[ 81 ] III .
On the Structure and Affinities of Fossil from the Palceozoic Rochs.\#151 ; III .
On Medullosa anglica , a New Representative of the Cycadofilices .
By I ) .
H. Scott , M.A. , Ph. D. , F.R.S. , Honorary Keeper of the Jodrell Laboratory , Royal Gardens , Kew .
Received December 21 , 1898 , \#151 ; Read January 26 , 1899 .
[ Plates 5-13 .
] Contents .
1 .
Introduction ... ... ... ... ... . .
2 .
General structure ... ... ... ... 3 .
The steles ... ... ... ... ... ... .
4 .
The leaf-trace bundles . . .
5 .
The ground tissue and periderm 6 .
Accessory vascular strands . .
7 .
The leaves ... ... ... ... ... ... .
8 .
The roots ... ... ... ... ... ... . .
9 .
Summary and diagnosis . . .
10 .
General conclusions ... .
Explanation of plates . . .
Diagram and description . .
Page 1 .
Introduction .
For some years past the existence of a group of fossil plants , combining in their organization certain characters of Ferns and Cycads , has been recognized by several Palseobotanists.* Potonie , in his new " Lehrbuch der Pflanzenpalseontologie/ 'f has proposed the convenient name Cycadofilices for this grouji , in which he includes the genera Nceggerathia , Medullosa , Cladoxylon , and * See , for example , Williamson , " Organization of the Fossil Plants of the Coal Measures , " Part XIII , 'Phil .
Trans. , ' B , 178 , p. 299 , 1887 ; Solms-Laubach , 'Fossil Botany , ' 1887 , Engl .
Ed. , pp. 141 and 163 , and in many subsequent papers ; Williamson and Scott , " Further Observations on the Organization of the Fossil Plants of the Coal Measures , " Part III .
, 'Phil .
Trans. ' B , 186 , p. 769 , 1895 ; Scott , Address to Botanical Section of British Association , Liverpool , 1896 ; Seward , " A Contribution to our Knowledge of Lyginodendron , " 'Annals of Botany , ' March , 1897 .
t Leipzig , 1898 , p. 160 .
% Re-named Lyginopterisby Potonie .
VOL , CXCI.\#151 ; B. M 10.7.99 Protopitys .
Among these the genus Medullosa , with its allies Brongn .
, and Steloxylon , Solms , is by no means the least remarkable .
The large stems , reaching half a metre in diameter , have an extraordinarily complex -structure , containing a system of numerous vascular rings , of the most various dimensions and shapes , each ring growing in thickness by a cambium of its own .
The structure bears , at first sight , a striking resemblance to that of the anomalous Sapindacese , with which it was compared by Goeppert and Stenzel , # but more recent writerst have found a better analogy in the organization of a polystelic Fern-stem , each steal , in the case of the Medulloseae , having its own zone of secondary growth , a condition not known among recent plants .
The general character of the secondary tissues , and the anomalies presented by certain of the older stems , are strongly suggestive of Cycadaceae , though the ground plan of the structure is rather that of a Fern .
The investigations of Schenk , Solms-Laubach , and Weber and Sterzel , have left no doubt that these extraordinary stems bore , as their petioles , the fossils known under the generic name of Myeloxylon , Brongn.^ These leaf-stalks , which are often of huge dimensions , have , in essentials , the structure of those of Cycadaceae , but Benault , in his turn , has shown that they bore the foliage of Alethopteris and Neuropteris , genera of which the external characters are entirely Fern-like .
S The above remarks are only intended to indicate , in the briefest way , the remarkable combination of characters which the Medulloseae present .
Any further consideration of the existing literature on the group , and all discussion of its affinities , will be postponed to the end of the paper .
Although fragments of Myeloxylon are common enough among the fossil remains , with structure preserved , found in the calcareous nodules of the English Coal-measures , no specimen of Medullosa itself , has , so far as I am aware , as yet been recorded from this country .
This apparent absence of the stems , to which the leafstalks must have belonged , would be surprising , were we not familiar with the fragmentary character of such petrified specimens .
Becently , however , I have had the opportunity of investigating several fine specimens of the stems of a Medullosa , obtained by Mr. G. Wild and Mr. J. Lomax , * * * S * 'Die Medullosese , eine never Gruppe der fossilen Cycadeen , ' 'Palseontographica , ' vol. 28 , 1881 .
f Zeiller , ' Etudes des gites mineraux de la France , Bassin houiller et permien d'Autun et d'Epinac , Flore fossile , ' Part I. , p. 286,1890 ; Solms-Laubach , * iiber Medullosa LeucJcarti , ' ' Bot. Zeitung , ' 1897 , p. 182 .
| Schenk , ' iiber Medullosa elegans , ' Engler 's ' Botan. Jahrbiicher , ' vol. 3 , 1882 ; ' iiber Medullosa und Tubicaulis , ' ' Abhandl .
d. K. Sachs .
Gesellsch .
d. Wiss .
, Math. Phys. Cl. , ' vol. 15 , 1889 .
Solms-Laubach , iiber Medullosa Leuckarti , loc. cit. The original discovery appears to have been due to " VVeber ; See Weber and Sterzel , ' Beitriige z. Kenntniss der Medullosese , XIII .
Bericht der Naturwiss .
Gesellsch .
zu Chemnitz , ' 1896 .
Myeloxylon , Brongn .
, Stenzelia , Goepp .
, and Ren .
, are synonyms .
Cotta 's Medullosa elegans is , of course , a Myeloxylon .
See p. 111 .
S Renault , ' Course de Botanique Fossile , ' vol. 3 , p. 159 , 1883 .
Stur , however , inclined to regard these leaves as Cycadaceous .
OF FOSSIL PLANTS FROM THE PALAEOZOIC ROCKS .
from the colliery at Hough Hill , Staly bridge , a locality already well known to the student of British Carboniferous plants .
The late Professor W. C. Williamson , F.It .
S. , received a transverse section of one of these stems ( which we will call Specimen I. ) from Mr. Lomax , about the year 1892 , but it was never incorporated in his collection .
The slide , which is labelled , in Professor Williamson 's handwriting , " Very large , crushed , and a crushed Myelopteris , " came into my possession , with other rejected material , after his death , in 1895 .
The section was then a rough , unfinished one , and it is not surprising that Williamson had failed to recognise its importance .
On having it thinned and covered the true nature of the fossil became evident.# Subsequently ( in September , 1897 ) , Mr. Lomax sent me another section , evidently cut from the same specimen , and I then learnt that the block was still in his possession .
It appears that it had been originally discovered by Mr. Wild and himself , when collecting at Hough Hill .
Mr. Lomax sent me the principal piece of the block for macroscopic examination , and prepared from it a great number of admirable sections , in various directions .
From first to last not less than 100 sections , transverse and longitudinal , from this one specimen have been examined .
The investigation of this specimen was already far advanced , when Mr. Lomax , in June of this year ( 1898 ) , was so fortunate as to discover two more stems in material obtained by Mr. Wild from the same colliery .
One of these ( Specimen II .
) is the magnificent specimen , a portion of which is represented in its external aspect in Plate 9 .
The third specimen was peculiar in having entirely lost its external cortex .
Still more recently ( in September last ) yet another stem was found by Messrs. Lomax and Wild , once more in the Hough Hill material .
This ( Specimen IV .
) again shows the external characters very well , and is otherwise of importance .
Another fragment accompanying this was probably a detached piece of the same .
From all these specimens numerous preparations have been made for me by Mr. Lomax .
The skill and precision with which he has cut the necessary sections , often of large size and forming long consecutive series in definite directions , deserve the highest praise .
In addition to the four principal specimens just enumerated , I have to thank Mr. J. Butterworth , of Shaw , and Mr. G. Wild , for the loan of other specimens , which , though fragmentary , have had an important bearing on special points in the enquiry .
For the investigation of the roots , other material again has been brought into requisition .
The fossils in question are of very considerable interest , not only as being the first recorded British specimens of the genus , but also from their geological horizon , which is considerably more ancient than that of the continental Medullosete ; the latter are essentially plants of the Permian and Upper Coal-measures ; the British examples , This section is now No. 226 in my collection .
See Plate 5 , photograph 2 .
like the other calcareous vegetable remains of Lancashire and Yorkshire , are from the Ganister heads of the Lower Coal-measures , and are associated in the Hough Hill material , with such fossils as CalamostachysBinneyana , Rachiopteris Oldhamia , and R. hirsuta .
The structure of the specimens shows , as we shall see , that they represent a true Medullosa , differing however in certain important respects , and chiefly in the direction of greater simplicity , from any species previously described .
The significance of these differences , and the question of the relation of our fossil to Brongniart 's genus Colpoxylon , will be discussed at a later stage .
As it is necessary to distinguish the British form by a specific name , I propose to call it Medullosa anglica .
2 .
General Structure .
The main features of the organization of Medullosa anglica are best shown in Specimen II .
, on which the following general description is primarily based , while the anatomical details have been more especially studied in Specimen I. The latter , indeed , had been worked out before the other specimens were discovered .
The block containing Specimen II .
was originally about a foot long , and is traversed from end to end by the Medullosa .
The specimen was cut , in approximately transverse planes , into three pieces.* In the lowest of the three the stem is embedded in the matrix , but the whole of the upper two-thirds of the block has split in such a way as to expose the surface of the plant on one side .
( See Plate 9 , fig. 1 , which represents , in natural size , the whole exposed surface of the stem .
) Sections were cut from all the transverse planes , and a piece at the top of the specimen ( not shown in the figure ) was utilized for obtaining a series of longitudinal sections .
The transverse section shown in photograph l , f was taken from the top of the lowest piece , and consequently immediately below the level at which the surface of the stem becomes exposed .
The crack shown in the transverse section represents the line along which , in the upper portion of the block , the split took place , and so enables us to see exactly what part of the specimen is exposed by it .
The stem , where the section shown in photograph 1 passes through it , is somewhat flattened , having a diameter of 10'5 X 3*7 centims. It has a roughly triangular sectional outline .
In a position corresponding to the three angles are three large and prominent masses of cortical tissue , together occupying almost the whole periphery of the stem .
These masses clearly represent the bases of leaves , still in continuity with the stem .
These leaf-bases have in all respects the typical Myeloxylon structure .
The hypoderma , composed of several series of sclerenchymatous strands , with interspersed * A short piece was also removed from the top ; see below .
t The explanatory diagram on p. 126 should be compared throughout with photograph 1 , to which it serves as a key .
OF FOSSIL PLANTS FROM THE PALAEOZOIC ROCKS .
gum-canals , is characteristic of the type named by Renault M. Landriotii* The inner parenchyma is traversed by numerous gum-canals , and by collateral vascular bundles , agreeing exactly , as will be shown below , with those already so well known in Myeloxylon .
Two of the leaf-bases ( a and , photograph 1 ) are oidy marked off from the interior of the stem by the internal periderm , which extends all round the stem .
The third leaf-base , however ( c ) , is delimited on its inner side by an interrupted sclerenchymatous band , similar to , though not so thick as the hypoderma .
This leaf , then , should be the first of the three , as traced upwards , to become free from the stem .
An examination of the exposed surface of the specimen shows that this is the case .
Up to about the line z the surface c ( Plate 9 , fig. 1 ) is somewhat concave , and as is shown by the position of the crack in the transverse section ( Plate 5 , photograph 1 ) , represents the inner surface of this leaf-base .
At about z the striae bend off to the left , and no doubt it was in this region that the petiole of c became free from the stem .
It is lost , for the outer part of the block had here split away .
Above z the surface becomes slightly convex , and , as shown by comparison with the transverse section at the level 3 , here represents part of the externa surface of a new leaf-base , appearing in the angle between b and c. At the level of the section shown in photograph 1 , this leaf-base ( be ) is only represented by a thickening of the hypoderma .
It extends to the top of the specimen ( see Plate 9 , fig. 1 ) .
The leaf-base , remains in connection with the stem all through , though marked off from it in the upper sections by an internal band of sclerenchyma .
In the upper third of the specimen this leaf-base is much flattened , and forms the greater part of the exposed outer surface .
The leaf-base b ( which , of course , is not seen in surface view , as it lies on the embedded side of the specimen ) is marked off by internal sclerenchyma at the level 3 , and at the top of the specimen has become free , but is here only represented by some shreds of tissue .
Although large free petioles of Myeloxylon occur in the surrounding matrix , they have not so far been followed to their connection with the stem in this specimen , but in the case of one free petiole of Specimen I. the connection has been traced .
The petioles usually appear to have been torn from their bases ; there is no sign of a regular abscission .
At the level 3 , in Specimen II .
, another new leaf-base ( ab ) makes its appearance between a and b. It will be seen from this specimen that the leaves were spirally arranged ; the facts agree well with a 2/ 5 phyllotaxis ( see diagram , p. 12G ) .
The leaf-bases were evidently adherent to the stem for a long distance .
That marked a can be traced with certainty for more than five inches before it detaches itself from the stem .
In this length only two leaves become free , namely , c , in the lower part , and b near the top .
The vertical distance between successive leaves may he roughly estimated at about four inches .
Judging from this specimen , there was but little free surface of the stem between the prominent bases of the leaves .
The other specimens are less favourable for * Renault , 1 Etude du genre Myelopteris'Mem .
pres , par divers savants a l'Acad .
de Paris , vol. No. 10 , 1875 ; also ' Course de Botanique Fossile , ' vol. 3 , p. 159 ( petiole of Alethopteris ) , 1883 .
deciding this point .
Specimen I. has lost a great part of its outer cortex and leaf-bases .
The portion of hypoderma at oc ( Plate 10 , fig. 2 , on the left ) might be regarded as belonging to the stem , but may also be explained as part of a flattened leaf-base .
Specimen III .
has lost everything outside the periderm .
Specimen IV .
shows the outer surface well , but is very much flattened .
It agrees , however , in general appearance and structure , with Specimen II .
, and the external surface , so far as can be seen , belongs almost entirely to the leaf-bases .
The habit of the stem , clothed with the long , almost vertical , overlapping leaf-bases , may have been not unlike that of some of the Tree-ferns , such as Alsophila In the narrow furrows between the leaf-bases the hypoderma appears to have been interrupted , and it was no doubt from these furrows that the adventitious roots sprang , as will be shown below .
So much for the more external characters of our fossil .
We have so far learnt that the stem was densely clothed by the spirally disposed , adherent bases of its petioles .
The interior tissue is surrounded by a broad ring of periderm ( see photograph 1 ) , which is strongly developed in all the specimens except Specimen I. , in which it only forms a thin layer .
On other grounds also this stem seems to have been the youngest of the four .
In Specimen III .
, all the outer tissues , including the bases of the leaves , have disappeared , leaving the periderm exposed .
This may have been the result of a natural casting-off of the outer bark with advancing age , or may have been due merely to post-mortem maceration .
This question we will consider later .
The periderm cut off something more than the leaf-bases , for between the periderm and the internal sclerenchyma of the leaf-base a band of cortical tissue , containing vascular bundles , is present .
We now come to the most important feature in the stem\#151 ; the structure of its vascular system .
As a rule every transverse section of the stem shows three distinct masses of wood and bast , which , as we shall see , represent so many separate steles .
This is the case throughout Specimens II .
and IV .
, so far as observed .
All the sections of Specimen III .
are incomplete ; two steles only are usually shown , but there are fragments of a third .
In Specimen I. there are three , but one of these divides into two parts for a short distance ( see Plate 5 , photograph 4 ) , reuniting above and below .
The three steles are shown with great clearness in photograph 1 , from Specimen II .
, and in photograph 5 , from Specimen I. ( see also Plate 10 , fig. 2 ) .
Each steal always has an elongated sectional form , corresponding to that of the shorter " plate-rings " of Goeppert and Stenzel , but it often shows irregular curvatures and protrusions , thus approaching the " snake-ring " form .
The shape of the steles , in spite of frequent fractures , does not seem to have been essentially altered by pressure , at least in the better-preserved specimens , I. and II .
