the sub-divisions ; the names were disowned , and others suggested .
Still , the best palreontological authorities at home and abroad felt that for practical purposes none better were forthcoming than HJECKEL 's or MILNE EDWARDS and HAIME 's , and have continued to give their countenance to names which express only an outworn idea .
My own opinion is that the most intelligent way of studying corals is not to approach them by means of any fixed system of sub-divisions , but by some such way as I have tried in drawing up the Phylogenetic scheme below , where the worker first studies the anatomy of the soft parts and the structure ( microscopic ) of the skeleton in one common type of each of the living families of corals .
According to my classification there are ten of these .
Then , having a general idea of the physical conditions characLeristic of the six geologic periods enumerated above , it is no very difficult matter to work back through the outstanding types in each of the past ages .
Thus one cannot fail to note how the types approach each other in the preceding ages , and how the whole many-branched order of living Madreporaria springs from probably two main trunks .
[ I have borrowed two names , the Zaphrentoidea arid the Cyathophylloidea , to signify the two parent trunks , each with its ramifications of younger families .
These names were given by Professor NICHOLSON* to designate Palreozoic groups of families ; u. third Palreozic group erected by him was the Cystiphylloidea .
At the same time , as it may be open to misunderstanding if I include Mesozoic and Tertiary families under names which were previously of Palreozoic application , I think it well to suggest alternative terms for the two parent trunks .
The Zaphrentoid families may be called Madreporaria Haplophracta in reference to the simpler structure of the septa , while the Cyathophylloid families may be called Madreporaria Pollaplophracta , , referring to the more complicate structure of the septa : Zaphrentoidean Families or Madrepora1\#183 ; ia .Haplophracta : Zaphrentidre , Turbinolidre , Amphiastrreidre ; Stylinidre , Oculinidre , Poci1loporidre , Madreporidre .
* As used by Professor NICHOLSON , the terms Zaphrentoidca and Cyathophylloidea are to all intents and purposes identical with the use made in this paper of the names Zaphrentidoo and Cyathopbyllidre .
The family names , Zaphrentidce and Oyathophyllidce , are not used in the same latitude of meaning by different authors .
Professor NICHOLSON makes each of them only one member of a Palreozoic group of families , and to these groups he gives the names respectively of Zaphrentoidea and Oyathophylloidea .
In my arrangement of families I have accepted the usual wide sense of the family names , Cyathophylloidean Families or Madrepo1\#183 ; aria Pollaplophracta : Cyathophyllidre , Astrreidre , Fungidre ; Cystiphyllidre , Eupsammidre .
The Haplophracta have retained very simple trabecular structure in their septa .
The trabeculre are small and uniform in size , the fibres are bisymmetrically arranged , the centres of calcification very closely set in the median plane of the septum .
The septa of the forms included here have been usually described as " smooth-edged , " and either with " smooth septa } surfaces , " or with " rows of granulations parallel to the septal edge .
" The Pollaplophracta have attained greater variation and complication in the trabecular structure of their septa .
The cEmtres of calcification , instead of being uniformly spaced along the median plane of the septum , are aggregated more thickly at certain intervals .
When the " centres " are placed serially in each aggregate , the fibres form a long elliptical figure in cross-section , but when the " centres " crowd still more closely together , the fibres radiate out and form practically a circular figure around them .
Thus we pass from the bi-symmetric to the radio-symmetric trabecula , and the progressive stages are shown in the early members of the Pollaplophracta .
Such elliptical and circular groups of fibres really correspond to a number of combined trabeculre , and give rise to prominent granulate ridges on the septal surfaces , ending frequently in apical teeth at the septal margin .
The elaborate structure of these septa is a skeletal feature in keeping with more elaborate musculature of the mesenteries in the living polyp , just as the simpler , smoother septal structure corresponds with less specialization of the mesenteries .
I divide , then , the Madreporaria into two great groups-an old-fashioned , primitive group , conservative of type , slow of advance , prone to retrogression-and a p1\#183 ; 0- gressive group which has always been in the forefront of evolutionary advance , quick to adapt , seldom retrograde .
These are chg , racteristics true not only of the structural features of their septa , but throughout the whole calyx .
The affinities of the reef Poritids must remain doubtful until an exhaustive anatomy of their soft parts has been accomplished. .
It is possible that they may have sprung from an independent parent trunk , the same from which the Triassic Spongiomorphidre , and certairi Palreozoic types , e.g. , the Theciidre took origin .
--12th October , 1896 .
] ~ q l ' ' h \#183 ; asrraeidae SJy ; , \#183 ; .
A~~1n , dqe o~ a. t : .
" \lt ; ~ ' : , c a , ... .
C'\~ _ : , --------- -- --~ u , 1 5~-\-\Jt ___ ~ i --------------- .
, , e , % .
, ..~ . .
, .
~ ~ Qo \#183 ; !l ' f ... , ~~$ , t } .
~ ... , ~ ofT ' ~ ~ ~ 1 ~ .._'t \#183 ; , !
'J ~'ti .
( ' , ... ~ ~ ~ ~('lo~ --- ------- - - ---::1 : ; g :1 .-.:.'fJ ~~ ' ~(i,.~ \ \ , , ~- go ~- ~'-\lt ; .
---- ~~ -~ ; : , ~ ~(J ~t : , ; , \ - \lt ; ~ , , , ' ~ ~ ~ ~~- ~ ~ '$ --- ... ; S..__ .tr !
Vj'\G''ti ... ... . .
~ .
0 f\gt ; \ .
\ b ~ ~ ~d \ .
, ' ~ 1 ; -t ~ _ ... ~ ti .
Q\gt ; ~~ ~ l - ... ... $$-~ ~ ... ... . .
0 ~. .
, ..~ .
_ '}t'~-:::- -03. .
" % ~ ..~ ~ a. er . !
' !
I,.~ ~ , ~ ~'Ob ~ ~'t , ~ ' ... X ' ~ ... ... d'- ? .
A~ Q ) ooo ... . .
- : : : ; .%~~-.:-~- .
~ " . !
t. CJ ) ?
i ' ... c. 3 '"O -~ -~ ~ ~~-~ cyt " uQ : ' , \lt ; , \lt ; ? .
: o , ... .
o.- : : .