On the other hand , they have no doubt often suffered considerable displacement from their natural positions .
The whole system of steles must have formed collectively a somewhat triangular figure in the normal condition , as is still clearly shown in Specimen II .
( photograph 1 and diagram , p. 126 ) .
OF FOSSIL PLANTS FROM THE PALAEOZOIC ROCKS .
There is an entire absence of the small " star-rings " ( slender cylindrical steles ) so characteristic of other species of Medullosa .
In this respect M. anglica resembles the genus Colpoxylon of Brongniart .
A most satisfactory feature of our specimens is the very perfect preservation of the internal tissue of the steles , a part which in other Medullosese is so often damaged and obscure .
The whole interior of each steal is occupied by a mass of tracheides , forming confluent groups , interspersed with thin-walled conjunctive parenchyma ( see Plate 6 , photographs 5-9 ) ; usually the tracheides are much more abundant than the parenchyma , and even where the latter is more largely developed it never forms a coherent medulla , so that the expression " partial pith , " so often loosely applied to the interior of the " rings " in Medullosese , is wholly inappropriate here .
The whole of the internal mass constitutes the primary wood ; the spiral elements form scattered groups near its periphery .
Surrounding the primary xylem is a zone of secondary wood , of variable thickness , amounting in some places , as in parts of Specimen II .
, to more than forty elements in a radial series .
Beyond the wood we come to the phloem , which is best preserved in Specimen I. , though even here only at certain places .
It is evident that in this fossil we have before us a stem , fully comparable to that of many Ferns , except that here each steal is surrounded by its own zone of secondary wood and bast .
The structure of any one steal in Meclullosa , is almost identical with that of the single central cylinder in the stem of So far as the organization of the stem is concerned , it would be quite admissible to describe this form of Medullosa simply as a polystelic The comparison is not new , f but the relatively simple structure of the English species brings out the relation between the two genera far more clearly than has previously been possible .
From the steles , leaf-trace strands are given off , which at their base are of large size , and probably of concentric structure , but which , in passing through the cortex , break up into smaller collateral bundles of the type ( see Plate 6 , photographs 6-9 ; Plate 7 , photographs 14 and 15 ) .
These bundles pass out in great numbers into the massive bases of the leaves ( Plate 5 , photograph 3 ) .
The stems bore , between the leaf-bases , numerous adventitious roots , usually triarch in structure ( Plate 8 , photograph 19 ; and Plate 13 , fig. 20 ) , and characterized by a great development of periderm .
Young and old roots and rootlets have been investigated , both free and in connection with the stem , and will be fully described below ( p. 102 ) .
The preliminary survey which we have now taken leads to three main conclusions : ( 1 ) that our fossil was a polystelic stem , with secondary tissues developed around * * Cf .
Williamson , " Organization of the Fossil Plants of the Coal-Measures , " Part IV .
, 'Phil .
Trans. , ' 1873 , PI .
28 , fig. 30 ; Part XIII .
, 'Phil .
Trans. , ' B , 1887 , PI .
21 , fig. 1 .
t See Zeiller , " Etudes des gites mineraux de la France ; Bassin houiller et permicn d'Autun et d'Epinac ; Flore Fossile , " Pt. 1 , 1890 , p. 286 ; Solms-Laubach , " fiber Medullosa ; " ' Bot. Zeitung , ' 1897 , p. 182 .
each steal , ( 2 ) that this stem bore leaves with petioles like those of Landriotii , and ( 3 ) that it gave rise to triarch adventitious roots .
We will now proceed to examine the various structures more in detail , and then to consider the facts observed in their bearing on the affinities of our plant and of the Medullosese generally .
3 .
The Steles .
The individual steles , as seen in a transverse section of the stem , vary much in dimensions , reaching from 2 to 3 centims. in their greater , and from 6 to 10 millims. in their lesser diameter .
Their form is sometimes fairly regular , as in one of the steles represented in Plate 6 , photograph 5 ; while in others , as shown in the same preparation , the transverse section is irregularly lohed .
The whole interior of each steal , as already stated , is occupied by the primary wood , which consists of elements of two kinds\#151 ; tracheides and parenchymatous cells .
The arrangement and relative proportions of the two constituents vary much in different steles , and in different parts of the same .
At some places , the tracheides form a continuous system over a considerable area , only interrupted by scattered strands of parenchyma ( Plate 6 , photograph 8 ) .
In other cases the tracheides are arranged in definite groups , separated from one another by a network of parenchyma ( photograph 6 ) , while sometimes the cellular tissue predominates , and is traversed by comparatively few tracheides , isolated , or in small groups.* This is shown in Specimen I. , in a bifurcating steal ( photograph 4 ) , and has also been observed elsewhere ( Plate 10 , fig. 4 ) .
In no case is there any differentiation of a true " partial pith , " or even any marked increase in the amount of parenchyma towards the middle of the steal .
The great majority of the primary ligneous elements are pitted tracheides of large size ( averaging about T5 millim. in diameter ) ( Plate 7 , photograph 12 ) .
That they are really tracheides and not vessels , is rendered probable by their pointed , tapering ends , and by the absence of any appearance of perforation in their terminal walls .
This is a point , however , on which it is rarely possible to obtain conclusive evidence in dealing with fossil plants .
The pits on the tracheal walls are very numerous , and are arranged in transverse or inclined series ( see Plate 10 , fig. 5 ) .
They have borders like those of the secondary tracheides , to be described below , but the latter are usually better preserved and more favourable for the observation of details .
Such pitted tracheides extend throughout the whole internal region of the primary wood , and at many points also reach its outer limit , bordering on the secondary zone .
In the peripheral primary wood , however , other strands of tracheides occur , which have a different structure .
These peripheral strands are not always easy to recognize in transverse * Precisely similar variations occur in HeterangiumGrievii .
See Williamson and Scott , " Further Observations on the Organization of the Fossil Plants of the Coal-measures , " Part III .
, 'Phil .
Trans. , ' B , 186 , 1895 , p. 748 .
OF FOSSIL PLANTS FROM THE PALAEOZOIC ROCKS .
section , but are conspicuous enough when met with in longitudinal sections ( Plate 10 , fig. 5 ) .
They consist of tracheides smaller than the pitted elements , with densely spiral , or finely scalariform , markings on their walls .
The delicacy of the sculpturing on these elements contrasts sharply with the coarser structure of the pitted tracheides , though transitional forms are not wanting .
These peripheral strands pass out to form part of the xylem of the leaf-traces , but in doing so they are accompanied by a number of the pitted tracheides .
The spiral and scalariform elements were no doubt among the first to be differentiated .
In favourable cases the actual protoxylem can be detected , in the form of one or two narrow , laxly spiral tracheae , lying near the outside of the xylem-strand , but not at its extreme margin ( see Plate 10 , fig. 5 ) .
Thus the whole arrangement of the primary tracheae , as well as their minute structure , is almost identical in each steal of the Medullosa .
with that in the single steal of Heterangium* The conjunctive parenchyma consists of short thin-walled cells , as is best shown in Plate 10 , fig. 4 , which is drawn from a part of one of the steles of Specimen II .
, where the parenchyma is exceptionally abundant .
In this case it will be noticed that the spiral tracheides are not limited to the periphery of the primary wood .
The section , however , cuts through a narrow protrusion of the steal , and not through its central portion .
Where the intertracheal bands of primary parenchyma are thin , they may offer a deceptive resemblance to medullary rays , as seen in tangential sections of the secondary wood .
Sometimes large , thick-walled elements are found among the delicate cells , but such appearances may be due to pathological changes , for the tissues in question are often infested with fungal spores and fragments of mycelium .
Around every steal a broad zone of secondary wood and bast is constantly present ; the former is at once distinguished from the primary wood by the radial seriation of its elements ( see the photographs on Plate 6 ; also Plate 10 , fig. 3 ) .
The thickness of the secondary wood is very variable ; it often greatly exceeds the diameter of the primary mass .
It might be expected , from the analogy of other species of , that the secondary thickening would be greatest in each steal on the side towards the exterior of the stem , but no such rule holds good here , and indeed the reverse is often the case .
At the points where leaf-traces separate from the steal , the secondary zone is interrupted .
Immediately below the gap thus formed , the secondary wood is both smaller in quantity , and composed of smaller elements than elsewhere ( see photographs 6-9 ) .
The radial series of tracheides are as a rule more numerous than the medullary rays ; usually from two to four series of tracheides intervene between two rays .
The latter are in most cases of great vertical height , and from one to three cells in thickness , but short rays , one cell thick , sometimes occur ( Plate 10 , fig. 3 , and Plate 11 , fig. 8 ) .
At some places the secondary , like the primary wood , is very parenchymatous , each ray occupying as much space as several rows of tracheides .
The cells of the rays are * Cf .
Williamson and Scott , loc. cit. , pp. 749 and 760 , Plates 26-29 , figs. 24 , 25 , 32 , 33 , and 34 .
N VOL. CXCI.---B .
DU .
1 ) .
II .
SCOTT ON THE STUUCTUliE AND AFFINITIES longest in the radial direction , and have the usual muriform arrangement ( see Plate 10 , fig. 4 ) .
The principal rays , which extend throughout the whole thickness of the secondary zone , are connected towards the interior with the primary conjunctive tissue of the steal .
Secondary rays , starting far out in the wood , are found where the latter reaches a considerable thickness .
Where , as is usually the case , the medullary rays are narrow , the tracheides have a very regular and fairly vertical course ( see Plate 11 , fig. 8 ) .
Where , however , the rays are of great breadth , the tracheides follow a more devious path , curving round the ends of the broad rays , and even passing through them in an horizontal direction ; in a radial section of the wood such tracheides may be cut more or less transversely ( see Plate 7 , photograph 12 ) .
The significance of these deviations from the normal arrangement is no doubt to be sought in the necessity for keeping up communication between the different water-conducting channels .
Similar irregularities are frequent in the wood of recent Cycads .
The sculpturing of the secondary tracheides consists of multiseriate bordered pits , with narrow slits , which are often extremely well preserved ( see Plate 7 , photograph 13 ; and Plate 10 , fig. 5 ) .
The pits are as a rule limited to the radial walls , and appear only to be present on the tangential surfaces when the regular arrangement of the secondary elements is disturbed , as happens near the exit of a leaf-trace , or where the medullary rays are of exceptional width .
Where the pits communicate with the cells of a ray they are , as one would expect , unilateral , the thickening being developed only on the tracheal side of the wall .
It occasionally happens that the continuity of the secondary wood is interrupted , the radial series of elements starting afresh at a certain distance outwards , and no longer corresponding to those previously formed .
The resemblance to an annual ring is merely superficial , and the peculiarity is evidently due to some local interference with the cambial activity .
The phloem is only preserved here and there , but in some parts of Specimen I. is well shown ( see Plate 10 , fig. 3 ; and Plate 11 , fig. 9 ) .
The phloem-elements are arranged in radial rows , corresponding to the series of tracheides in the wood .
They appear to have very thick cell-walls , so that the lumen , as seen in transverse section , is often reduced to a mere dot .
Precisely the same peculiarity has been noticed by Count Solms-Laubach in the phloem of Medullosa Leuckarti* It is remarkable that these elements should present the same appearance in the calcified English material as in the silicified specimens investigated by Count Solms .
Probably this observer is right in regarding them as representing sieve-tubes and not bast-fibres , for their walls , which are very light in colour , present a totally different aspect from the undoubtedly sclerenchymatous fibres of theh ypoderma ( cf. Plate 12 , fig. 14 ) .
And further , if the phloem-elements were of a fibrous nature , it is almost certain that they would be preserved much more generally than is actually the case .
* 'Ueber MedullosaLeuckarti , ' p. 179 .
OF FOSSIL PLANTS FROM THE PALAEOZOIC ROCKS .
01 The argument used by Count Solms , that there are no other elements in the phloem which could be the sieve-tubes , if these are not they , applies with equal force to the English Medullosa .
Probably the cell-walls of the elements in question were comparatively thin during life , but swelled up greatly under the influence of maceration and decay in the interval preceding petrifaction .
It is also possible that the thick-walled appearance may be altogether delusive , depending rather on some change in the cell-contents .
In this case the apparent " middle lamella , " which is sharply defined , would really represent the whole thickness of the cell-wall .
Longitudinal sections show that the sieve-tubes , as we may provisionally call them , were of great length , with tapering ends .
Their course and relation to the medullary rays are well shown in Plate 11 , fig. 9 , drawn from a tangential section of the phloem at one end of the same steal , which is shown in transverse section to the left of photograph 5 ( st.1 ) .
The phloem-rays , occupying the meshes between the strands of sieve-tubes , are somewhat broader than the corresponding xylem-rays , and consist of short-celled parenchyma .
Beyond the phloem is a zone of ill-preserved tissue , in which mucilage-canals are contained ; this zone may have been of the nature of a pericycle , but no details can he made out .
The cambium between wood and bast is never well-preserved , though some traces of it can be seen in Plate 10 , fig. 3 .
The regular radial seriation of the elements leaves no doubt as to their mode of development .
There is evidence that the steles in their course through the stem underwent both fusion and division , though probably only at long intervals .
Thus , if we compare photograph 5 with photograph 2 , from two approximately transverse sections about two inches apart , we find that the tissues have undergone a certain amount of rearrangement .
Photograph 5* is from the lower , and photograph 2 from the upper , of the two sections .
One of the steles ( st.1 ) goes through , not , indeed , unaltered , but without joining either of the others .
Steal 2 of photograph 5 has , however , fused with the adjacent end of steal 3 , to form a new mass which we may call steal 2 ' .
The remaining portion of the original steal 3 has reorganised itself as a complete steal s ' in photograph 2 ) .
It is unfortunate that , owing to accidental ruptures of the tissues , no section shows the actual fusion clearly , but the course of the steles has been satisfactorily made out by carefully following them through six intermediate transverse sections cut between those shown in the photographs cited .
The prevalence of the number three for the steles may indicate that at the same level three leaf-bases were usually receiving bundles from the stem , a suggestion which finds some support from Specimen II .
in particular ( photograph 1 , and diagram , p. 126 ) .
The steal marked st.1 in photograph 5 and in the diagram ( fig. 2 ) is peculiar , in that it undergoes division for a short distance into two parts , which have just re-united * See also diagram , Plate 10 , fig. 2 .
In photographs 2 and 5 the right and left are reversed .
N 2 02 DR. D. IT .
SCOTT ON THE STRUCTURE AND AFFINITIES before reaching the level of photograph 2 .
Photograph 4 is from one of the intermediate sections , and shows the steal divided .
Only one of the halves becomes completed by the closing in of the secondary zone .
The other half never becomes converted into a complete steal , but its primary wood remains open on the side towards the twin-half , passing over , without any sharp break , into the intermediate parenchyma .
Just above the division all the middle part of the steal is unusually rich in parenchyma , and this condition appears to have allowed of an extension of the cambium across the primary wood .
The whole process is exceptional ( though something similar lias been observed in another specimen ) , and does not seem to be related to the normal bifurcation of the steles .
4 .
The Leaf-trace Bundles .
The question of the relation between the foliar bundles and the vascular system of the stem in Medullosa , is quite cleared up by the specimens under consideration .
The leaf-traces are met with , in the various transverse sections , in all parts of their course , from their first leaving the steal to their finally passing out into the petiole .
The series of successive transverse sections which have been prepared from every specimen , render it possible to follow the same strand for long distances , while in many cases the leaf-traces shown in the corresponding transverse and longitudinal sections can be identified with one another .
The leaf-traces are always given off from the free faces of the steles , i.e. , from the sides or ends directed towards the cortex ; they are never found in the central space between the steles .
Where , however , a steal projects beyond its neighbours , leaf-traces may arise from both sides of the projecting part .
Where a leaf-trace first starts from one of the steles , it constitutes a large strand , more or less circular in transverse section , and surrounded , completely or partially , by a zone of secondary wood and bast ( see Plate 6 , photographs 6-9 ; Plate 11 , fig. 10 ) .
Further out in the cortex the original leaf-trace divides repeatedly , to form several smaller bundles , at the same time losing its secondary tissues ( see Plate 6 , photograph 9 ; and Plate 11 , fig. 12 ) .
In the outer cortical layers these branch-bundles have assumed collateral structure , and now precisely resemble the typical Myeloxylon-bundles of the petioles ( see Plate 7 , photographs 14 and 15 ) .
The junction of the leaf-traces with the steles is shown at many places in all the specimens , and often with the greatest clearness ( see photographs 6 and 8 ) .
A large strand , at the periphery of the primary wood , detaches itself from the main mass , and passes very gradually outwards , accompanied on its outer side by a crescent of secondary tissue .
The tracheides of the outgoing strand diverge but slightly from the vertical direction , and the arrangement of the elements in the rest of the steal is not disturbed .
I have seen nothing here of the horizontal or very oblique strands of tracheides , which Count Solms-Laubach observed in the case of M. Leuckarti , and OF FOSSIL PLANTS FROM THE PALAEOZOIC ROCKS .
believed to be connected with outgoing bundles.# Where such irregularities occur in the primary wood of our specimens , they have no relation to the leaf-traces , which pass out in the same orderly manner as in H The trace passes gradually out through the secondary wood , which then closes in behind it ( see photographs 6-9 ) .
The secondary tracheides accompanying , or in the immediate neighbourhood of a leaf-trace , are usually smaller than in other parts of the steal .
As soon as the bundle is free , its own secondary tissues extend all round it ( photographs 7 , 8 , 9 ; fig. 10 ) ; they may be irregular , but are developed indiscriminately on all sides of the strand .
It is probable that the leaf-traces , in this part of their course , were really concentric bundles , though this cannot be strictly proved , as it is impossible to determine with certainty whether the phloem was , from the first , continuous all round the xylem .
We know that in Lyginodendron Oldhamium , where the leaf-traces while within the stem are certainly collateral , the secondary tissues , in certain exceptional cases , extended all round the bundle ; t in Medullosa anglica , however , this condition is constant .
The distribution of the secondary tissues around the leaf-trace , and its general appearance , certainly point to a primarily concentric structure.^ The gradual exit of the leaf-trace from a steal can also be well followed in some of the longitudinal sections ( see Plate 7 , photograph 11 ) .
The intercalation of secondary wood behind the trace , in connection with the latter and with the steal , is well shown , where the direction of the section is favourable .
The structure of an undivided leaf-trace near its base , bears much resemblance , on a small scale , to that of the steal itself ( Plate 11 , fig. 10 ) , and recalls the " stellate rings " described in other Medullosece .
The primary xylem includes a certain amount of parenchyma .
Most of the tracheides are large and pitted , like those which form the bulk of the primary wood in the steles ( see Plate 11 , fig. 11 , which is from the same leaf-trace , l.t.2 , which appears in transverse section in Plate 6 , photograph 8 ) .
One or more groups of smaller elements are present , and are shown by longitudinal sections to consist of spiral and scalariform tracheides ( fig. 11 ) .
Where the bundle first leaves the steal , its protoxylem is directed outwards ; irregularities in orientation , which one sometimes observes , may be due merely to displacement .
The secondary wood of the leaf-trace often contains , in addition to the ordinary elements , a number of extremely short tracheides with square ends ( fig. 11 ) ; their longer axis may be tangential to the wood .
In two cases it was noticed that a leaf-trace after diverging from the steal was connected with it for a considerable distance , by radial bridles of short tracheides .
There is reason to believe , as we shall see below , that this peculiarity was correlated with the insertion of an adventitious root .
* " Ueber Medullosa LeucJcarti , " 'Bot .
Zeit .
, ' 1897 , p. 180 , Plate 6 , fig. 1 .
f Williamson , 'Organization , \amp ; c. , ' Part IV .
, Plate 22 , fig. 1 , Williamson and Scott , 'Further Observations , \amp ; c. , ' Part III .
, p. 724 .
1 Cy .
Solms-Laubach , 'Ueber Medullosa Leuckarti , ' p. 194 , Plate 6 , fig. 4 .
The phloem can usually be recognised round the whole circumference of the bundle , but its preservation is imperfect ( Plate 11 , fig. 10 ) .
Several leaf-traces are often given off near together , as shown , for example , in photograph 9 .
Considering the great size of the leaf-bases borne on the stem , there can be no doubt that a number of the principal traces went to supply a single leaf .
The steal marked st.1 in photographs 2 , 4 and 5 and in the diagram ( fig. 2 ) can be traced through 18 transverse sections , representing a length of about 3J inches ( 9 centims. ) .
In this part of its course it gives off at least 10 leaf-traces .
Considering the great vertical height of the leaf-base , which , as we know from Specimen II .
, exceeded 5 inches ( 13 centims. ) , there can be no doubt that many of these , perhaps even a larger number , were destined for the same leaf .
There is evidence , as we shall see below , that each leaf-base continued to receive bundles from the stem up to the point where it was shut off by internal sclerenchyma .
It has not been possible to determine whether each leaf received bundles from a single steal only , or whether the adjacent parts of two steles contributed .
The number of bundles entering a leaf was largely increased , as we shall now see , by the subdivision , of the original traces in passing through the cortex .
The first stage of such a subdivision is shown on comparing photograph 8 with photograph 9 .
The vertical distance between the two sections represented was about a centimetre .
In the lower section ( photograph 8 ) the larger leaf-trace , l.t.2 , is undivided , but in the upper ( photograph 9 ) it is separating into two strands .
The secondary arc is still complete on one side , but the primary xylem is severed into two halves by the intrusion of parenchyma .
An intermediate section shows the first beginning of the fission , while in a section higher up in the series the division is complete .
Various other cases have been observed .
In Specimen III .
a leaf-trace was followed , in 13 successive transverse sections , through a vertical distance of about 2| inches ( 6 centims. ) .
It gradually lost its secondary tissues in its upward course , and divided twice .
It appears that the leaf-traces underwent further subdivision in the upper part of their course .
Thus Plate 11 , fig. 12 , shows two groups of bundles , each of which must have arisen by the division of a single strand , for in a section about a centimetre lower down , the bundles of each group are still connected with each other .
These are strands which have lost all .
trace of secondary tissues , and which lie outside a band of internal sclerenchyma , so that they were evidently still undergoing division when already far advanced on their outward course .
The leaf-traces as they divide , lose their concentric structure and become collateral .
At the stage of division shown in fig. 12 , this change has already taken place .
Probably the collateral structure was assumed lower down , for it is found in bundles which are still in the cortex of the stem , as distinguished from that of the leaf-bases .
This is beautifully shown , for example , in the pair of bundles ( probably arising from the division of a single one ) represented in transverse section , Plate 7 , photograph 14 .
These bundles , with many others like them , lie well within the internal sclerenchyma OF FOSSIL PLANTS FROM THE PALAEOZOIC ROCKS .
( see Plate 6 , photograph 5 , v.b. The phloem is excellently preserved , and is evidently limited to the outer side of the xylem .
The smallest xylem-elements are adjacent to the phloem , and in fact the structure is in all respects that of the well-known " exarch " vascular bundles of Myeloxylon .
Photograph 15 , showing a similar pair , is taken from a more external part of the cortex , lying within the triangular space indicated in Plate 10 , fig. 2 , between the hypoderma , o.c. and the internal sclerenchyma This space , as proved by the comparison of successive transverse sections , represents the base of a leaf , which at a higher level begins to become free from the stem ( photograph 2 , Pet. ) .
We see , then , that the structure characteristic of the vascular bundles of the petiole is already assumed , not only in the leaf-base , but within the cortex of the stem .
As shown by the longitudinal sections ( Plate 12 , fig. 15 ) , the collateral strands , resulting from the division of the original leaf-traces , no longer possess any pitted elements in their xylem , which here consists entirely of spiral or finely scalariform tracheides accompanied by a few parenchymatous cells .
The latter appear , in the collateral bundles , to lie only at the edge of the wood , and not in the midst of the tracheides .
Where the spiral bands of two contiguous walls cross one another an appearance of reticulation is produced .
A narrow laxly spiral tracheide of the actual protoxylem can be recognized in the radial section figured , separated only by one layer of cells from the delicate and elongated elements of the phloem ( see fig. 15 ) .
In every detail we find the closest agreement with typical Myeloxylon bundles.* The conclusions arrived at respecting the leaf-trace system of Medullosa may be summed up as follows :\#151 ; ( 1 ) The leaf-trace strands are given off from the external side of the steles , a number of such strands contributing to supply a single leaf .
( 2 ) On first leaving the steles , the leaf-traces are large , probably concentric bundles , each surrounded by its own zone of secondary wood and bast .
( 3 ) In passing through the cortex the leaf-traces lose their secondary tissues , and divide up to form several smaller bundles , which assume the collateral , exarch structure characteristic of Myeloxylon .
5 .
The Ground-tissue and Periderm .
The central ground-tissue , lying between , or immediately surrounding the steles , is generally much disorganized .
Between the steles there was probably but little intermediate tissue during life , the pericycles of the vascular masses coming near together ( see Plate 7 , photograph 10 ) .
The tissue immediately outside the group of steles consisted of thin-walled parenchyma traversed by numerous leaf-trace bundles and * Cf .
Seward , " On the Genus Myeloxylon , " * Annals of Botany , ' vol. 7 , Plates 1 and 2 ; especially figs. 16 and 17 ( 1893 ) .
9(5 DR. D. H. SCOTT ON THE STRUCTURE AND AFFINITIES gum-canals .
The latter occur in all parts of the extrastelar tissues as well as in the pericycles ; those shown in Plate 12 , fig. 14 , from the hypoderma , will serve as examples .
In Specimen I. the more external part of the ground-tissue is very well preserved , especially on one side of the stem ( Plates 5 and 6 , photographs 2 and 5 ) , where the infiltrating substance was different from that of the rest of the specimen , as shown by its less easy solubility in acetic acid .
This tissue , however , probably belongs for the most part to the leaf-bases , for it agrees very closely with the tissues of the evident leaf-bases shown so well in Specimen II .
( photographs 1 and 3 ) .
It is clearly impossible to draw a sharp distinction between leaf-base and cortex , for the one merges imperceptibly into the other , except in the upper part of the leaf-base , where the internal sclerenchyma serves to mark the limit .
The cortical parenchyma shown in Plate 5 , photograph 2 and neighbouring sections , is very poor in both vascular bundles and gum-canals , and therefore probably belongs to the region at the bottom of a leaf-base .
Further description of the more external tissues will be postponed until the leaf-bases and petioles are dealt with .
In Specimen I. sclerenchymatous bands sometimes appear isolated in the ground-tissue ( see Plate 10 , fig. 2 , diagram ) , and , so far as the transverse sections show , are not in connection with the hypoderma .
They , no doubt , represent the downward continuation of the sclerenchyma which , at a higher level , passes out to form the inner surface of the petioles.* The relation of the internal sclerenchyma to the leaf-bases is clearer in Specimen II .
( see photograph 1 , and diagram p. 126 ) .
The development of an internal periderm is a striking feature in most of the specimens ( see Plate 5 , photographs 1 and 3 ; Plate 7 , photograph 16 ; Plate 10 , figs. 6 and 7 ) .
It forms a continuous though very irregular zone , running round the stem a little within the bases of the leaves .
The regular radial arrangement of the cells at once characterizes this zone as a secondary formation ; it attains in places a thickness of at least 25 elements .
The cells , as seen in radial or transverse section , have a tabular form ; their walls are of moderate thickness .
Seen in tangential section ( fig. 7 ) , the arrangement of the peridermal cells is , as one would expect , quite irregular , but the zone is at once recognized by the uniformity and marvellous preservation of its elements .
Towards the interior the periderm passes over into the primary cortex , and , on the boundary , cells are found with very delicate tangential walls , which are evidently the result of the last divisions , and thus indicate the position of the phellogen ( fig. 6 ) .
The great bulk of the periderm was evidently formed to the outside of its generative layer , thus corresponding morphologically to cork .
It is extremely probable , from the very perfect preservation of its cell-walls , that the tissue was actually suberized .
There was little , if any , phelloderm to the interior of the cork-cambium .
In Specimen I. , unlike the other three , the periderm is very little developed , and cannot * Cf .
Solms-Laubach , 'Ueber MLeuckarti , ' p. 191 .
OF FOSSIL PLANTS FROM THE PALEOZOIC ROCKS .
be traced as a continuous ring .
Evidently it was only beginning to form in this specimen ( see Plate 11 , fig. 12 , pd .
) .
From the first , the formation of the periderm was very irregular ; thus in Specimen I. , where it is so little developed , it already appears here and there in the form of loops or rings enclosing leaf-trace bundles , or portions of sclerenchyma .
The same thing is even more conspicuous in the more advanced stems , where prominent bulges or deep invaginations of the periderm are of constant occurrence , and sometimes a closed ring of periderm appears , quite isolated from the main zone , though probably connected with it at a different level ( diagram , p. 126 ) .
It is in the excision of effete tissues that we must seek the significance of these anomalous developments .
The most extreme case of the kind is found where the periderm cuts deeply into the interior of the steles themselves .
This pathological process has gone on to a great extent in Specimen III .
, which was evidently in a senile condition .
The periderm can be traced into the steles , from the cortical tissues , through the leaf-trace gaps ; clearly the cell-division had spread from one cell to another , until it invaded the parenchyma of the primary wood , with the result that the latter has become cut up in various directions by bands of intrastelar periderm .
At one place in another specimen ( No. II .
) still more extraordinary phenomena have occurred , for here not only has periderm extended deeply into the interior of a steal , so as to obliterate most of its primary wood , but it has even replaced a great part of the secondary wood , completely changing the whole structure of the steal affected .
The abnormality is quite local , and was probably due to some injury at an early stage of growth .
It is of interest chiefly as a warning , for in less perfect material such pathological symptoms might easily be confused with normal processes of growth .
Those who have worked at the anatomy of recent Cycads will readily recall parallel cases .
Certain very definite outgrowths of the periderm , which occur in connection with the bases of the adventitious roots , will be best considered below , together with the roots themselves .
The normal zone of periderm must obviously have cut across all the vascular bundles on their way out to the leaves ; we consequently often find them imbedded in the thickness of the periderm .
Possibly they may have kept up communication between the inner and outer tissues for some time , for Specimens II .
and IV .
show that the leaf-bases remained on the stem , without any apparent withering , after a thick zone of periderm had been formed .
Specimen III .
, as mentioned above , is peculiar in having lost all its tissues exterior to the periderm .
The specimen was old , compared with the others , as shown by the great thickness of its secondary wood , and by the wide-spread occurrence of pathological new-formations in its tissues .
It is , therefore , very probable that it may have reached a stage at which the exfoliation of all the external tissues normally took place .
It is difficult to see the object of the well-marked periderm-formation characteristic of the species , unless exfoliation eventually occurred , but at the same time we cannot absolutely prove that the loss VOL. CXCI.--B .
O 'J8 DR. D. H. SCOTT ON THE STRUCTURE AND AFFINITIES of the external tissues , in such a case as that of Specimen III .
, may not have been due merely to maceration and decay .
G. Accessory Vascular Strands .
In addition to the regular constituents of the vascular system of the stem , namely the normal steles and leaf-traces , we sometimes find accessory strands of wood and bast , which do not fall under either of these categories .
Their occurrence is very inconstant ; where present at all , they are always found in the cortex , outside the regular group of steles .
An anomalous structure of this kind is present in Specimen I. , a little to the exterior of the subdivided steal described above ( see Plate 5 , photograph 4 , an ) .
Another example ( from Specimen II .
) is shown in Plate 12 , fig. 18 , on a larger scale .
The structure is essentially similar in both .
The strands are of considerable size ; thus , in the case first mentioned ( photograph 4 ) , where there are two of these anomalous rings side by side , the larger of the two has a maximum diameter of 7 millims. , its long axis lying tangentially .