\lt ; '.".- : , \lt ; : -\lt ; \gt ; , ~~ - ; . .
~- ~ oo ~ \gt ; u , t : " ~-=. .
, ; _t ti"ii ; " " b o ~ tt \#183 ; !t.J ~ ; / v~ ~'\lt ; J~ , ,. .
\#183 ; o. 0 , .
, o ~~---G~"q. .
( '.~ .
- ~ ; .
; ... ~~oo . .
, .
:r:::o~--a~~ c ; .
~aa..~~r ; : ; ~ c~~ ~~-~ ~ G'ti~fJ~ ' -~ . .
':\ , , . .
" - ~ ~o , !
, .
, ro\lt ; " ; . .
..~ . .
, " ' ~:.- .
, / \lt ; ~:t ~ ~ ~ %~ ~\lz-~~~i~\#183 ; i\#183 ; ~-l- --H.ff- !
l f ~ , l q : '- , , l - .
, " .~ ; - !
" '\ , ~ ( ' 0. .
... , , .
, ,le'"~ .
... . .
~\lt ; \gt ; 1,.~~ ' ?
' % " " " " \lt ; " ' ~ ~ ~ " =- ; ; ; , ~isg , as"'i i ~ \lt ; f:~~"'-~'l-l I ~ .
, #~~ , ~i ' .
v , \amp ; .\,.v"-.\i , . .
, ~o , 'o..- : .
, o. , J--~:~o--=.o : , :\lt ; :- .
, c , ~ ~d O ..~ . .
... .- a.''ao~ii~ 1_ '-lt .
J~ ~ ~~ ... ... # ~ ( , ,~ ':\gt ; \ov , " ( \\ .
e. ' - / O~ i:'~- , ; =-::0 " " ' o~ ~ o , ,\lt ; 1 ' ~. . !
.~- ~ - ~ n ~ -Q , \#183 ; : , 'i , ~ ... ... i''b ~~'t\gt ; '- , , ~\ r , \ } ~c\ " e ... . .
/ , ~ " ' ' ?
_. .
, , , ; :J " i , ~ O , -- - : : .
90 O " =\gt ; : : : , .
, " " ' , , .
tP.'\1'\amp ; -- d ' ~ d~ ' ?
-- " " ,. .
' ; .
''------ -- 'b S ~ ... . .
cP ' ' o. ( \~ , , , s. 1\ ' ' lA , ~ ~ , \ , ~,1~ . .
, .
, ,. .
, , o - .
, - .
, , , .
, _ - : : ; .
_-i : ; ;-n ----\lt ; : : i:l ' f:\gt ; O , ; , \#183 ; \\V " ~ , -X\ , , \ , 'Ir ~v , : , " 'J~_ .
, .
/ '\lt ; '..-:~- -- ~ " ' , . .
.- ' ?
, ~ ~ -- ; , , .
\#183 ; t\gt ; , ,\#183 ; .- ' ' \lt ; \gt ; ~~- f- ' 'o~\O~/ \t , o\lt ; .
'Q " ~ ' Cct %-q , ~Q o ~ ~ G : ) ~ -. .
' f : , ,\/ .
~ '\ ' " , .
~ .
1\ .
~( ) ~- - , u. .
~ ,.o/ .
,. .
, / d ' , ' o-~.S:-- : : : , , .
.,..,. .
, C C ( Ji ' , oq , t.f ~ , ?
\gt ; .('("eK .
, ~t~..,. .
ji(:-'t !
..fl ~e , \\1 ~ \I.I'\ eh ' 7 : .
~q"'o/ / , ' 0/ ' o\gt ; ~ , , 3 ~ ~ f ~~ -~~ \ t\t .
~ ~~\o.'\a .
f e. f l ) .
-~tr\amp ; : ; ; , ,"Oc-so , 'F.r , , , .
, , , , , , I / , , , cl , ~- : -~-~ .
, , , , .
.S ' ' , , , , , . .
'tli~ .
, ~~t c , on .
M5 !
!3e6 t:J ~~Co " f ' f\lt ; l/ '/ 2 " ,1 I , ' / 't-J\gt ; ~~ ~Oft .
' t , ,~01 ( ..\"l'\0 \ e , , .
, _ , n r,. .
c-.Y~ , ,.~ u- .
, h , ' ~ / ~ # ' , .
\ , 't\gt ; " rr~e .
, 0..0,.~ , " ' " " ... .'.l :-4 : , , e '\'\lt ; '- ' , , , , e. , \~\#183 ; 15~\amp ; \ ' tl Sch .
99-d',.._ , , r~e ' o':Ydr , / , ' 0 .
\ , \-JS .
\.'Q : ' ~ IJ .
: !
f.J't2/ ' , t , v ; .s , fos , ,d/ Q : ; .
, J ; .oc .
/ / \\ C ) \#165 ; -cP ' '\ 3,1 , ~re .
1s~wraea \ ,.e , a.rt''~rf\le.4 \lt ; , ' 11ocy , u , Pt , ,/ ~ " ' '-'09 , , .I'\ .
; \lt ; " .~ ' \ c ; yP~ e4 ' : ?
J."'e , o , { ~ , ,4 Cyc{c ? .
, , .
, , , , , .
, o , .
, c : , _~. .
, , , , ; _ ; -s , , .
, / ~ " \ .
1-~r , a , r1~. .
\ \ tJ 1a .
\~ , a~ " - 'o"o111of.tn1/ ,13 S , :J'\lt ; tr , t , L/ .
, :Y\lt ; 11"/ 2 " ' " , \ \ ' srra.e.a \ ttet14 " 1-.Y4 r1Jv .
, rel'ho I ~ 1 ... .r,. .
' 5r~1 !
ae.4 , 'tra.e.a \ P4rdcv a1.1 , Sty1 , o.c / 104/ 111/ , a f , ' ~ \ 'fo5a.5rr f'ie\1ll .
B , i " I V , S ; ')'dl"hu .
, I I .._ , 1 , .
COi'\ e,3 .
\ ' cu : : / ~ uo'coeri11J t1eliasrra , ~ .f Cer , / otiochu , ' .
/ / l\t P~Y \ \ Th. " ' , rraea ; ; o , Pidtytr'och Con .