The general structure in these cases somewhat resemljles that of a normal steal .
The secondary tissues form a more or less irregular ring , enclosing a parenchymatous central region , in which tracheides are scattered .
All the elements are small , compared with those of the regular steles ; the internal tracheides are often quite short , with square ends , and reticulated walls The secondary wood has broad rays ; cambium and phloem are sometimes fairly preserved ( Plate 12 , fig. 18 ) .
The anomalous strands in Specimen I. ( photograph 4 ) are more irregular than that figured in detail from Specimen II .
; thus on one side of the ring in the former case the secondary wood is undeveloped for some distance , and is here replaced by an excessive growth of phloem or secondary parenchyma .
Longitudinal sections of Specimen I. , cut immediately below the transverse section from which phot .
4 was taken , show that the anomalous strand fused , in the downward direction , with the adjacent normal steal .
This fact may have some relation to the division of the latter about the same point .
In other cases also accessory vascular tissues have been found , which sometimes take the form of irregular bands , and not of closed rings .
These anomalous vascular strands are probably best regarded as extrafascicular new formations , comparable to the cortical bundles of which they much resemble , or to the irregular strands which sometimes occur in the extrafascicular region of Macrozamia .
* We know , from the investigations of Weber and Sterzel , that some { e.g. , M. stellata , var .
gigantca ) formed successive extrafascicular zones of wood and bast outside their stelar system , just as we find in , and * See Worsdell , " Comp. Anat .
of certain Genera of the Cycadaceae , " ' Linnean Society 's Journal\#151 ; Botany , ' vol. 33 , 1898 , pp. 441-444 , Plate 20 , figs. 6-8 ; " Anatomy of the Stem of ' Annals of Botany , ' vol. 10 , 1896 , p. 610 , figs. 2 , 9 , and 10 .
OF FOSSIL PLANTS FROM THE PALAEOZOIC ROCKS .
lartos at the present day.# It is therefore not surprising that in our species we should meet with other characteristic Cycadean anomalies .
The accessory strands in M. anglica are certainly quite different from the normal steles and leaf-traces ; neither can they be identified with the strands supplying adventitious roots , which had a more horizontal course .
There is no indication of their connection with any other form of lateral appendage .
7 .
The Leaves .
The structure of the leaf-bases in connection with the stem is best shown in Specimen II .
( Plate 5 , photographs 1 and 3 ) , though also evident in parts of Specimens I. and IV .
The outer cortex or hypoderma , as seen in transverse section , is represented in detail in Plate 12 , fig. 14 .
The fibrous strands , which have a rounded transverse section , are ranged three or four deep , the smallest towards the outside , with dense , large-celled parenchyma filling the interstices between them .
In some places a more delicate parenchyma , forming an external layer outside the fibrous zone , is preserved , and no doubt this was always present during life .
When this external parenchyma is well preserved , it is found to consist of smaller and thin-walled cells , with abundant cell-contents .
The outermost cortical layer , as well as the actual epidermis , had a well-marked palisade form .
This external zone probably formed the assimilating tissue of the organ ( see Plate 11 , fig. 13 ) .
Gum-canals are abundant in the hypodermal parenchyma , and often abut on the fibrous strands .
As a rule the actual epithelium of the canal has perished ; in a few cases some remains of it can be traced .
A carbonaceous strand , representing no doubt the fossilised secretion , often traverses the canal .
Tangential sections of the external cortex show that the fibrous strands have a straight course , without anastomosis ( see Plate 7 , photograph 17 ) .
This is extremely evident , even to the naked eye , on the exposed surface of the specimen , as shown in Plate 9 , fig. 1 , where the prominent ribs , representing the fibrous strands , can be traced for several inches without finding any fusions .
For this form of hypodermal tissue , which is characteristic of " Myeloxylon " among other fossils , Count Solms-Laubach has proposed the name ( l Sparganum cortex , " f taken from an obsolete genus established by Unger for certain fern-petioles .
The " Dictyoxylon cortex " ( typically shown in Lyginodendron Oldhamium)differs from the Sparganum type in the reticulate arrangement of the fibrous strands .
The fibres are of great length , with thick walls , while the intermediate parenchyma consists of very short cells , often broader than long .
The internal parenchyma of the leaf-bases , which is very well preserved in some cases ( see Plate 5 , photograph 3 ; * Weber and Sterzel , ' Bcitriigc sir Kenntniss der Medullosese , ' p. 27 , Plate 8 , fig. 2 .
t " Pflanzenreste des Unterculm von Saalfeld , " ' Abhandl .
d. K. Preuss .
Geol .
Landesanstalt ; ' Never Folge , Heft 23 , p. 49 , 1896 .
O 2 Plate 7 , photograph 17 ) , consists uniformly of large but short cells , with rather thin walls .
There are no fibres in this part of the tissue , except in connection with the vascular bundles .
The parenchyma is traversed by numerous gum-canals , like those of the external cortex , but often larger .
In some cases , however , the cavity of the canal is small , or even almost absent .
When this is so , the surrounding cells had evidently undergone active cell-division , by cell-walls tangential to the cavity , as is shown both in transverse and longitudinal sections .
Canals in this condition appear at first sight to be at an early stage of development .
It is , however , improbable that new canals should have been in course of formation in the base of a leaf already separated from the inner tissues by a broad zone of periderm ( photograph 3 ) .
A more probable interpretation is that the canals in question were undergoing obliteration , by dilatation and division of the surrounding cells .
The numerous vascular bundles in the leaf-bases have the collateral exarch structure already described .
The phloem is only occasionally preserved , but its position on the outside of the bundle is quite clear .
The xylem has its smallest elements next the phloem ; it is separated from the surrounding parenchyma by a sheath of small thick-walled elements , which are shown by longitudinal sections to be fibres .
The fibrous sheath does not extend round the phloem .
Traced longitudinally , it begins to appear in the base of the leaf , hut is best shown in sections of free petioles , which are abundant in the blocks containing Medullosaanglica .
The bundles , in their course through the leaf-base and petiole , sometimes anastomose .
The number of bundles to be counted in the transverse section of a leaf-base varies according to the level at which the section is taken .
Thus , in the leaf-base marked c in photograph 1 , which , as we know , -was near its point of separation from the stem , nearly 80 bundles were found , while in that shown in photograph 3 there are only about 30 .
The latter is not yet marked off from the stem by internal sclerenchyma , and we know that it did not begin to become free for more than 3 inches above the plane of section .
Another leaf-base , cut still lower down , only showed about 15 bundles .
It is evident then that new sheaves of bundles successively entered the leaf-base at different levels , probably up to the point at which the appearance of internal sclerenchyma formed a harrier between leaf and stem .
In the leaf-base shown in photograph 3 the more external bundles are ranged in a very definite horse-shoe curve , just wdthin the hypo-derma ; these probably represent the first sets which entered the leaf-base , while those grouped more to the interior have passed in at a higher level .
This regularity of arrangement is gradually lost as the leaf-stalk receives its full complement of bundles .
In some transverse sections , the hypoderma appears of exceptional thickness , and is evidently cut obliquely ( see Plate 7 , photograph 16 ) .
There is little doubt that in these cases the plane of section passes immediately above the axil of a leaf .
The sclerenchy-matous strands which , in the photograph referred to , are seen passing out from either side of the thickened hypoderma , must belong to the upper free surface of the petiole , the rest of which is lost .
OF FOSSIL PLANTS FROM THE PALAEOZOIC ROCKS .
The larger free petioles accompanying the stem are absolutely identical in structure with the leaf-bases still attached to it .
The large leaf-base , part of which is shown in photograph 2 ( from Specimen I. ) , is already almost free from the stem , and becomes quite free , as traced upwards in successive sections , without showing any change of structure .
The petiole shown in transverse section in Plate 8 , photograph 18 , is from another part of the same block , and probably at some distance from its attachment to the stem .
The mean diameter of this fragment is about 175 centims. , whereas the leaf-bases at their insertion on the stem are about double that size , taking transverse measurements .
Except for its smaller dimensions the petiole referred to shows in all respects the same structure as the leaf-base ; the hypoderma is distinctly of the type of Myeloxylon Lanclriotii , Pen .
The parenchyma has here become crushed along certain lines , forming a coarse net-work , precisely as in some of M. Renault 's sections .
Among the larger petioles are smaller ones of various dimensions .
Some of these , with a diameter of 4 or 5 millims. , may represent the rachis of the main pinnae .
These have the same structure as those already described , on a smaller scale .
The hypoderma is , however , here comparatively thin , with a single row of wedge-shaped strands of sclerenchyma .
Still smaller branches have the structure shown in Plate 12 , fig. 16 , drawn from a rachis only 1 '25 millims. in diameter .
It has distinct wings on one side , no doubt representing the upper surface .
A continuous band of sclerotic tissue runs round the periphery ; in the parenchyma many gum-canals are contained .
In the middle is a group of five or more bundles , apparently surrounding a canal .
As far as can be seen , they are of the usual type , with their phloem , which is crushed , directed outwards .
This may well have been the rachis of a secondary pinna , bearing the ultimate pinnules.# Sections of leaflets , such as are usually found associated with Myeloxylon , are very common among the fragments of the petioles and rachis .
One of the best is shown in transverse section in Plate 12 , fig. 17 .
The midrib contains two bundles close together , which , so far as the imperfect preservation allows one to judge , appears to have been collateral , with the phloem directed outwards .
The lamina , which is incurved at either margin , has a distinct palisade layer , and is manifestly of bifacial type .
A vascular bundle of the lamina , which runs near the lower surface , is well shown in longitudinal section .
I have not yet found absolute proof that these leaflets belong to the ( i Myeloxylon"rachis , but the constancy of association , and general agreement in the structure leaves no real doubt .
The leaf must have been very large ; in one slide a rachis ( probably of a primary joinna ) can be traced in longitudinal section for fully 5 inches , apparently without branching .
In other cases branching is shown .
The facts just detailed show that Medullosa anglica bore large , highly compound leaves , the petioles of which have the structure of the Myeloxylon Landriolii of Renault .
The latter is not to be regarded as a species in the same sense as in * In another section a small rachis , like that shown in fig. 16 , is seen arising as a branch from one of larger dimensions .
speaking of recent plants , but rather as a type of petiolar structure , common , no doubt , to a certain group of the Medullosete .
M. Renault found reason to believe that Myeloxylon Landriotii represents the petiole and rachis of leaves belonging to the form-genus The organization of the leaves of our Medullosarenders it highly probable that they too would fall under the same category .
The resemblance in structure to M. Renault 's specimens is , in fact , a very close one/ " The identification of the leaf of Medullosa cannot , however , be finally settled without further investigation , on which I am now engaged .
8 .
The Roots .
It appears that the structure of the root has not been described as yet in any species of Medullosa .
The material from Hough Hill contains a number of specimens of a very characteristic root , of triarch structure , with well-marked secondary wood and bast , and a conspicuous zone of periderm ( see Plate 8 , photograph 19 , and Plate 13 , fig. 20 ) .
The larger roots attain a mean diameter of about half an inch ( 12 millims. or more ) , and are accompanied by rootlets of various dimensions and similar structure .
Of the specimens examined , many were found in close association with the Medullosa stems , Avhile others occurred in separate nodules from the same locality ; of these Mr. Lomax also prepared the sections .
In all cases the roots are associated with fragments of the " Myeloxylon " petioles , which we now know to have belonged to the Medullosa .
The histological structure of the wood in these roots and in the stem of Medullosa anglica was found to agree so closely , as to suggest at once the probability that the organs belonged to one and the same plant , a conclusion which has now been confirmed by direct proof of continuity between root and stem .
The structure of the roots and rootlets will first be described , after which the evidence will be given which demonstrates their connection with the stem of Medullosa anglica .
If we examine a well-preserved specimen of one of the larger roots , such as those shown in transverse section in photograph 19 and fig. 20 , we find the following structure .
The root is bounded externally by a broad zone of radially seriated parenchymatous tissue , having a thickness in different specimens of from eight to twenty-four or more cells .
This tissue is evidently of the nature of periderm ; the narrowest cells , with the thinnest tangential walls , lie towards the interior of the zone , so we may conclude that its development was mainly centripetal , the bulk of the tissue having been formed , like cork , on the outer side of the phellogen .
The cell-walls of the large peridermal cells are sharply defined , though of no great thickness ; it is highly probable that they were suberized .
The periderm of the root bears a strong general resemblance to that of the stem of our Medullosa .
Within the periderm but little * Renault , ' Course de Botanique fossile , ' vol. 3 , Plate 27 .
OF FOSSIL PLANTS FROM THE PAL/ EOZOIC ROCKS .
parenchyma is preserved ; what there is occurs chiefly at the corners of the compressed root , where it has escaped crushing .
This tissue was probably pericyclic , for , as we shall see presently , the periderm was of internal origin .
The radial sections show the seriation of the peridermal cells with equal clearness ( see Plate 8 , photograph 20 ) , and indicate that the vertical height of these elements was about equal to their tangential width .
Almost the whole interior of the root is occupied by the vascular cylinder , which is well preserved .
In all the clear transverse sections the structure is evidently triarch ; in one somewhat doubtful case only it appeared to be tetrarch .
The three arms of the primary xylem are long , and run out to slender points , as seen in transverse section .
Longitudinal sections confirm the obvious conclusion that the protoxylem ( represented by very small densely spiral tracheides ) lies at these points .
In the primary wood the tracheides , which are accompanied by much conjunctive parenchyma , are largest about the middle of the strand , and diminish in size towards the angles .
The larger tracheides have multiseriate bordered pits distributed equally on all their walls ; as the protoxylem is approached the markings become more scalariform .
The secondary wood , which in all the larger roots much exceeds the primary in amount , forms three well-defined arcs , corresponding to the three sides of the triarch strand .
Opposite the xylem-angles little or no secondary wood is developed .
The radial rows of the secondary tracheides correspond in size to the primary elements on which they abut .
Thus those series which start from the more central part of the primary wood consist of large elements , while the series abutting on the smaller tracheides of the arms gradually become narrower as the angles are approached , thus giving a very characteristic habit to the transverse sections .
The secondary tracheides , like the larger elements of the primary wood , are marked with multiseriate bordered pits , but here the pits are limited to the radial walls ( see Plate 8 , photograph 20 ) .
The medullary rays can , in favourable cases , be traced through the whole thickness of the secondary zone , from the conjunctive parenchyma to the outside of the phloem .
The latter , when it is preserved , consists of narrow elongated elements .
On the outer border of the phloem are traces of secretory organs , which may have been either sacs or canals .
The sections occasionally show the roots in the act of branching ( Plate 8 , photograph 19 , rt .
) .
The branches are always rootlets of much smaller size than the parent root .
In the example illustrated , the xylem-elements of the rootlet abut directly on one of the protoxylein-angles of the main root .
Very short tracheides are often found ensheathing the wood of the rootlet at this point .
The phloem and pericyclic parenchyma of the rootlet are well preserved , but its endogenous origin is by no means evident .
This is no doubt due to the fact that the periderm which envelops both root and rootlet was an internal one , so that the primary cortex of both is lost .
Kootlets much smaller than the main roots hitherto considered , are often met with in transverse section ; in favourable cases their structure leaves no doubt as to their identity .
An especially clear specimen , 1*5 X 1 millim. in diameter , is shown in Plate 8 , photograph 21.The periderm is here 10 to 12 cells thick , and agrees exactly with that of the main roots ; the phellogen is beautifully preserved .
The triarch steal has begun to form secondary wood , which has only reached a thickness of about three elements .
The phloem , and even parts of the cambium , are in good preservation .
The phloem-groups are evidently placed between the angles of the xylem , so that the structure is typically that of a root .
The thin-walled , tabular cells of the phellogen lie so close to the phloem as to leave no doubt that the periderm arose in deep-seated tissue , and probably in the pericycle .
Direct proof of this fact is afforded by a still smaller rootlet ( with a mean diameter of scarcely a millimetre ) , shown in transverse section in Plate 12 , fig. 19 .
This rootlet retains its pr imary cortex , while the formation of per iderm has already begun .
The xylem , though slightly deformed , is evidently triarch , and agrees well with that of the other specimens , but irr this case no secondary growth had yet taken place .
Immediately beyond the phloem , which is partly preserved , is a zone of larger cells which were evidently undergoing active tangential division .
This zone , in fact , represents the periderm at an early stage of its development .
Adjoining it , on the outside , is a single layer of large cells , which do not fit orr to those of the periderm .
We carr hardly interpret this layer otherwise than as an endodermis , though I could not satisfy myself as to the presence of the characteristic marks on the radial walls .
Beyond the endodermis comes the primary cortex , which resembles that of the root of Lyginodendron Oldhamiumin its lax , lacunar structure , the presence of " secretory sacs , " and the definite double layer of cells at the periphery .
The rootlet might , in fact , be taken for a young triarch rootlet of , if it were not for the very distinct formation of periderm , which I have never found so well marked , even in the oldest roots of the latter plant.# The specimen , on the other hand , agrees so closely with the Medullosa roots now described , that there can be no doubt of its identity . .
It thus proves , in conjunction with other specimens , the pericyclic origin of the periderm in these organs .
It is unnecessary to take up more time with a description of the rootlets , which occur in great numbers , associated with the larger roots and with the stems of Medullosa anglica .
In many cases their identity is certain , while others are more dubious .
It is evident that the roots in question branched freely , and that their ultimate rootlets were extremely slender .
The root-nature of these organs does not seem open to question .
The anatomy is in all respects that of a root .
The example of Sphenophyllum warns us , indeed , that a very root-like anatomy may be found in a stem , but in the present case there are no indications whatever of cauline structure .
Nothing like a node is to be found ; the * Williamson and Scott , 'Further Observations , \amp ; c. , ' Part III .
, p. 740 .
The root of Lyginodendron Oldhamium is shown in that paper to be identical with the old Kaloxylun of WILLIAMSON .
OF FOSSIL PLANTS FROM THE PALAEOZOIC ROCKS .
appendages are all of one kind , and are manifestly rootlets ; that they are not obviously endogenous is fully explained by the early appearance of periderm , which so soon removed the whole of the primary cortex .
The organs are not , so far as I have been able to ascertain , identical with any fossil roots previously described .
They differ from those of in the presence of periderm and in the prevalence of triarch structure , which is exceptional in that plant .
In both the points mentioned our specimens resemble Amy don radicans , but they differ from it in the whole character of the wood , which in Amyelon is much denser , with smaller tracheides and much less parenchyma . .
The pitting of the tracheae is also quite different , nor does the periderm itself agree in detailed structure .
Amyelon requires further investigation , but in the mean time there can be no doubt that it is quite distinct from the roots under consideration .
It is not impossible that the finer rootlets above described may turn out to be identical with one or other species of llhizonium , but at present I have found no proof of this .
The agreement in histological detail between these roots and the stem of Medullosa anglica ( shown especially in the structure of the wood and periderm ) , together with the constant association of the stems and leaves of the Medullosa with the roots , strongly suggested that the one belonged to the other .
We have now to examine the direct evidence for this connection .
A specimen found by Mr. G. Wild , in the same material which yielded the stems of Medullosa , throws important light on the subject .
Plate 8 , photograph 22 , represents a section of this specimen , kindly lent me by the owner .
Four large roots , evidently of the same nature with those already described , are shown in transverse section .
Each root is surrounded by a broad zone of periderm , within which is a layer of parenchyma , containing secretory organs , and within this again the phloem can be traced .
The secondary wood is divided up by broad medullary rays , and agrees in all respects , including the pitting of the tracheae , which can be seen where the section is oblique , with that of the free roots .
The primary wood is rather obscure , owing no doubt partly to pressure , and partly to natural irregularity ; in two out of the four roots the general form of the primary strand , and the position of the principal rays , point to triarch structure .
The roots are crowded together , and their outline is more or less irregular* Between them are packed masses of tissue , some of which are of a different nature from the roots themselves .
Both these intermediate masses and the tissue bounding the group of roots laterally , are continuous in places with the adjacent roots .
This tissue ( M.c. in photograph 22 ) shows , in longitudinal section , the structure of the Medullosa cortex , including sclerenchymatous strands , gum-canals , and vascular bundles with tracheides like those shown in Plate 12 , fig. 15 .
The only possible interpretation of the facts is that the section is a tangential one of the cortex of a Medullosa stem , passing through the base of a group of adventitious roots .
By Mr. Wild 's permission eight more sections of the specimen were cut for me by Mr. Lomax , parallel to the first .
These , as it proves , cut successively deeper VOL. CXCI.\#151 ; B. P 10G DR. D. H. SCOTT ON THE STRUCTURE AND AFFINITIES into the stem , and render it possible to trace the connection between the tissues of the roots and those of the parent axis .
The peridermal zones of the various roots become confluent with each other and with the periderm of the stem itself , while the vascular cylinders of the roots open out , and join on to longitudinal strands of vascular tissue , which are themselves continuous with a steal , seen in longitudinal section , and showing in all details the structure of a steal of the Medullosa stem .
Plate 8 , photograph 23 , is from the innermost section of the series , and shows the steal ( at st. ) with which the roots have united .
The fragment was evidently a part of one of the stems of Medullosa anglica ( and , in all probability , of Specimen I. ) .
It is the only specimen I have seen which shows large and mature adventitious roots in connection with the stem , hut in the case of younger roots the connection has been established repeatedly .
A good instance of this also is afforded by Mr. Wild 's specimen , just described .
The same section ( photograph 23 ) which passes through one of the steles of the stem , also shows the cortical tissues , which are obliquely cut .
Towards one side the inner cortex forms a vertical row of conspicuous outgrowths in contact with one another , where the arrangement has not been disturbed by the intrusion of the ubiquitous Stigmarian rootlets .
These outgrowths are each coated on three sides by periderm , which is also continuous between their bases .
The cortical tissue of the outgrowths consists of thin-walled parenchyma , in which many of the elements have dense carbonaceous contents , and may be provisionally called secretory sacs .
Within each outgrowth is a vascular cylinder , seen in oblique section , and extending inwards , towards the interior of the stem .
In addition to the outgrowths , two free appendages are seen in transverse section , one of which is included in photograph 23 ; others appear in an adjacent section .
These structures are clearly roots ( see Plate 13 , fig. 21 ) ; most ot their cortex has been thrown off by the formation of internal periderm ; the parenchyma ( presumably pericyclic ) lying within the periderm , is of considerable thickness , and resembles the tissue of the cortical excrescences .
The vascular cylinder ( which has little or no secondary thickening , showing that the roots were still young ) is asymmetrically triarch , two of the arms forming a well-developed vertical band , with a shorter horizontal arm .
Small rootlets accompany the larger roots .
At one place ( fig. 21 ) the base of a similar root is still in connection with a cortical outgrowth , of which it forms one lobe .
As we shall see presently , it was often the case that two roots sprang from each cortical excrescence .
Mr. Wild 's specimen thus affords evidence that young roots may spring from the same part of the stem which bears mature roots , which have far outstripped the former in their development .
The growth of the immature roots may very probably have been arrested .
Still clearer evidence , as regards the mode of connection between the young roots and the stem , is found in another specimen , namely , that which I distinguish as Specimen IV .
This stem is much compressed , but otherwise well-preserved , and shows with perfect OF FOSSIL PLANTS FROM THE PALEOZOIC ROCKS .
clearness , on both sides , the external characters of the cortex , # agreeing in all respects with those of Specimen II .
, as represented in fig. 1 .
The base of a petiole is also shown , and the internal anatomy is identical with that of the other stems of M. cmglica .
It is in longitudinal sections , taken from the lower end of the specimen , that the adventitious roots appear .
One of these sections ( Plate 8 , photograph 24 ) passes almost radially through the undulating layer of periderm .
To two of its protrusions adventitious roots ( r ) are attached , the continuity of the tissues being uninterrupted .
Both these roots are cut in an obliquely tangential direction , the section passing at such a depth as to show the dense internal periderm of the root , outside which is the somewhat lax tissue of the primary cortex , containing numerous short secretory sacs .
Opposite the third peridermic protrusion , but not in connection with it , is an appendage seen in oblique transverse section , which is no doubt another root .
All these roots were evidently quite young , as shown by their intact primary cortex , though even at this stage periderm had begun to form .
From the root-bases strands of tracheides extend inwards into the stem , and attach themselves by means of an irregular network ( Reseau radicifere of Van Tieghem ) to a nearly vertical bundle , which , from its structure , is clearly a leaf-trace ( It .
) passing out gradually from the steal ( si .
) , which is seen further towards the interior .
The leaf-trace , as it diverges from the steal , remains connected with it by means of a system of short tracheides .
This radial connection between leaf-trace and steal ( referred to aboye , p. 93 ) has been observed in another specimen ( No. II .
) , where it was probably also connected with the presence of adventitious roots .
Provision appears to have thus been made for abundant water-communication between the roots and the stem which they supplied .
Another radial section of Specimen IV .
shows two cortical outgrowths in approximately median section ( Plate 8 , photograph 25 ) .
One of these contains a bifurcating strand of vascular tissue ; one fork is cut oft ' obliquely , but the other runs out into an evident root , of small size , cut sufficiently near the median plane to show the xylem .
In this case it is clear that the cortical outgrowth bore two roots , one above the other .
These outgrowths may best be regarded as the pedicels of the adventitious roots , t The other pedicel also shows a forked bundle , but less clearly .
Immediately outside this pedicel , and evidently belonging to it , a root is well shown in transverse section ( Plate 13 , fig. 22 ) .
The steal is asymmetrically triarch , just as in the young roots of Mr. Wild 's specimen above described .
In the present case , however , the broad zone of primary cortex is well preserved to the outside of the internal periderm .
* * These characters , so well marked in the fossil state , by no means represent the natural appearance of the plant during life .
We know that the ribs of sclerenchyma , which form the main feature of the fossil surface , were covered in nature by a layer of cellular tissue , which must have hidden them from view , or at most have allowed but slight elevations to appear externally .
t Cf .
Van Tieghem , " Origin des Membres Endogenes , " ' Ann. des .
Sci. Not .
( Rot .
) , ' Ser. 7 , vol. 8 , 1888 ; p. 540 .
The specimens described prove conclusively that triarch adventitious roots were borne on the stem of Medullosaanglica , and it is unnecessary to dwell on other cases in which the evidence is less complete .
It is interesting , however , to point out that these observations explain the nature of a very curious specimen , lent me some time ago by Mr. J. Butterworth , of Shaw , the significance of which long remained a problem .
A section of this specimen is represented in Plate 13 , fig. 23 .
It shows a four-lobed structure ; the lobes are in a row , and though in close lateral contact are only confluent with each other at the base ( towards the left of the figure ) .
The lobes are coated with periderm , the cortex within this is rich in " secretory sacs " ; three of the lobes contain each a steal , seen in obliquely transverse section , and in two this structure is very well preserved .
The resemblance of the lobes to the cortical excrescences or root-pedicels , as shown in Plate 8 , photographs 23 and 25 , is obvious .
The lobes of Mr. Butterworth 's specimens are rather larger than those compared with them , the vertical diameter of each lobe being about 6 millims. in the Butterworth specimen , and about 4 millims. in the others ; but in Specimen II .
of the Medullosa we find similar pedicels still larger than the former , so the differences in size go for nothing .
Of the two steles clearly shown , one ( that in b ) has a pith , or central parenchyma , in which , however , some tracheides are scattered .
Around this is a well-preserved zone of secondary wood and bast .
The other ( c , in Plate 13 , fig. 23 , shown in detail in fig. 24 ) has a fairly distinct triarcli primary xylem , with its tracheides much compressed , perhaps by dilatation of the adjoining parenchyma ; beyond this is a broad zone of secondary wood , succeeded towards the outside by cambium and phloem .
Fragments of leaflets , such as are usually associated with , occur in the matrix .
Mr. Butterworth kindly prepared five additional sections of the specimen , cut successively in planes parallel to the first .
The next section to that figured shows the steal of the lobe c losing its triarch character and acquiring a medulla containing scattered tracheae .
The succeeding section is still more important , for it shows two of the steles anastomosing in an oblique direction with a vertical vascular strand , in which the tracheae are pitted like those of our Medullosa .
Other sections , cutting still deeper into the parent organ , no longer show the pedicels as distinct organs , but pass through an irregular periderm , within which are fragments of secondary wood .
The preservation of this part is imperfect , but comparison with the sections already described , showing the insertion of the roots in Medullosa , leaves no doubt as to the nature of the specimen , which evidently consists of the bases or pedicels of a series of adventitious roots , in connection with a fragment of a Medullosa stem .
There is a decided general resemblance to the arrangement of the adventitious roots in Heterangium , as figured in a previous memoir , # but the agreement is much closer with our Medullosa .
Whether the Butterworth specimen is specifically identical with M. anglica may remain an open question . !
* Williamson and Scott , ' Further Observations , \amp ; c. , ' Part 3 , Plate 27 , fig. 28 .
t It is fair to my friend , Mr. Seward , to mention that he was the first to notice the resemblance of the OF FOSSIL PLANTS FROM THE PALAEOZOIC ROCKS .
Before leaving the roots , a word may be said as to their place of insertion on the stem .
There is nowhere any sign of their breaking through the strong armour of fibrous sclerenchyma by which the leaf-bases were protected .
Considering how much of the external surface is exposed in Specimens II .
and IV .
, besides what is shown in tangential sections , the passage of roots through the hypoderma , if it occurred , could scarcely have failed to come under observation .
In some cases , on the other hand , there is a distinct interruption of the hypoderma between the bases of adjacent leaves , as shown in Specimen II .
( Plate 9 , fig. 1 ) , and also in Specimen IV .
I have no doubt that it was at these places that the vertical series of adventitious roots were inserted , though , as the specimens are usually somewhat crushed , it has not been easy to determine the exact relation of the roots to the leaf-bases .
Some of the transverse sections of Specimen II .
, however , show a root embedded in the cortex , exactly between the bases of two leaves ( see Plate 5 , photograph 1 , and diagram , p. 126 , r ) , and all the other preparations showing roots are quite consistent with such a position .
We may sum up our knowledge of the roots of as follows ( 1 .
) The stem of Medullosa anglica bore numerous adventitious roots , ranged in vertical series , and escaping between the leaf-bases .
( 2 .
) The roots were borne on short pedicels , often two on each pedicel , and were connected with one of the steles of the stem , either directly or by means of - a leaf-trace bundle .
( 3 .
) The roots developed at different periods , for young and mature roots may occur on the same stem at about the same level .
( 4 .
) The roots were almost invariably triarch ; the steal formed a large amount of secondary wood and bast .
The primary cortex was cast oft* early , owing to the formation of periderm in the pericycle .
The periderm attained a great thickness in the older roots .
( 5 .
) The roots branched abundantly , the rootlets repeating in essentials the structure of the main root .
9 .
Summary and Diagnosis .
The account of the general structure of our plant given in Section 2 ( pp. 84-88 ) renders any detailed summary unnecessary .
It will suffice to recapitulate very briefly the chief points established .
Medullosa anglica ( like the other species of the genus ) had a polystelic stem , with secondary wood and bast developed around each steal .
Lutterworth specimen to the triarch roots of Medullosa anglica , when I showed him the sections at which I was working .
Mr. Butterworth himself was inclined to refer his specimen to turn\#151 ; a very reasonable supposition before the English Medullosa was known .
I may further mention that I was indebted to Mr. Butterworth for my first acquaintance with Mr. Wild 's specimen of the roots .
These two observers recognized that their respective specimens were of a similar nature .
In this species the steles are few ( 3 or 4 ) and uniform , having a somewhat elongated and irregular sectional form .