, ,m , f ; d lJt , - ' . .
1"h .
n..rr ... LeprJphy/ ltd 3 c:- , - IJS ' dS/ Smi/ ia - ~t , a.~ I ,. .
/ j : , : ; , .
Haploph !
/1 ; .
B ; atyH .
Axo , m ; t ; , . ! .
- .
f hdrTl'l " .
, / na oyclo / o. ~ .
t .
f Tri~mf !
i11 G , gan(ostyfis ! ?
; 7nr , ; ; , ; ; , -..'..:':~f ; i , Q__ Toich~_ , r ; aea_ M1c !
o50 e \ ~ ~ Turbinpltd Turbinofid Sty Jot'rochu : J c. , OmphdlophylliiJ Troc , hdfdBt1 : ; ; ; : 1 Cl ) / ofycetJphyllid , , , ' . !
t i ; N\lt ; 1e~ndrd rded ; I i co 1 . .
\ i.sco~eris Cotrloaer is I -f \#183 ; .
~ Ampn\he l. .
\ / Protoaeri .
, t .co~eria LopM , aer , a ll.\gt ; \oph ; peH .
\ 0 .
r , , , " ' C \ .
, , .
, ~on:os8r , ~ ; ;:arhpaer , ~ Und4f14 ::_ QC , U f~t'~ 'oGoef"lo : p hefO .
Seris Cro , ser , - .
, N , I ::\ I \ ~ !
'.~firi\lt ; .
' .
he 'n\gt ; atose , .
Cycioje , ; .
, C ea,11{ro.erts 0 \ .
}o. .
' ' , , , .
\ ; , - , .
' 'B r , , , b , ,loae , \#183 ; './ 'Cl'{ .
er , - .
, , i \amp ; tro \"l~oGo \ :zi : , ?
IJ"J .
, ,11 fl.t,.C1 : '.S , ' \ , , .
\#183 ; , .
51.P , \ / : , - , ,-1\lt ; \gt ; ~ , " .o , ; , ~ , e 03e , , ; , -'1ic,01qc , - .
, , rrQ , ; .C~oe " \ , " .ott / ; :a~~ 8 , de , " 11 , ; id F. I ~'~prfe\ ... 01a .
\ t \amp ; ~ , , , : : : , / ,61\lt ; 1 ; 00.s,. .
, .
7~0 ?
" u11 .
, ; , 3 u11 , p , ; , !
)li\o .
o(il- ~~~e ; ~-:S:- , : ; o~-'%\#183 ; ,./ ;~ / 'de~ ~ , o'\p ~\\.\#171 ; ' ' ~ .
; ; : , -~0 / ' " 0 , , , / I Sr~ , 3 / , -\lt ; 1~ \ .
, ~v..\. .
' '\ , ... Q .
~ lo ~ , ' ; " 'r'o , a , ?
\lt ; ' " ' .
, , , O .'y',4J .
t. .
, q oh 1'o .
' ' : .
, \lt ; ' ' ' , c.~ .
e,1 ~~ \lt ; " 06 .
I :r~ , , 8 .
, , , , ' , .
, , , .
, , , .
:w , - , .
, , .
ol''- ,. .
' ' ; ,.t ' ?
fl " " ' " ' " s.o. , , , , ~ " o , , , _ .
, Jo , c , , , ~ ~o " - o\lt ; .
i . . . .
, ~ . ?
: ~(yo 9 , 'b " " ~ : ; -1 ' , , .
, 0_ , , , , .
/ CJa/ ''/ ,.j , .
, ,.oP , '-- ' ' ' .
o. ~\#183 ; \lt ; .
o. .
,._Q '.f1beo , r-rt- .
-\lt ; 'O\li~ ~o , ,.i : , . .
, ,. .
, _,1 , " \lt ; " 04t .
' :r0 .
, r \''\ _1a .
\ ' e-\lt ; d ' '\ .
, ' 0 ~ X\gt ; : o ~ , l ' ' o~\J ' \ .
' Cl\~ , o'l ' .
\ ; c , 0 0- , P , .
c~ o'Q'"\gt ; ;,0- \\lt ; a.t , \lt ; , Y , o ~ a..~ / Q0\amp ; ::.ou'\lt ; t 0 , \lt ; 1 e. " ) ~ ' !
) '- f , ':\gt ; ~ ~ .
, f ' l. V I 9c 0 '\gt ; .J .
00 , '1 '\lt ; . .
' ' \lt ; J\gt ; 'c YPeo JV-\'(}~~ , # , ~ o_ , / " '00 ) ' - : ; ;. .
~-\#183 ; " , .
0 , ; , of ' c~r , .
, '-~e 9 ' ,.Wlo : .
, c , -- ?
, .
' .
, .
, ... Q ' : .
, ( ::I\ ~ l ,.:S\gt ; ~ ( ' !
, { ) , .-.- , ,0 \lt ; 1:- " ' ' -~ 04o ' .
, _ .". .
~ ' ; , .\#162 ; .~t\ ' ' : J ~ " -~ \amp ; , a o'\\ " 1o \I ) ' .- ; , '.\amp ; ~::2\gt ; eo\gt ; , ~ .
-c , \amp ; ~1 'tl- , '''z .
e~Lq c~ .
'.( ; )7 ~ ~ .
/ i , ; ~Q , ' !
\gt ; \#190 ; \gt ; e. , .
~ , ~').'\~ ec ~ _ , -f\,9 ; ~-'Zc ~- ~ c : : : ::::s ~ ' \#183 ; ~ Q , U1 " 1 ' .
U1 .-3 ~ ~ p , - 0 U1 ct b 1-\lt ; l 0 i-:tj , H oS ~ ~ p.- 0 ~ ttj '1:i 0 : ; :o p. : ; : o H p.~ .
f-'J , \gt ; -\lt ; 1 '1:i : ttj U1 0 1- ; rj : 0 0 t ; : : : p.t : " ' 1/ 1 co co t--'\#183 ; 83 :2 DH ... HARlA M. OGILVIE ON 'l'HJ\#163 ; l\HC1:WSCOPIC AND CHAPTKB , XII.-GENERAL SCHEME OP CLASSIU'ICATION .
ORDER MADREPORARIA .
Family :-Zaphrentidce .
E.H. Solitary corals .