No " star-rings , " in the sense of Goeppert and Stenzel ( i.e. , small internal steles with circular transverse section ) are present .
The interior of the steles is occupied entirely by their primary wood .
The structure of each steal in Medullosa anglicaresembles in all points that of the single vascular cylinder of a Heterangium .
The steles divide and anastomose at long intervals .
They give off large leaf-trace bundles , which , on first leaving the steles , are concentric strands , with secondary wood and bast .
As they pass through the cortex , the leaf-traces divide repeatedly , lose their secondary tissues , and assume the collateral exarch structure of " My eloxylon " bundles .
Accessory vascular strands , somewhat similar to the normal steles , were occasionally developed in the cortex .
Externally the stem was clothed by large , spirally arranged , decurrent leaf-bases , having in all respects the structure of Myeloxylon , Ren .
The leaf-bases received numerous vascular bundles from the stem ; the vascular supply of each leaf consisted of the groups of bundles resulting from the subdivision of a number of the primary leaf-traces .
The surface of the fossil is marked by a system of fine longitudinal ribs , representing the fibrous hypodermal strands of the leaf-bases .
An internal periderm was developed around the stem , and may perhaps have ultimately caused the exfoliation of the leaf-bases and outer cortex .
The leaf , as shown by the ramification of the petiole and rachis , was a highly compound one .
The structure of the fine branches of the rachis and of the pinnules shows a general agreement with the corresponding parts of leaves , as described by Renault .
Gum-canals are numerous in the hypoderma and parenchyma of the leaf-bases , in the cortex of the stem , and in the pericycle of the steles .
They extended throughout the rachis of the leaf .
The stem bore numerous adventitious roots of triarch structure , with abundant secondary wood and bast , and an early development of internal periderm .
The roots branched freely .
For purposes of diagnosis , a still more concise abridgement of the characters must be attempted .
The specific value of the types of Medullosa , to which names have been previously assigned , is admitted to be doubtful .
Weber and Sterzel , * to whom we owe the best account of the genus , prefer to speak of " form-cycles " ( Formen-kreise ) rather than of species .
In the case of our English form there is no reason to doubt that all the specimens hitherto investigated ( with the possible exception of Mr. Butterworth 's fragment ) are specifically identical .
* ' Beitriige sir Kenntniss der Medulloseue , ' p. 88 .
OF FOSSIL PLANTS FROM THE PALEOZOIC ROCKS .
Ill Medullosa anglica , sp. nov .
Stem vertical , clothed by large spirally-arranged decurrent leaf-bases , perhaps cast off in old stems .
External surface of leaf-bases longitudinally striate .
Vascular system of stem consisting of a few ( usually three ) uniform steles , somewhat elongated and lobed , as seen in transverse section .
Star-rings absent .
Interior of each steal wholly occupied by primary wood .
Secondary wood and bast of moderate thickness , developed on all sides of the steles .
Tracheides usually with bordered pits .
Leaf-traces concentric on leaving the steles , branching and becoming collateral in traversing the cortex .
Leaf-bases and petioles with the structure of Ren .
Leaves highly compound .
Gum-canals abundant in the petioles and leaf-bases , and in the cortex , and around the steles of the stem .
Adventitious roots borne in vertical series , triarch , with secondary wood and bast , and periderm .
Stem , with leaf-bases , about 7-8 centims. in mean diameter .
Petioles , about 2*5-4 centims. in diameter at base , diminishing to about 1 millim. in the ultimate branches of the rachis .
Leaflets about 3 millims. wide .
Roots reaching 12 millims. in diameter .
Locality.\#151 ; Hough Hill Colliery , Stalybridge , Lancashire .
Horizon.\#151 ; Lower Coal-measures .
Found by Messrs. G. Wild and J. Lomax , 1892-1808 .
10 .
General Conclusions .
The specimens of Medullosa anglica just described appear to be the most complete examples of the genus hitherto investigated ( though the crucial point , the fructification , is unfortunately still lacking ) , and are certainly the most ancient .
They are , therefore , of considerable importance for the elucidation of the affinities of the group as a whole .
This is not the place to give a history of our knowledge of the Medullosese , but the chief steps in its progress may be indicated .
The first description of Medullosa is due to Cotta , * and dates from 1832 .
He described two species of the stem ( M. stand M. porosa ) , and , by an odd coincidence , included in his genus a third species , M. elegans , which , after having-long been separated as type of the genus , has eventually proved to be the petiole of a Medullosa .
Cotta 's inclusion of M. elegans in its right genus was * ' Die Dendrolithen in Bezug auf ihren inneren Bau , ' Dresden , 1832 .
no more than a lucky accident ; petiole and stem are , as we know , totally distinct in structure , and it was only owing to a complete misinterpretation that Cotta placed them together , for he confused the hypodermal sclerenchyma of the petiole with the peripheral wood of the stem , an error which Brongniart set right.# It was almost half-a-century before the botanical study of the stems made any essential advance , although additional specimens were , of course , described in the mean time .
In 1881 Goeppert and StenzelI showed that the whole vascular system of the stem , as seen in transverse section , consists of a number of closed rings , the peripheral wood being built up by a variable number of flattened " plate-rings , " differing from the central " star-rings " in size and shape , but not in structure .
This discovery showed at once that the structure of was essentially different from that of the stem of any recent Cycads .
J The organization revealed by the researches of Goeppert and Stenzel needed interpretation , and an important step in this direction was made by Schenk , who , in 1889 , discovered the presence of primary wood in the interior of each ring .
S His account , correct as far as it goes , of the structure and course of the leaf-trace bundles in M. stellata , also marks a real advance .
The interpretation of the structure of the stem as thus at once bringing it into line with the stem of most Ferns , is probably the most important step yet made in the anatomy of these fossils , and is due to Zeiller.|| The value of this conclusion lies in its giving a new clue to affinities , and is quite independent of the special importance we may attach to Van Tieghem 's anatomical terminology .
Zeiller 's interpretation received some support from the comparison between the structure of Medullosaand that of Psaronius , drawn by Weber and Sterzel,1T but its complete confirmation is due to Count Solms-Laubach .
This observer showed in 189G that in the Cladoxylon of Unger , a genus of Fern-stems from the Culm , each of the many steles may have its own zone of secondary wood.## The clear analogy with Medullosese is pointed out by the author , who in a more recent publication has shown that the primary wood of each steal in Medullosa Leuckarti may best be compared with that of the single steal of a Heterangium , Medullosa thus standing in the same relation to Heterangium as Primula Auricula , \amp ; c. , to the normal species of Primula , tf His description of M.Leuckarti is altogether the most complete that we have hitherto * * Tableau des Vegetaux Fossiles , ' Paris , 1849 , p. 60 .
t Goeppert u. Stenzel , ' Die Medullosese ; eine never Gruppe der fossilen Cycadeen , ' 1881 .
X Cf .
Solms-Laubach , ' Fossil Botany , ' 1887 , p. 102 ( English edition ) .
S A. Schenk , " Ueber A Tcdullosdu .
Tubicauli'Iv .
Sachs .
Gesellsch .
d. Wiss .
, Math. Phys. Cl. , ' vol. 15 , p. 535 , 1889 .
|| Lor .
tit .
, 'Flore Fossile d'Autun et d'Epinac , ' Part 1 , 1890 , p. 286 .
11 Lor .
tit .
, 'Beitr .
z. Kenntniss d. Medullosese , ' 1896 , p. 100 .
** Solms-Laubach , ' Pflanzenreste des Unterculm von Saalfeld , ' p. 55 , 1896 .
ft Solms-Laubach , " Ueber MedvllosaLeuckarti " 'Bot .
Zeit .
, ' 1897 , p. 182 .
OF FOSSIL PLANTS FROM THE PALAEOZOIC ROCKS .
possessed of any species of the genus , though certain points , such as the origin of the leaf-traces from the steles , could not be made out in this form .
An interesting addition to our knowledge of the anatomy of the genus was made by Weber and Sterzel , who showed that in one exceptionally large specimen ( M. stellata , var .
gigantea ) , and possibly in one other case ( M. Solmsii , var .
lignosa ) successive extrafascicular zones of wood and bast were developed outside the whole system of steles , so that in these cases the anomaly characteristic of zamia , and Encephalartos coexists with the polystely proper to the genus .
While the anatomy of the stem was thus being elucidated by the discoveries of successive observers , the nature of the leaf was also becoming known .
The discovery that Myeloxylon is the leaf-stalk of Medullo , originally made by Weber , was referred to by Schenk in 1882 , but the stem bearing the Myeloxylon was by him simply described as that of a Cycad.# The connection with Medullosa was recognised as probable by Count Solms-Laubach in 1887,1 and further confirmed by Schenk in 18894 Zeiller , however , in 1890 still treated the question as an open one .
The memoir of Weber and Sterzel ( 1896 ) for the first time gave the evidence fully , and , as it seems to me , established , beyond reasonable doubt , the fact that the species Medullosa bore as its leaf-stalks a Myeloxylon of the type M. Landriotii of Renault .
S The latest investigations of Count Solms-Laubach have fully confirmed their conclusion , though that author reserves his judgment as to the specific identification of the Myeloxylon in question .
|| The connection of Myeloxylon with Medullos having been finally established , the previous discovery by Renault that the Myeloxylon petioles belonged to compound leaves with the characters of Alethoptensand ,1 became of the greatest importance , as determining the nature of the foliage of Medullosa .
It now became clear , as Stur had previously suggested , that the fronds of these genera ( to which others may probably have to be added ) were not those of true Ferns , but of the peculiar semi-Cycadaceous family , Medulloseae .
We have now to consider the fresh light thrown upon the question of the affinities of Medullosese by the new species Medullosa , in which , for the first time , it has been possible to examine the entire vegetative structure of a plant of this genus in all its parts .
The first point of cardinal importance is the close agreement in stem-structure between our fossil and Heteran gium .
It is not a mere case of analogy , but of a * A. Schenk , " Ueber Medullosa elegans , " Engler 's ' Bot. Jahrbuecher , ' vol. 3 , 1882 .
t 'Fossil Botany , ' p. 161 ( English edition ) .
| ' Ueber Medullosa u. Tubicaulis .
' S Weber and Sterzel , loc. cit. , p. 43 ( 84 ) , Plates 4 , 5 , and 9 ; text-figs .
14-22 .
|| 'Ueber Medullosa Leuckarti , ' loc. cit. , p. 187 .
H Renault , ' Course de Botanique fossile , etc. , ' 3 , chaps .
12 and 13 ; Stur , ' Morph , u. Systematik der Culm und Carbonfarne , ' p. 638 , 1883 .
VOL. CXCI.---B .
Q detailed correspondence so exact as to be explicable only by a true affinity .
In the structure of the steal , as regards both primary and secondary wood , every detail agrees , including the mode of pitting of the tracheides , their distribution relative to the conjunctive parenchyma , the position of the protoxylem-groups , and the relation between the leaf-traces and the steal .
The is polystelic and the angium monostelic , but that difference , as we know from the case of Selaginella and other genera , is consistent with the closest affinity .
The structure of the hypoderma is another point , though one of secondary importance , in which our Medullosa and Heterangium agree , and the arrangement of the adventitious roots is also similar.* It is , however , on the agreement in the stelar structure that stress is to be laid .
This comes out much more clearly in M. anglica , even than in M. Leuckarti , partly because we know the details better in the former species , and partly because its structure is much simpler , and so allows of more direct comparison .
The affinity which , as it seems to me , is now demonstrated between Medullosa and Heterangium , is of much interest , for the genus Heterangium is one of the most Fernlike of the Cycadofilices .
Apart from the secondary thickening , the structure of the stem in Heterangium is closely similar to that in the recent while the foliage is altogether Fern-like , and in fact almost certainly identical , as Stur 's description shows , with that of Sphenopteris elegans .
t It would thus appear that , though so complex and highly specialized in many of its representatives , may well have been derived , by fairly direct descent , from a simple Fern-stock .
This does not , of course , imply that any known form of Heterangium was the ancestor of any Medullosa \#151 ; a most improbable supposition\#151 ; but that the Medullosese generally may have sprung from plants related to , and with the same type of stem-structure as Heterangium .
Such a view agrees better with the facts than would an attempt to derive the Medullosese from the Cladoxylese .
The latter , indeed , conspicuously resemble the former in having polystelic stems with secondary growth , but the structure of the individual steles is strikingly different , notably in the fact that the primary wood of consists exclusively of scalariform tracheae , without conjunctive parenchyma .
I should , therefore , incline to regard the Cladoxylese as representing a parallel line of development to that of the Medullosese , rather than as primitive members of the same series .
In the organisation of the leaves , and in the course and structure of the leaf-trace bundles , Medullosa anglica diverges widely from the Heterangium type , and stands on the same level with M.Leuckarti , assuming , as I think the evidence justifies , that the petioles of the latter really had the structure of Myeloxylon Landriolii .
It * The arrangement of the adventitious roots in Medullosa is also closely similar to that in some of the recent Tree-ferns .
See the figures of Cyathea in De Bary 's ' Comp. Anat .
, ' figs. 140 and 141 .
f Stur , " Die Culmflora , " ' Abhandl .
d. K. K. Geol .
Reichsanstalt zu Wien .
, ' vol. 8 , 1877 .
See also Kidston , " Fructification and Internal Structure of Carboniferous Ferns , " ' Trans. Geol .
Soc. Glasgow , ' 1889 , p. 49 .
I Solms-Laubach , ' Pflanzenreste des Unterculm von Saalfeld , ' p. 54 .
OF FOSSIL PLANTS FROM THE PALAEOZOIC ROCKS .
is rather surprising that our plant appears to have been quite as highly differentiated with respect to its leaves , as any of the Permian Medullosese , while its stem-structure was so much simpler .
This suggested at the outset a doubt whether the single specimen which had then been examined might not represent the first-formed , basal portion of a stem which had not yet attained its full complexity .
This idea , however , appears to me no longer tenable .
All the four specimens show just about the same degree of complexity , all alike having three ( or locally four ) similar steles .
And further , it is significant that no appreciable difference in structure is apparent between the upper and lower ends of the individual specimens .
In Specimen II .
the part actually sectioned was about nine inches long , and in Specimen I. nearly the same .
It must therefore , I think , be taken as proved that the structure of the stem in our species was really far simpler than in any other known , as regards the small number of the steles , and the absence of differentiation among them .
This fact agrees well with its greater antiquity .
Possibly some botanists might regard this difference as sufficient to justify generic separation .
I do not think so myself , bearing in mind the close agreement with other species of Medullosa in leaf-characters , and in the structure of the individual steles .
In both points the agreement appears to be closest with M. Leuckarti .
This leads us to consider the possible relation of Medullosa to the xylon aeduense of Brongniart , which is known to have much in common with M. Leuckarti , while showing a far simpler anatomical structure .
, to judge from the admirable figures and description in M. Renault 's splendid contribution to the ' Fossil Flora of Autun , ' # was monostelic in parts of the stem , the large single vascular cylinder having a very irregular outline .
In other parts there are two such cylinders , and their number may even reach six or seven at the summit of the stem .
M. Renault correlates this multiplication of the cylinders with a repeated dichotomy of the stem , a view which is not accepted by Count Solms-Laubach , who has examined the specimens . !
Scattered groups of tracheae are present in the interior of the cylinder ; " star-rings " are altogether absent , as is also the case in our species of Medullosa .
In many details , such as pitting of tracheides , structure of the leaf-trace bundles , and of the cortex , and in external characters , there is a striking resemblance to Medullosa anglica , and I do not doubt that the resemblance indicates real affinity .
Until , however , something more is known of the leaves of it would not be admissible to unite our species with that genus : of specific identity there can of course be no question .
It may be doubted whether Colpoxylon will ultimately maintain its claims to generic rank , though the attempt of Weber and Sterzel to identify if specifically with Medullosa Leuckarti seems quite inconsistent with its very distinct structure .
Tn so simple a form of Medullosa as M. anglica , the Fern-characters come out so * Loc .
cit. , vol. 2 , p. 299 , Plate 68 .
t 'Ueber M.Leuckarti ' p. 196 .
distinctly that one might be inclined to doubt , with M. Zeiller , whether these plants were after all anything more than peculiar and highly modified Ferns.# Yet even in this species clear indications of Cycadean affinities are present .
The whole structure of the petiole and leaf generally is strikingly like that of the Cycads ; just as much so , in fact , as in any other " M. " f The occasional formation of accessory bundles in the cortex of the stem is suggestive of Cycadean anomalies , comparable with those observed by Weber and Sterzel in other species of The structure of the roots , never before observed in any member of the group , weighs appreciably on the Cycadean side , though , on the other hand , the abundant production of adventitious roots is in itself a decidedly Pteridophytic character .
The course of the leaf-trace bundles\#151 ; a number of the principal traces passing out into each leaf , and at the same time increasing their number by repeated sub-division\#151 ; is quite peculiar , differing essentially both from the system in existing Cycadace8e , and from that in the extinct Bennettite0e.| A better analogy may be sought in those poly-stelic Ferns which have numerous leaf-trace bundles , markedly smaller than the steles of the stem .
S On the whole , however , it is evident that Medullosa anglica , like other Medullosese , combined Cycadean with Filicinean characters , though in this comparatively simple species the resemblance to Ferns is more prominent than in other members of the group .
The fact , now , as it seems to me , established , that the leaf-traces of Medullosa were concentric in structure on first leaving the steles , throws a welcome light on Seward 's Rachiopteris Williamsoni , a fossil petiole which that author separated ' from Myeloxylon on the ground , among others , that its vascular bundles are concentric and not collateral .
|| The general resemblance to Myeloxylon is manifest , and there is no reason to doubt that it indicates near relationship , for it is perfectly conceivable that in some of the Medullosese the foliar bundles may have retained concentric structure in the petiole , instead of losing it in their passage through the cortex .
Similar variations among recent Ferns will at once occur to any botanist .
The wood of the bundles in Rachiopteris Williamsoni is quite comparable in structure to the primary part of the undivided leaf-traces in Medullosa , and , as sections in my possession show , the pitting on the tracheae is of the same kind .
We may * * * S * Zeiller , 'Flore Fossile d'Autun et d'Epinac/ vol. 1 , p. 289 ; " Observations sir quelques Fougeres des Depots Houillers d'Asie Mineure , " ' Bull .
Soc. Bot. de France/ vol. 44 ( 1897 ) , p. 201 .
t I have not found anything but exarch bundles ( with centripetal wood only ) in the leaf-stalks .
In one or two cases mesarch ( diploxylic ) bundles have been described in other " Myeloxylon " petioles and in the cortex of Colpoxylon , but as a rule the structure is exactly as in our species .
t Solms-Laubach , " Die Sprossfolge der Stangeria , " 'Bot .
Zeitung./ 1890 , p. 177 .
S Of .
De Bary , ' Comp. Anat .
of Phanerogams and Ferns/ English Edition , pp. 284-296 .
|| Seward , " On Rachiopteris Williamsoni , a new Fern from the Coal-Measures , " ' Annals of Botany/ vol. 8 , 1894 , p. 207 .
OF FOSSIL PLANTS FROM THE PALAEOZOIC ROCKS .
reasonably suppose , then , that Rachiopteris was the petiole of some unknown Medullosean stem .
If we now compare the organization of our with that of the species previously known , we find that the essential difference consists in the fact that in M. anglicathe few steles are similar to one another , while in the other species they are differentiated into a central and a peripheral system.* Polystely together with secondary thickening is a somewhat unpractical combination , which could never have reached any high development in the crude form in which it exists in Medullosa anglica .
The steles , as they grew in thickness on all sides , must ultimately have got in one another 's way .
The structure , as shown in all the specimens investigated , is , in fact , not adapted to the attainment of arboreal habit .
The difficulty is met in the four species , M. stellata , Cotta ; M. porosa , Cotta ; Schenk ; and M. Leuckarti , Gopp .
et Stenz .
, by the differentiation of a distinct system of peripheral steles forming a circle round the stem.f It was on the free outer side of these steles , where the space for new growth was unrestricted , that the chief formation of secondary wood and hast took place ; on the inner side of the peripheral ring , and around the smaller internal steles , the growth in thickness was necessarily limited .
Where , as in old specimens of M.stellata ( var .
of Weber and Sterzel ) , the external secondary zone attained an enormous relative thickness , | and especially where the " pith " of the stem was small to begin with , the Medullosa characters sink into the background , and the structure might almost be taken for that of a somewhat anomalous stem of the ordinary Cycadean type .
Can we regard such an organization as having really led on to the stem-structure of true Cycads ?
This hypothesis has been clearly stated by Mr. W. C. Worsdell , with special reference to the relation between Medullosese and Cycadacese .
He is led to the " provisional suggestion , " " That the anomalous structures in Cycadean stems ... ... ... ... ... .are remnants of some ancient structure once common to a large group of plants .
That this structure consisted of rings or layers of concentric vascular strands .
That , as time went on ... ... .the cambium of the inner portion of each such concentric strand gradually became less and less functional , that of the outer portion , on the contrary , more and more active , " and so on .
" The result , " he adds , " is finally the structure , as we at present know it , in the stem of Cycas , JEncephalartos , and Macrozamia " S Mr. Worsdell is speaking especially of the anomalous zones of wood and bast in those genera , but would , I believe , extend his theory to the normal vascular ring .
Potonie arrived independently at a similar conclusion .
|| Personally I regard these * Compare the illustrations of the various types in Weber and Sterzel 's monograph .
t The peripheral ring often becomes gamostelic in Van Tieghem 's sense , and in some cases appears to have constituted a single annular steal , like that in the section of Dicksonia , and many other Ferns .
| See , e.g. , Weber and Sterzel , loc. cit. , Plate 7 , fig. 3 .
Here only one internal " star-ring " is present .
S W. C. Worsdell , " Anatomy of the Stem of Macrozamia " 'Annals of Botany , ' vol. 10 , 1896 , p. 618 .
|| ' Lehrbuch der Pflanzenpalseontologie , ' 1897 , p. 168 .
views as fallacious , because the primary ground-plan of the stem-structure of a polystelic Medullosa was fundamentally dilferent from that of the monostelic Cycadaceae , though the ultimate result , in old stems , may have a certain similarity .
If we are to suppose that the Cycadaceous type of stem came from that of the Medulloseae at all ( which is not my own view ) , I should prefer to derive it from some monostelic form , such as we find indicated in Colpoxylon .
It is not impossible that such a stem , by gradual disappearance of the scattered internal tracheides , might have given rise to Cycadean structure .
Extrafascicular zones occur in the same form in some Medulloseae , as in certain recent Cycads , so it appears unnecessary to derive this part of the structure from a reduced system of rings .
The opinion is expressed by Count Solms-Laubach that the Medulloseae may probably form a group dilferent at once from the Ferns and from the Cycads , and may possibly represent the last derivative ( becoming extinct in the post-carboniferous strata ) of a series springing from the common parent-stock of those two classes.* On present evidence I am disposed to accept this view of the Medulloseae as a short divergent branch of the phylogenetic tree .
That they have real affinity with Cycads ( in the wide sense ) is certain , but it does not seem probable that they lie on the direct line of descent of any of the later Cycadales at present known to us .
As regards the " common parent-stock " referred to by Count Solms , I think it is probably to be sought among simple Ferns , or Fern-like plants , with a structure more like that of Heterangium than of any other known genus .
As regards the origin of the later groups of the Cycadales , I still think that Lyginodendron affords the best clue ; that is to say , I regard plants with that type of stem as nearer the direct line of descent than the Medulloseae .
From we may perhaps trace two divergent series , the one advancing in the direction of polystely through such a form as Medullosa to the more complex Medulloseae , the other advancing by elaboration and modification of the single steal , through Lyginodendron Oldhamiumto L. robustum , Cycadoxylon , and Ptychoxylon .
f The two last genera may well have been already true Cycadales , with foliage perhaps of the Pterophyllum or Sphenozamites type , and the fructification of Gymnosperms .
It is useless at present to speculate on what the fructification of the Medulloseae themselves may have been like , but it is quite possible that they also had attained to the formation of some kind of seed .
I have not thought it necessary to discuss the views of M. Renault , who still appears inclined to derive the whole of his Cycadoxyleae ( including Medulloseae ) from Lycopods rather than from Ferns .
Considering the clear points of agreement which Heterangium , Lyginodendrony and Medullosa all present with the Ferns , in the structure of the stem , the structure and form of the leaf , and in habit , while there are * 'Ueber Medullosa LeucJcarti , ' p. 198 .
t See Renault 's account of the Cycadoxyleae in vol. 2 of the 'Flore Fossile d'Autun et d'Epinac .
' Also , Seward , " A Contribution to our Knowledge of Lyginodendron " 'Annals of Botany , ' vol. 11 , 1897 .
OF FOSSIL PLANTS FROM THE PALAEOZOIC ROCKS .
no clear points of agreement with Lycopods , I cannot regard this view as consonant with the present state of our knowledge .
It is very interesting to find how the evidence is gradually accumulating which connects the Cycads ( and by a less direct line , probably the other Gymnosperms also)# with the Ferns .
The fossil evidence , very strong by itself , is supported by the morphology of the reproductive organs , and by the great discovery of the multiciliate spermatozoids which we owe to Ikeno and Hirase , and to Webber .
We thus see that external resemblance is not always fallacious ; the Fern-like foliage of or of the " Maidenhair Tree , " pointed in the right direction long before botanists were able to follow the clue .
The illustrations to this paper are partly photographic , partly from drawings .
Of the photographs , some were taken for me by Dr. E. C. Bousfield , others by Mr. L. A. Boodle , while the drawings are the work of Mr. Boodle and of Mr. G. T. Gwilliam .
The exact share taken by these three gentlemen in the illustration of the memoir is stated in the explanation of the plates .
I should like especially to express my thanks to Mr. L. A. Boodle for the help he has given me in various ways during the preparation of the paper .
Explanation of Plates .
Plates 5-8 , photographs from the actual sections ; many of these need to be examined with the aid of a lens .
Photographs 2 , 5 , 6 , 7 , 8 , 9 , 10 , 12 , 13 , 14 , 15 , 19 , 20 , and 21 are by Dr. E. C. Bousfield ; the remainder by Mr. L. A. Boodle .
PLATE 5 .
Phot .
1 .
Complete transverse section of Specimen II .
, taken from the lower part of the specimen , immediately below the portion shown in Plate 9 ; the three leaf-bases .
At a higher level a new leaf-base ( be ) , shown in surface-view in Plate 9 , fig. 1 , and in section in photograph 3 , appears between b and c , and still higher another ( ab ) arises between a and b. The order of the leaves from below upwards is b , be , ab , indicating a f phyllotaxis .
pd .
, internal zone of periderm ; within this the three steles , each with primary and secondary wood , are evident .
X If .
S. 737.t ( See p. 84 .
Compare diagram on p. 126 .
) Phot .
2 .
General transverse section of Specimen I. Pet. , part of a large petiole , just becoming free from the stem , o.c. , outer cortex or hypoderma of another leaf-base ; ct .
, cortical parenchyma ; on this side the hypoderma is lost ; * Cf .
Coulter , " Origin of Gymnosperms , " 'Bot .
Gazette , ' vol. 26 , 1898 .
f Figures preceded by the letter S indicate the number of the slide in the author 's collection .
.si1 , st.2 ' , stz ' , the three steles ; is almost perfect , the others more broken .
X 2 .
S. 226 .
( See p. 86 .
) Phot .
3 .
Part of a transverse section of Specimen II .
, at the level 3 marked in Plate 9 , showing the leaf-base ( of which the external surface is seen in Plate 9 , fig. 1 ) , and part of b ( phot .
1 ) .
The leaf-base shows the characters of Myeloxylon LandrPen .
b. , one of the vascular bundles ; m.c. , a gum-canal ; p part of the periderm of stem ; sc .
, internal sclerenchyma between the leaf-base b , and stem .
X 3 .
S. 740 .
( See p. 100 .
) Phot .
4 .
Part of a transverse section of Specimen I. , at a level below that of phot .
2 , and above that of phot .
5 .
The steal st.1 is here divided into two parts ; cm .
, accessory vascular strand outside st.1 ; st.2 ' , part of the next steal .
A number of leaf-trace bundles are also shown .
X 2|- .
S. 578 .
( See p. 91 .
) PLATE 6 .
Phot .
5 .
Transverse section of Specimen I. , including all three steles .
Right and left-hand reversed as compared with phots .
2 and 4 .
The level is about 2 centims. below that of phot .
4 , and about 4 or 5 centims. below that of phot .
2 .
The two halves of st.1 have reunited .
At its right hand end the phloem is preserved .
The other steles are also well shown , and their limits evident , in spite of fractures .
o.c. , hypoderma of a leaf-base , identical with Pet. in phot .
2 ; sc .
, its internal sclerenchyma ; numerous leaf-trace bundles are shown ; v.b. , the two bundles shown enlarged in phot .
14 .
For details , see phots .
7 and 9 .
X nearly 3 .
S. 579 .
( See pp. 86 and 91 ) .
Phot .
6 ( Specimen I. ) .
Part of steal 1 in transverse section .
x , primary wood , consisting of tracheides and conjunctive parenchyma ; secondary wood ; l.t. , leaf-trace bundle just separating from steal ; it has a fan of secondary wood on its outer side .
X 18 .
S. 581 .
( See p. 92 .
) Phot .
7 ( Specimen I. ) .
The same part , from a section about 1 centim. higher up the stem , px .
, probable position of a protoxylem group ; l.t. , the leaf-trace shown in phot .
6 , which has here become free from the steal , and is surrounded by secondary wood .
X 18 .
S. 579 .
( See p. 92 .
) Phot .
8 ( Specimen I. ) .
Parts of steles 1 and 2 in transverse section ; primary and secondary wood evident , and phloem of steal 2 partly preserved .
leaf-trace beginning to detach itself from steal 2 ; large leaf-trace already free .
Other leaf-traces are shown .
X 10 .
S. 581 .
( See pp. 92 and 94 ) .
Phot .
9 ( Specimen I. ) .
The same part , from a section about 1 centim. higher up the stem .
The leaf-trace , l.t.1 , is here just free from its steal ; is beginning to divide .
Other leaf-traces shown .
X 10 .
S. 579 .
( See pp. 92 and 94 ) .
OF FOSSIL PLANTS FROM THE PALAEOZOIC ROCKS .
Phot .
10 ( Specimen I. ) .
Radial section from a series cut a little below the transverse sections figured in Plate 6 , crossing steles 1 and 2 at right angles .
shown in tangential section through the secondary wood ; , shown in radial section .
Leaf-traces are seen in the cortical tissue on either side .
X 3 .
S. 596 .
( See p. 95 .
) Phot .
11 .
From an adjacent section to the preceding , showing part of steal 2 in radial section , with a leaf-trace , passing out from the steal .
X 4 .
S. 595 .
( See p. 93 .
) Phot .
12 ( Specimen I. ) .
Part of a longitudinal section , from a series cut between transverse sections 578 ( phot .
4 ) and 226 ( phot .
2 ) , parallel to long axis of steal 1 , a part of which is shown in radial section .
primary wood , showing the strands of tracheides and conjunctive tissue ; x2 .
, secondary wood in radial section .
The arrow marks the limit between the two .
X 10 .
S. 576 .
( See p. 88 .
) Phot .
13 .
Secondary tracheides from the same radial section , to show bordered pits .
X 90 .
( See p. 90 .
) Phot .
14 ( Specimen I. ) .
Part of the transverse section 579 , to show the two collateral leaf-trace bundles marked v.b. in phot .
5 .
xylem of one of the bundles , with its protoxylem towards the phloem , ; cortical tissue of stem and a fragment of internal sclerenchyma ( sc .