Septa always pinnate in their mode of increase ; well-developed , occasionally represented by series of horizontally-placed spines projecting inwards from the wall .
One to four fossulre present .
Base of the calyx formed by broad tabulre ; \#183 ; a very narrow zone of dissepiments is sometimes present peripherally .
Tangential thecal deposit* present within a well-developed epitheca .
Family :-Turbinolidce .
E.H. ( emend .
OGILVIE ) .
In this family I include , for the first time , a brge number of genera which have been previously classified amongst the Eusmilince ( farn .
, Astrc:eidre , E.H. ) .
I also place here various Palreozoic and Triassic genera previously called Cyathaxonidce ( pp. 269 , 270 ) , and I have re-arranged the Turbinolid genera in Sub-families .
Solitary rarely colonial , corals .
Radial insertion of septa typical , pinnate insertion occurs m a few Palreozoic genera .
* Compare my remarks on " true theca " ( p. 320 ) .
The absence of any distinct " thecal " centres of calcification in primitive Madreporaria indicates that invagination took place originally only in radial , i. e. , septal , lines .
The " epitheca " was the primitive base and periphery in one .
Calcareous laminal , if deposited on the complete inner surface of the epitheca , would give rise to solid laminar deposit at the base of the calyx , and to so-called " secondary thickening " around the calyx .
In other words , the basal deposit of '1.'urbinclidro , such as we find it in living types , is probably only a remnant of the primitive epithecal lining of Madreporaria , and is certainly an exact homologue of the wall of Zaphrentidai .
If we follow the history of this primitive epithecal lining farther , we find that while retained in the basal portion of the Turbinolid calyx , its form became modified round about the calyx , owing apparently to the pouching or rolling up of the soft tissues peripherally , as well as radially , during the upward growth of the polyp .
Thus , too , the epithecal layer lagged behind , and a " true theca " became specialized , whose " centres of calcification " marked the position of the peripheral folds .
The only difference between the " true theca " of Turbinolids and the primitive theca of Zaphrentids , is that the former is a double set of calcareous laminal , the latter a single .
Their structural relations to the septa are quite the same .
The epithecal pellicle of Turbinolidai has usually the appearance of wound threads of calcareous substance , or it may be s1ightly coloured and look like a rough bark .
In no case is it so purely crystalline as the calcareous deposit of the rest of the skeleton .
Apparently , it contains a large amount of organic matter , as in the case of the median plane ( " dark line " ) of septa and wall .
SYSTElMA1.'IC STUDY OH THE MADREPORARIAN 'l'YPES OP CORALS .
333 Base of the calyx never tabulate ; usually a solid laminate deposit is formed , more rarely dissepiments occur .
A columella or pseudocolumella is almost always present .
A true theca or pseudotheca is present peripherally , in all cases coalescent with the epitheca .
Sub-family 1.-TurbinolinaJ .
Solitary corals with round calyces and large styliform or fasciculate columella .
No pali-septa , never lobed , their edges smooth .
Basal deposit laminate and solid , or scanty dissepiments .
Bilateral or radial symmetry of septa.* Sub-family 2.-TroclwcyathinaJ .
Solitary corals , rare occurrences as colonies .
Calyx round , septa radially arranged .
Columella present or absent .
Pali always present in one or more whorls .
Basal deposit laminate , or scanty dissepiments present . .
True theca or pseudotheca present peripherally .
Thecal centres of calcification are frequently zig-zagged round the calyx .
Epithecal basal plate presentrootlets never developed .
Suh-family 3.-TrochosmilinaJ .
Solitary corals with elliptical or elongated calyces .
Columella lamellar , may be replaced by pseudocolumella or be absent .
No pali .
The upper edges of septa are frequently crested at the calicinal lip .
True theca always developed .
Tribe :-DesmophyllaceaJ .
Basal deposit absent or in the form of solid laminate thickening of skeletal parts .
Rootlets very often present ( gen .
, Desmophyllum , Flabellum , Sphenotrochus , Smilotrochus , etc. ) .
* The double row 0\#163 ; " dimples " in the wall of Turbinolia , between any two neighbouring septa and the additional costre between the double series , is an interesting morphological feature , since it is one also presented by the cranenchymatous types , Tiwbinaria and Madrepora .
A Miocene Australian ( solitary ) genus T1\#183 ; ematot1'ochus has pores in its theca in the 11ame relative position as the " dimples " 0\#163 ; Turbinoli1i and the " pores " 0\#163 ; Turbinaria .
The few colo11y-building Turbinolid genera are said to have buds which " ascend in growth and join others by their epitheca .
" This is precisely the process which the study of structural features led me to accept in explanation of the cranencbyme in Oculinidre , Pocilloporidoo , and M adreporidoo .
The calyces in all those families show closest relationship with the Turbinohdoo ( cf. chap .
viii .
) .
Tribe :-Trochosmiliacew .
Basal deposit present as dissepiments .
Rootlets never occur ( gen .
, Trochosmilia , Pleurosmilia , Placosmilia1 Cfrcophyllia , etc. ) .
Family :-Amphiastrwidw , * OGILVIE .
Mainly colony-building , seldom solitary corals .
Colonies massive or bushy , rarely meandroid .
The whole colony as well as the individual corallites covered wi~h strong epitheca .
Septa never pinnately inserted in mature calyces ; well developed or represented by horizontally-placed spines .
No fossula present ; but bilateral symmetry of the calyx always well marked by means of a main septum .
True theca present .
Base of the calyx tabulate or dissepimental , the dissepiments in the peripheral part are usually placed at a high angle and give rise to a false " inner " wal1 .
Pinnate insertion of the septa has been observed in the young calyces of a few examined cases ( cf p. 291 ) .
And frequently the septa are pinnately arranged in the calyx with reference to a main septum , so that a group of longer septa is present posteriorly , and short septa are present in the anterior part of the calyx .
In this family I would include the Pala:iozoic genus columnaria , which has hitherto been so variously placed in classification .
The AmphiaRtra:iida:i may be regarded as colony-building types directly descended from typical Zaphrentid genera .
The families Zaphrentida:i and Arnphiastra:iida:i are the only two in which bilateral symmetry of the septa is a constant systematic feature .
In all other families its occurrence is occasional .