) also shown .
X 35 .
( See p. 94 .
) Phot .
15 ( Specimen I. ) .
Two similar bundles from the base of a leaf , x. , xylem ; ph .
, phloem of one of the bundles .
X 55 .
S. 581 .
( See p. 95 .
) Phot .
16 .
Part of a transverse section from a detached piece of stem , probably forming part of Specimen IV .
, showing the axil of a leaf , st. , part of a steal ; pd .
, periderm of stem ; o.c. , hypoderma , in oblique section , continuous with the torn hypoderma of the petiole , pet .
X lyg .
S. 792 .
( See p. 100 .
) Phot .
17 ( Specimen II .
) .
Tangential section , from a longitudinal series , cut immediately above the level 4 at top of fig. 1 , Plate 9 , and parallel to the free surface of specimen .
The section passes chiefly through the hypoderma , showing " Sp\gt ; arganum"structure and gum-canals .
To the right it cuts more deeply into the tissues , showing the parenchyma and vascular bundles ( v. b.)of a leaf- base .
X If .
S. 742 .
( Seep .
99 .
) PLATE 8 .
Phot .
18 .
Partial transverse section of a free petiole from the same block as Specimen I. Structure as in Myeloxylon Landriotii , Ren .
me .
, gum-canals , also conspicuous in hypoderma .
v.b. , collateral vascular bundles .
X 4 .
S. 685 .
( See p. 101 .
) VOL. CXCI.\#151 ; B , B , Phot .
19 .
Transverse section of a large triarch root , the three prominent protoxylem-angles , two of which have broken away from the steal .
In the middle is the primary wood , with secondary wood and bast between the protoxylem-groups .
rt .
, vascular strand of rootlet , connected with \#166 ; px .
; pd .
, periderm .
X 8 .
S. 5G7 .
( See p. 102 .
) Phot .
20 .
Longitudinal section of two roots , fd and r.2 , the former complete , but somewhat compressed , pd .
, periderm ; x. , wood of root .
The secondary wood is partly in radial section , showing pits and medullary rays .
X 20 , S. 569 .
( See p. 103 .
) Phot .
21 .
Transverse section of triarch rootlet , showing the steal , with primary and secondary xylem , phloem-groups , pericycle , phellogen , and broad zone of periderm .
At rt a lateral rootlet branches off , opposite one of the protoxylem-angles .
X 50 .
S. 569 .
( See p. 104 .
) Phot .
22 .
Tangential section from a fragment of stem , showing the bases of four adventitious roots ( rd-rd ) .
M.c. , cortex , with Medullosa structure , connected with the roots on either side .
X 2 .
From outermost section of Mr. Wild 's specimen , lent by him .
( See p. 105 .
) Phot .
23 .
Part of another tangential section of Mr. Wild 's specimen , cut further to the interior , st. , a steal of the stem , with which the roots rd-rd in phot .
22 have united ; pad .
, series of root-pedicels , in oblique section ; r. , two young roots , one free , the other attached to a pedicel .
These roots are shown on a larger scale in Plate 13 , fig. 21 .
X 3^ .
S. 720 .
( See p. 106 .
) Phot .
24 .
Part of a radial section from Specimen IV .
of anglica , to show insertion of adventitious roots , st. , a steal of the stem ; l.t. , leaf-trace bundle connected with steal by radial system of tracheides ; pd .
, periderm ; r. , young roots in tangential section , connected with cortex ; rd , free root .
X 4 .
S. 782 .
( See p. 107 .
) Phot .
25 .
From another radial section of Specimen IY .
st. , part of a steal ; ped .
, root-pedicel in median section , showing bifurcating bundle ; r. , young root borne on pedicel ; rd , larger root , in transverse section , shown on an increased scale in Plate 13 , fig. 22 ; o.c. , sclerenchyma of a leaf-base .
X 4 .
S. 783 .
( See p. 107 .
) Plates 9-13 .
Figures from drawings .
Figs. 1 , 2 , 4 , 8 , 10 , 12 , 14 , 18 , and 20 are by Mr. G. T. Gwilliam ; the remainder by Mr. L. A. Boodle .
PLATE 9 .
Fig. 1 .
External view of Specimen .
II .
, showing all that part of the block in which the surface of the Medullosa is exposed .
2 , 3 , 4 , planes from which transverse sections have been prepared .
Phot .
1 is from a section cut OF FOSSIL PLANTS FROM THE PALAEOZOIC ROCKS .
at the top of the lowest piece ( not shown ) , immediately below the plane 2 .
The longitudinal sections were cut just above 4 , from a piece not shown .
a. , external surface of a leaf-base which extends the whole length of specimen ( cfphot .
1 ) .
In the upper part it is much flattened out by pressure , c. , inner concave surface of another leaf-base , which had split away ( see crack in phot .
1 ) .
Note the finer striation of the internal sclerenchyma .
Between a. and at the bottom , the inner cortex is exposed , be .
, a new leaf-base , which appears about at the line Z ; its external surface is convex .
This leaf-base is shown in transverse section , at the plane 3 , in phot .
3 .
The leaf-base , , shown in photograph 1 , is not seen here , as it lies at the back of the specimen .
Natural size .
( See p. 84 .
) PLATE 10 .
Fig. 2 .
Diagrammatic transverse section of Specimen I. , at a level just below that of phot .
5 .
st.1 , st.2 , and st,3 , the three steles , as shown in detail in phot .
5 .
l.t.1 , large undivided leaf-trace bundles ; smaller leaf-traces after sub-division ; many of both kinds are shown , hypoderma ; sc .
, internal sclerenchyma .
The triangular space between o.c. and sc .
( towards C ) is part of a leaf-base , identical with the petiole shown , at a higher level , in phot .
2 .
On the side A the hypoderma is lost , hut the ground-tissue better preserved .
On the side B is a part of another leaf-base .
The ground-tissue is shaded , where preserved .
X 2 .
S. 580 .
( See p. 86 .
) Fig. 3 ( Specimen I. ) .
Transverse section of secondary wood and bast of steal 3 .
x.2 , secondary wood ; ph.2 , secondary phloem , the larger elements probably the sieve-tubes ; slight remains of cambium are present between wood and phloem .
X about 50 .
S. 578 .
( See p. 89 .
) Fig. 4 ( Specimen II .
) .
Badial section through part of the primary and secondary wood of a steal , x. , primary wood , with tracheides and abundant parenchyma .
One primary tracheide is pitted , the rest spiral or scalariform .
x.2 , secondary wood , showing medullary rays .
X 35 .
S. 751 .
( Seep .
89 .
) Fig. 5 ( Specimen I. ) .
Badial section through part of the primary and secondary wood of steal 1 .
px .
, spiral elements of protoxylem ; inner part of primary wood ; some tracheides pitted ; parenchyma only partly preserved ; x.2 , secondary wood ; bordered pits of 'tracheides shown in places .
X about 110 .
S. 597 .
( Seep .
89 .
) Fig. 6 ( Specimen II .
) .
Part of periderm of stem in radial section , with phellogen and cells of primary cortex to the left .
The transverse section is quite similar .
X 43 .
S. 747 .
( See p. 96 .
) Fig. 7 ( Specimen III .
) .
Small portion of periderm of stem , in tangential section .
X 43 .
S. 770 .
( See p. 96 .
) PLATE 11 .
Fig. 8 ( Specimen II .
) .
Tangential section of secondary wood .
tracheides ; medullary rays .
X 44 .
S. 750 .
( See p. 89 .
) Fig. 9 ( Specimen I. ) .
Tangential section of phloem from steal 1 .
strands of sieve-tubes in secondary phloem ; phloem-rays ; gum-canal in pericycle ( ?
) ; hy .
, fragments of fungal hyphae .
X 40 .
S. 595 .
( See p. 90 .
) Fig. 10 ( Specimen I. ) .
Transverse section of a large undivided leaf-trace bundle , given off from steal 1 .
x. , primary xylem , consisting of large tracheides , with a little conjunctive parenchyma ; protoxylem ; secondary wood , extending all round , except for accidental ruptures .
On the border of primary and secondary wood , the pitted walls of short tracheides are seen .
pli .
, phloem , imperfectly preserved .
X 50 .
S. 582 .
( See p. 92 .
) Fig. 11 ( Specimen I. ) .
Padial section of the large leaf-trace , shown in transverse section in phots .
8 and 9 .
x. , primary wood of leaf-trace , consisting chiefly of pitted tracheides , becoming scalariform and spiral towards the protoxylem , px .
; x.2 , secondary wood , of very unequal thickness on the two sides .
Towards the left many short tracheides are shown .
X about 45 .
S. 601 .
( See p. 93 .
) Fig. 12 ( Specimen I. ) .
Transverse section , showing a number of small leaf-trace bundles , probably belonging to a leaf-base , as they lie outside the internal sclerenchyma , sc .
The two groups v.b.1 and v.b.2 each arise from the subdivision of a single strand .
pddouble layer of periderm enclosing part of the sclerenchyma .
X 12 .
S. 580 .
( See p. 94 .
) Fig. 13 ( Specimen I. ) .
Small portion of the external tissue of a leaf-base , in transverse section .
cu .
, probably cuticle .
The epidermis and next layer have palisade form , sc .
, part of hypodermal sclerenchyma .
X 88 .
S. 735 .
( See p. 99 .
) PLATE 12 .
Fig. 14 ( Specimen I. ) .
Transverse section of the hypoderma of a leaf-base .
The outer surface ( from which .the external tissues have perished ) is directed upwards .
The fibrous strands and intermediate parenchyma are obvious .
m.c. , gum-canals .
X 23 .
S. 580 .
( See p. 99 .
) Fig. 15 ( Specimen I. ) .
Padial section of a collateral vascular bundle from a leaf-base .
x. , xylem , containing scalariform and spiral tracheides only ; px .
, protoxylem ; pli .
, phloem .
X about 85 .
S. 588 .
( See p. 95 .
) Fig. 16 .
Transverse section of the rachis of a pinna , from same block as Specimen I. OF FOSSIL PLANTS FEOM THE PALAEOZOIC EOCKS .
Note the wings on upper surface ; P.5 .
, vascular bundles , forming a ring , with phloem outwards , around a gum-canal ; other gum-canals .
The hypoderma is almost continuous .
X about 50 .
S. 687 .
( See p. 101 .
) Fig. 17 .
Transverse section of a leaflet ( associated with rachis , petioles and stem of Specimen I. ) .
v.b. , vascular bundles .
The two in the midrib have their phloem directed laterally ; those in the lamina are in longitudinal or oblique section .
Note the palisade tissue towards upper surface .
X 36 .
S. 691 .
( See p. 101 .
) Fig. 18 ( Specimen II .
) .
Accessory vascular strand from a transverse section of the stem , cut at the level 2 ( cf. Plate 9 , fig. 1 ) .
tr .
, tracheides in the central parenchyma , x.2 , secondary wood ; some of the tracheides are horizontal .
ph .
, remains of phloem and cambium ; i.c. , inner cortex .
X 28 .
S. 738 .
( See p. 98 .
) Fig. 19 .
Transverse section of small rootlet . .
, the three protoxylem-groups .
pd .
, pericyclic periderm beginning to develop .
Beyond this are the endodermis and primary cortex , as described in text .
X 90 .
S. 696 .
( See p. 104 .
) PLATE 13 .
Fig. 20 .
Transverse section of a large triarch root , associated with Specimen I. the three protoxylem-angles .
To the interior is the primary wood .
x2 , secondary wood , interrupted opposite the protoxylem-groups ; ph .
, phloem ; pd .
, periderm .
X 25 .
S. 731 .
( See p. 102 .
) Fig. 21 .
From Mr. Wild 's specimen , showing the two roots marked in phot .
23 .
One root is free , the other is still seated on its pedicel , which forms part of the cortex of the stem , px .
, the three protoxylem groups of the two triarch roots .
( The long axis of the xylem is vertical in nature .
) pd .
, periderm ; c. , primary cortex of the free root , partly cast off ; rt .
, rootlets .
( In the free root the parenchymatous cells are shown rather too small and numerous .
) X 19 .
S. 720 .
( See p. 106 .
) Fig. 22 ( Specimen IV .
) .
Part of the root marked in phot .
25 , in transverse section .
px .
, the three protoxylem-groups of the asymmetric wood .
In the natural position the long axis is vertical , pd .
, internal periderm .
Beyond this is primary cortex , with " secretory sacs .
" X 36 .
S. 783 .
( See p. 107 .
) Fig. 23 .
Section of Mr. Butterworth 's specimen , showing four root-pedicels , a , b , c , d. In b , the steal has an apparent pith ; in it is triarch .
The pedicels are coated by periderm , and united at their bases .
My .
, associated Myeloxylon leaflets .
X 4| .
( See p. 108 .
) Fig. 24 .
Steal of pedicel c in the last figure , shown in obliquely transverse section .
px .
, one of the three protoxylem-angles of the triarch primary xylem ; x.2 , secondary wood ; ph .
, phloem .
X about 30 .
( See p. 108 .
) o.c. Medullosa anglica.\#151 ; Diagrammatic transverse section of Specimen II .
, to serve as a key to photograph 1 .
a , b , c , the three leaf-bases ; be , ab , position of the next two leaf-bases above ; the three steles , each with primary and secondary wood ; l.t. , leaf-traces given off from the steles ; accessory vascular strands ; pd , zone of periderm ; pd ' , isolated rings of periderm , one of which encloses an accessory strand ; o.c. , outer cortex or hypoderma ; sc , internal sclerenchyma delimiting the leaf-base c , on the inner side ; r , root passing out between two leaf-bases .
Compare Plate 5 , photograph 1 , which is from the same section , and on the same scale .
In the diagram the relative , but not the absolute , number of vascular bundles in the three leaf-bases is indicated .
The order of insertion of the leaf-bases , from below upwards , is c , b , be , ab , which corresponds to a f phyllotaxis .
Phil. Trans. , B , Vol. 191 , 1899 , Plate fj .
Scott .
Phot .
3 Phot .
4 msa mm BmSSJ Sm .
SIP r ^ 'SShLj MWM ||S|p \#187 ; r H ^Bw WMm ||p|ig|TC s SpB !
q| I yfcp* 11MW Medullosa Anglica .
B'n$field and Boodle , phot .
Jr ; -/ - 31J .V i \#166 ; ; /- , .
\#166 ; ' ' \#166 ; .
, vyf$$ " .
; V \#166 ; .
v \#166 ; v ' D ff .
Scott .
Phil. Trans. , B , Vol. 191 , 1899 , Plate 6 .
o. H. Scott .
Phil. Trans.j B , Vo/ .
191 , 1899 , 7pHOT.ll Phot.12 .
Vnoril Phot .
17 .
Phot.16 .
Medullosa Anglica .
o.c , Bousfield and Boodle , phot .
Phot .
19 fi f rt\#151 ; PhiLTrans .
, B , Vol. 191 , 1899 W12 .
M.c. Phot .
EI .
Phot .
2.3 .
PhoT.25 M.c. Bousfield \amp ; Boodle , phot .
Medullosa Anglica .
D. H. Scott .
Phil. Trans. , B , Vol. 191 , 1899 , 9 .
Medullosa Anglica .
||| GwiUiam , del .
I H. Scott .
Phil. Trans. , B , Vol. 191 , 1899 , Plate 10 .
Medullosa Anglica .
t L ' A. Boodle and G. T. Gwilliam , del Q / / .
Scott .
wm Phil. Trans. , B , Vol. 191 , 1899 fi .
t T Fi , .
12 .
i Boodle and G. T. Gwilliam , del , Medullosa Anglica .
j ) , B. Scott .
Phil. Trans. , B , Vol. 191 , 1899 , 12 .
fn .
c Medullosa Anglica .
L ' A ' Bood\gt ; e and G. T. Gwilliam , del .
J ) .
H. Scott .
B , Vol. 191 , 1899 , Plate 13 L. A. Boodle and .
G. T. Gwilliam , del .