In the same colony often one member may have it and another not .
Bilaterality is atavistic in such cases .
It may be inherited and constant in Seriatopora , certain genera of Madrepora , etc. , yet the hexameral symmetry of related types and genera is suffi~ient to show that bilaterality is giving place to radial symmetry even in those cases .
The lamellar columella in the Trochosmiliacere is undoubtedly the representative of * A1nphiastrreidce :--This newly-erected family will be found fully described in my " Monograph of the Stmmberg Upper Jurassic Corals " ( ' Palreont .
Mitth .
, ' Stut.tgart , 1896 ) .
One of its most interesting characters is the primitive nature of its modes of budding .
The most usual is so-c::iJled " tabular " and " endothecal " budding .
Or an incomplete inner wall is frequently formed within the calyx , and buds develop in the cut-off corners of the calyces .
In other words an " edge-zone " is formed within the calyx during periods of rapid vegetative budding .
Again the contraction of the calyx may give rise to one bud centrally ; :rna .
Iogous contraction of the mother calyx may still be observed in a recent Turbiuolid , Schizocyathus .
SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
335 a " main " septum ; in Jurassic species it is as a matter of fact a long septum .
Yet we see how it gradually becomes less developed , gives place in recent genera to a pseudo-columella , or is rudimentary .
Family :-Stylinid(13 , KLz .
Colonial corals , massive , astrreiform , or incrust.ing .
Septa never pinnately inserted and no fossulre present .
Septal symmetry always radial , septa alternating in length and developed strictly in cycles ( 4 or 8 , 5 or 10 , 6 or 12 = primary cycle ) .
The septa leave a central space free , occupied by tabulre or a columella .
Dissepiments present peripherally and throughout the cmnenchyme .
True outer theca sometimes , inner pseudotheca , between calyx and cmnenchyme always present ; cmnenchyme light , with costre strongest at the calycinal rims .
Epitheca well developed on the base of the colony and frequently encircling the individual corallites .
The genus Stylina and its allies were classified by EDWARDS and HAIME in the sub-family Eusmilinre , therefore in the family of Astrreidre .
I have erected these genera into an independent family for reasons referred to above ( pp. 157-160 , 221 ) but only di'scussed fully in my " Monograph of the Upper Jurassic Stramberg Corals .
" In that work I compared them with some Palreozoic genera , especial _y with the genus Decaphyllum , which I now include with the Stylinidre .
There too I noted the many signs of a common ancestry which were displayed by the Jurassic representatives of the Stylinidre and Amphiastrreidre .
The latter , however , still kept a certain old-fashioned look and sameness , while the Stylinidre , with their small star-like calyces , inner septal walls , outer costal and vesicular area , and infinite variety of colonial habit , were clearly .
in the forefront of that great influx of cmnenchymatous types which signalized post-Jurassic time .
Family :-Oculinid(13 , E. H. Colonial corals branching , massive or incrusting .
Cmnenchyme dense , compact ; fine papillary ridges are usually present on the surface , more rarely it is smooth .
Costre are distinct at the rims of calyces .
The calyces extend some distance above the surface of the cmnenchyme .
Septa well developed , few or numerous , bilaterally or radially symmetrical , Pali occasionally present .
Columella or pseudocolumella always present .
Tabulre or dissepiments may be sparsely developed .
Theca strongly developed , compact The calycinal cavity is greatly diminished by secondary thickening of all the skeletal parts .
Rudimentary epitheca apparent at the base of the colony .
The mesenteries of the polyp are continued into an edge-zone at the lip of the calyx .
The edge-zones of adjacent polyps are , however , rarely brought into direct contact , they become continuous with the common crenosarc of the colony .
A comparison of the fine microscopic structure of the calyces of Lophohelia and A mphihelia with typical Turbinolidre is . .
sufficient to prove the identity even in minute details .
The earliest Oculinids were the Enallohelias of Jurassic age , and these had typically a well marked main septum in their calyces and smooth septa , the cyclical number varying in number , 6 , 8 , 10 , 12 .
The calyces were definitely oriented and arranged in spiral or straight series on a dendroid stem .
Recent Oculinidre are radially symmetrical , and the calyces are approaching more and more to the Astrreid type , with irregular cyclical development and exsert serrated septa .
The ancestral forms of Oculinids are the " Lobophyllias " of QuENSTEDT , forms in which the mother polyp gave off whorls of buds at regular intervals .
The individuals of the whorl might one and all remain short , but retain their original connection .
with one another by means of the edge-zone of the mother polyp , or they might grow variously as branches .
Family :-Pocilloporid(E , E. H. , .
, Colonial corals branching , massive , or incrusting , with small calyces either sunk in the ccenenchyme or extending very little above it .
Crenenchyme compact or sub-compact , with echinulate surface .
Septa , few in number , arranged with bilateral or radial Bymmetry ; often represented by a series of horizontal septa !
spines .
Tabulre or compact laminate deposit , seldom dissepiments .
Theca compact .
The mesenteries of the polyp are not continued over the lip of the theca into an " edge-zone .
" * I have included the common genus Stylophora and its allies ln this family of Pocilloporidre ( cf. Phylogenetic scheme and pp. 308 , 313 ) , SYSTEMATJC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
337 Sub-family.-Pocilloporince .
Calyces round , irregularly distributed .
Ccenenchyme compact .
Tabulce present in the corallites ; solid laminar deposit may occur near the surface .
Septa often represented by spines .
Two septa , opposite one another , usually better developed than the rest .
Six of the mesenteries bear filaments , two of which arc reproductive , four rudimentary .
Sub-family.-Seriatoporince .
Calyces oblong , regularly placed in series along the branches .
Ccenenchyme compact .
Solid laminar deposit ; forms the base of corallites ; tabulm rare .
Septa usually twelve in number , two forming a mid-line coinciding with the long faxis of the calyx ; certain septa are rudimentary , associated with the presence of a pair of fossulm on opposite sides of the mid-line .
Only two of the mesenteries bear reproductive filaments , an opposite pair at right angles to the septal mid-line ; these filaments occupy the fossulm .
Sub-family .
--Stylophorince .
Calcyces round , irregularly placed .
No tabulm , occasional dissepiments present .
Ccenenchyme sub-compact ; unfilled spaces are left in rapidly-growing parts of the colony .
Six radially-arranged septa well developed ; a varying number of rudimentary septa .
Columella present as a simple axial style coalescent with the ends of the six primary septa .
Family : Madreporidce .
DANA .
Colonial corals , branching , palmate , or incrusting , frequently forming reefgrowths .
Calyces sunk or little prominent , small and with few septa .
Theca porous , ccenenchyme canaliculate , the canals communicating with the calyces by means of the thecal pores .
Septa bilaterally or radially arranged , compact ; sometimes represented by a series of horizontal spines projecting inwards from the wall .
Basal deposit in the calyces may be absent ; when present , it is in the form of tabulm or solid laminate thickening .
The mesenteries of the polyp are not continued over the lip of the theca into an " edge-zone .
" Sub-family.-J.v.ladreporinw .
Mainly branching colonies , with a radially symmetrical end-calyx .
Other calyces are usually oriented with resped to the axis of the branch , and vary in shape and in development .
They project slightly from the surface of the branch .
No basal deposit present in ca1yces .
Septa six , radially arranged ; two are usually specially developed , and form a calycinal mid-line ; a varying number of septa are present in rudimentary form .
Sub-family.-Montiporinw .
Colonies , dendroid , massive , or incrusting .
Calyces are placed in the surface depressions of the ccenenchyme .
No endcalyx is present in branching types .
No basal deposit in calyces .
Septa 6 or 12 ; seldom more than 6 are well developed , and all may be represented by series of spines .
Sub-family .
--A lveoporinw .
Massive colonies , with closely-placed calyces .
Thecal pores large ; ccenenchymal canals , little compacted .
Calyces sunk in ccenenchyrne ; cavities deep , tabulrn present in lower parts .
Septa 6 or 12 , all represented by series of spines .
Two opposite spines meet occasionally , as a calicinal mid-line .
More often the spines of several series coalesce and wincl round one another centrally , forming a pseudo-columella .
Su b-farnily .-Turbinarinw .
Colonies , low , massive , or dendroid .
Calyces at irregular distances , rising slightly above the surface of the ccenenchyme , rarely sunken .
Septa 6 to 30 , either well developed or rudimentary , but never represented by series of spines .
A calicinal mid-line may be marked out by the elongation of one of the principal septa , or its coalescence with the septum opposite .
Pali often present , formed by coalescence of the inner ends of neighbouring couples of septa .
The pali frequently have winding upward growth , and give the appearance of a large colnmella , present or absent , SYSTEMATIC STUDY Ol !
- , THE MADREPOR .
A.RIAN TYPES OF CORALS .
339 Family : Poriticlro .
DANA ( pro part ) .
Reef~growing , incrusting , or massive , rarely dendroid colonies , with solid welldeveloped epithecal basal-plate .
Calyces unequaJly distributed , in contact directly at their periphery , or \#183 ; with a development of ccenenchyme .
The whole skeleton is bui]t up on a system of upright trabeculre and stout cross bars , so that , in longitudinal section , it is scarcely possible to distinguish calyces .
Several trabecuhe may be partially compacted to form 5- 12 primary septa , or individual broad trabeculre may represent septa ; frequently one or more trabeculre are free centrally , as styliform columella and pali .
The ccenenchyme is not echinulate at the surface , but presents costate ridges , confluent or non-confluent .
Dissepiments richly developed throughout the skeleton ; tabulre may be present in the calyces .
The Palreozoic family of Theciidre ( Thecia , Protarroa , Stylarroa , Coccoseris ) seems to be the ancestral types of our recent reef-builders .
An interesting transitional genus is Astrroomorpha , a Triassic genus , whose thick septa and calicinal features show more the character of the Palreozoic types , yet the fine structure proves its affinity with the recent Porites ( cf pp. 305-307 ) .
Longitudinal sections of Astrroomorpha and Porites are essentially similar , while the intimate connection of the calyces at their periphery is characteristic in both .
The Theciidre and several allied forms of Palreozoic age represent , to judge from their calicinal structures , early colony-building relatives of the solitary Zaphrentidre , where tlrn epitheca of young broods came laterally into contact .
Primitive Madreporaria had no theca , hence tangential contact readily gave rise to a so , -called ccenenchymatous colony ( cf pp. 315-317)-whether porous or compact is a further point of modification .
The genus Astrroomorpha may be held provisiorrnlly as the representative of a sub-family , Astrreomorphinre .
The Palreozoic genera of the Theciidre require detailed examination before they can be rightly placed .
The Cretaceous and recent typ8s of the Poritidre may be provisionally treated as a sub-family , Poritinre .
The typical reef-building members of the Poritidre belong to this sub-family Poritinre , and I shall therefore enumerate here the characteristic structural features observed in Porites .
I am not in a position to thoroughly review the family of Poritidre or to discuss the possible affinities of Palreozoic types with Favositidre at present ; I limited microscopic investigation to Mesozoic and younger genera .
In the following list of features I shall indicate the superficial resemblances of Porites with other recent families of Madreporaria .
340 DR. MARIA M. OGILVrn ON THE MICROSCOPIC AND The mic1\#183 ; osc0ziic sl1\#183 ; ucture of the septa and pali in Porites is unlike that of any other Madreporarian type hitherto examined in detail .
All skeletal parts of Porites have the same microscopic structure , so that its most characteristic feature is the sim .
, ila1 \#183 ; ity in s1 ; 1 , e cincl d evelopment of each _growth-segment however it may happen to be placed in the skeleton vertically , obliquel3 , , or horizontally ( so-called " synapticul::e " ) .
The ccdyccs are radially symmetrieal , but the number of septa in the primary cycle is highly variable : 5 , 6 , 7 , 8 , 10 , 12 .
A sty]iform columella i8 usually present , surrounded by pa. Ii as in the Oculinid.=e and Trochocyathime .
The sepla are madu up of vertical or slightly sl:rnting trabecul:-e ; when spaces are left unfilled , the surface appearance of the septum is mnch like that of a porous Eupsammid septum .
The individual trabecuhe a.re often directed locally inwg , rds towards the axial part of the calyx , but there is no cwea of divergence at the periphe1\#183 ; y ( cf , pp. 222 , 22:3 ) of the calyx as Eupsarnrnid.=e , Astr.=eidrn , and Fungidm .
Moreover , the microscopic structure of the individual trabeculm in Poritinm is different from trabecular structure in these three families .
Numerous fine dissepiments are present .
The calycinal structures radiate outwards , considered from the centre , in deltoid fashion , and this feature is fuudamental in many genera belonging to the Trochoc_ yathinm and Enpsammidm .
The so-called wall of the Poritinm is a zig-zag ring of thecn , l deposit like tha , t in most of the Trochocyathinm .
The porosity of the " wall " is based on structural relations quite different from those in Eupsarnmia , but homologous with those in the Madreporidm .
Each new layer of growth is indicated by fresh epithecal rims round the cora.llurn , as is the case in encrusting types of Oculinidre .
Spongiomorphidce .
For the diagnosis of this family I refer to its founder Professor FRECH ( " Triassic Corals , " 'Palmontogr .
, ' 1890-91 ) .
I have had occasion to compare it with the Poritidm above ( p. 307 ) , and think it would be best regarded a : : ; a Subfamily , of equal rank with the Poritinm and included under the family of Poritidm .
Family :-Cyathophyllidce .
E. Fl .
Solitary and colonia.1 corals ; colonies massive , bushy , or branching-never meandroid .
Base of the calyx distinctly subdivided into a central tabulate or columellar area and a broad peripheral area with dissepiments .
Septa , arranged in the inature calyx usually with radial symmetry , indicatiol18 of bilateral symmetry seldom ; there may or may not be pinnate insertion of septa ; fossulm small or obsolete ; septa \#183 ; well developed , never repre\#183 ; sented by a series of horizontally placed " wall " spines .
Theca usually present , coalescent with the epitheca .
SYSTEMATIC STUDY OF THJ\lt ; J M.ADREPORARIAN TYPES OF CORALS .
341 Pseudothecal septal thickening occurs frequently , an\lt ; l at some distance from the true theca , forming an inner wall .
The P~lrnozoic types , as they have hitherto been classified , cannot be said to have been referred to families or even genera according to the fine microscopic structure of their septa .
Anrl it might , therefore , seem at first sight , as if the method I have advanced under the new sub-orders ( pp. 329 , 330 ) wou1d bent complete variance with all existing grouping of corals .
Curiously enough , it makes least change of aU amongst the Palrnozoic types .
For the microscopic structw\#183 ; e of the t1'aueculce and the relative a1Tnn , c ; ement of the trabeculce in the septal plane are both architectural features dependent upon the breadth of the oral disc and the required strength of the septa for the support of the corresponding aboral breadth .
The same polypal conditions have determirn~d the nature of the skeletal base .
Hence a classificatory arrangement of Palrnozoic corals , such as Professor NICHOLSON'S , where the chief weight is laid upon the particular relations of the calicinal base , is bound to coincide in its main features with the system I now follow .
Either classificatory basis can indeed merely act as a corrective on the other in the examination of the primitive Palrnozoic species and genera .
As I have already said , I use Zaphrentidce practically with the limits of Zaphrentoiclea , NrCI-IOLSON , similarly Cyathophyllidce as the equivalent in tlrn main of Cyathophylloidea , NICHOLSON ; but Cy : : ; tiphyllidce only with the special meaning which is attached to thefaniily of the same name by Professor NICHOLSON .
I have seen it stated , in several lately-published papers on Palrnozoic corals , that the genus Heliophyllum may be incorporated with Cyathophyllum .
It is said that the carinrn of Heliophyllum are not sufficient reason for a separation .
I place considerable importance on the carinre as a starting-point of a variation within the Cyathophyllidm which may have been the origin of the main differences in the structures of recent Astrrnids and Fungidrn .
But .
, in addition to this feature , there is the fact that in H.eliophyllum the pinnate insertion of the septa is strongly marked throughout growth , whereas in the typical species of Cyathophyllum , which ought to be regarded as the standard of the genus , pinnr1.te insertion is not observed in the mature calyx .
NICHOLSON , on the other hand , made two distinct families---Cyathophyllidrn and Heliophyllidrn .
I would certainly be inclined to regard them as subfamilies .
In the Clisiophyllidrn ( N rcr-r .
) , in addition to the columellar area , a point may again be made of the invariable pinnate insertion of septa .
The following families have been already treated in full systematic detail in Part I. , and their characteristic features will be found sumrrw.rized in the Table ( p. 252 ) .
Short diagnoses may be repeated here .
342 DH , .
MARIA :M .
OGILVIE O)'f THE MICROSCOPIC AND Family :-Astrceidce .
E. H. ( emend .
OGILVIE ) .
Mainly coloui.al , seldom solitary , corals .
Colonies massive , branching , brn ; hy , incrusting , meandroid .
No tabulre are present centrally , dissepimeuts are distributed throughout , and form a compact calycinal base .
Septa never pinnately inserted ; always arranged with radial symmetry ; give 1\#183 ; ise frequently to a pseudocolumella centrally ; compact ; trabecular structure regular .
True theca only present in a few of the older types , coalescent with the epitheca .
Pseudothecal septal thickening almost always present in the interior of the calyx .
Epitheca surrounds simple forms , also the branches of bushy and dendroid colonies , but is limited to the base of massive and meandroid colonies .
The sub-families of Astrreiclre are given ~bove ( Part I. , pp. 1G3 , 164 ) .
I have frequently referred , in the course of this work , to the significance of the pseudotheca in this group of corals , and its c0rrelation with the retrograde state of the more primitive theca and epitheca round individual members of a colony .
On this depends the homology between the broad outer disc , or " \lt ; lissepimental zone , " of Cyathophyllidre and the costate , so-called extracalycinal , zone of Astrreidre- both representing , in the skeleton of a massive astrreiform colony , the edge-zone , or " Randplatte , " which bps over the edge of the wall in the free branches of bushy colonies , or in a solitary coral .
Although I retain the long-familiar name of Astrreidre , the family is very differently defined from EDWARDS and HAil\rn 's original use of the name .
I have removed from it practically the whole of the sub-family Eusmi.linre , distributing these genera among the families Amphiastrreidre , Stylinidre , Turbinolidre , as given above ( pp. 159 , 162 ) .
Family :-Fungidce .
DANA ( emend .
OGILVIE ) .
Solitary and colonial corals .
Colonies massive , incrusting , palmate , meandroid , rarely dendroid .
Septa never pinnately inserted ; radially symmetrical in mature calyx .
Septa porous or compact , frequently coalescing , i.e. , the inner ends of the shorter septa coalesce with the side surfaces of longer septa ; trabecular structure regular .
Base of the calyx chiefly formed by synapticular bars ; dissepiments irregularly developed or absent .
No true theca is developed ; pseudotheca sometimes .
SYSTEl\1.ATIC STUDY OF THE MADREPOR .
ARI .
AN TYPES OF COR .
ALS .
343 Sub-family :--Funginro .
Low , solitary corals , or massive colonies , never meandroid .
Epitheca rudimentary , replaced in greater part by large basal spines .
Both " true " synapticuhe and pseudosynapticulm ( cf p. 198 ) richly developed .
Dissepiments well developed , rarely absent .
Septa usually compact , in some cases those of the higher cycles may L8 porous Coalescence of lateral surfaces largely developed .
Sub-family :-Thamnastrroinro ( cf pp. 186-190 ) .
Solitary corals of discoid or turbinate form .
Colonies massive , calyces confluent , frequently meandroid .
Epitheca well developed , forming a compact basal surface for the corallum .
No true synapticulro occur .
Dissepiments well developed .
Pseu , dosynapticulro developed in horizontal rows representing successive growth-periods ; the rows extend throughout the whole length of the interseptal and intercostal loculi .
Septa regularly fenestrated between individual trabecular parts .
The spaces become sometimes solidified by secondary deposit .
The divergenr , e of the septal trabecula : , is in a single fan-system .
Sub-family :-Lophoserinro .
Solitary corals , or low , foliaceous , and frequently meandroid colonie8 .
Epitheca , when present , is compact and not spinate .
Pseudosynapticulre richly developed at and near periphery of calyces , but not forming complete interseptal and intercostal series .
True synapticulre occasionally occur in the formation of a pseudotheca .
Dissepiments few or absent .
Septa usually compact ; when porous , the pores are not regularly dif:tributed between adjacent trabecnlre , they may be in the course of the trabeculrn themselves .
The divergence of the trabeculre takes place , as a rule , in several fan-systems .
Family :-Enpsammidro .
E. H. ( emend .
OGILVIE ) .
Solitary and colonial corals .
Colonies branching , massive , incrusting , bushy or stolonal .
Septa never pinnately inserted , radially symmetrical in calyx , rarely showing traces of bilateral symmetry .
344 DR. MARIA M. OGILVIE ON 'l'HE MICROSCOPIC AND Septa irregularly toothed and granulated , of irregular trabeculate structure , the " trabecular parts " often alternating in position and leaving pores of various sizes and sha , pes in the septal plane .
Lateral coalescence of septal surfaces at their inner ends usually pronounced .
No true theca is present .
Costre highly developed .
A. considerable area is marked out peripherally as a porous psendotheca ; the structure of the base here variefl from a rich vesicular deposit to a complicated synapticular network , and is associated with mural \#183 ; costrn ( cf " camenchymate types , " p. 316 ) .
Basal deposit within the calyx extremely various ; synapticulrn , pseudosynapticulrn , dissepiments , more rarely tabulm or solid laminar deposit .
A large amount of secondary thickening takes place on septa , pseudotheca , and base .
Epitheca , when present , is smooth or slightly rugose , never spined .
In this family I now include for the first time a number of Mesozoic genera which have hitherto had a somewhat uncertain systematic position .
The characteristic feature of loose and irregular trabeculate septal structure shov , 's itself throughout all .
The most important of these are the Jurassic genera-Haplarcea ( sub-fc1 , m. Pseucloastrreime irregulares , PRATZ ; fam .
Fungidrn ) ; Epistreptophyllum ( sub-fam .
Epistreptophyllime , ZITTEL ; fam .
Astrreidre ) ; Dermosmilici ( Dermosmilidrn , KoBY ) ; and the Triassic genera , Stylophyllum and Stylophyllopsis ( sub-fam .
Stylophyllinrn , FRECH ; fam .
Astrrnidre ) .
Family:- Cystiphyllid(J3 , E. H. Family :-Arch(J3ocyathidce , w ALCOTT .
These two Palreozoic families may be taken from standard text books SYSTEMATIC STUDY OF THE MADREPORARIAN TYPES OF CORALS .
345 .
LIST OF ABBREVIATIONS USED IN LETTERING THE FIGURES .
a. Aboral body-wall of polyp .
o. Oral body-wall of the polyp .
a.cl .
Area of divergence .
org .
Organic cell-remnant .
c. Costa .
p. Septal pore .
c.c. Centre of calcification .
pa. Pal us .
c.e. Costal edge .
ps.col .
Pseudo-columella .
en .
Crenenchyme .
ps.syn .
Pseudo-synapticula .
Gn.p .
Crenenchymal pore .
ps.th .
Pseudo-theca .
co .
Coalescence of septal surfaces .
r. Septal ridge .
col .
Columella .
s. Septum .
cor .
Coral lite .
SC .
Scale ( calcified calicoblast ) .
cs .
Crenosarc .
s.e. Septal edge .
c.tr .
Calcification centre in trabecula .
sec. Secondary fibrous deposit .
c.th .
Calcification centre in theca .
sp. Spiniform tooth .
d. Dissepiment .
sr .
Serra .
d.l. " Dark line " { Primarstreif ) .
S.S. Septal spine .
e. Edge-zone ( Randplatte ) .
str .
Stria on septal surface .
ep .
Epitheca .
syn .
Synapticula .
f. Fascicle of \#163 ; bro-crystals .
tab .
Tabula .
f.c. Fibro-crystals .
th .
Theca ( eutheca ) .
g.c. Growth-curve .
th.p .
Thecal pore .
g.l. Growth-lamella .
tr .
Trabecula ( simp .
= simple ; comp .
gr. Granulation .
= compound ) , g.s. Growth-segment of septum .
\#183 ; tr.ax .
Ideal axis of trabecula .
h. Hyphre ( Fungi , etc. ) .
tr-.p .
Trabecular pore .

