Echinoderm Larvae with Two Water-Vascular Systems .
323 direct improvement of conductivity or ( ?
\gt ; ) secondarily through the slowing of the rate of succession which they may induce , i.e. by lowering excitability , provided that this effect is not attended by a proportionate lowering of conductivity .
Adrenalin is notably useful in this respect , as indicated by the remarkable improvement in conduction often seen under its influence , especially evident in the auricles , where a strikingly slow contraction wave , present during gravely depressed conduction , may be replaced by an approximation or a return to the normal type .
Hence the special utility of adrenalin in dealing with forms of slow coarse fibrillation , already described , and also with fibrillar beats\#151 ; unless the damage in the latter case has been carried to an irreparable stage ( fig. 9 ) . .
, * The success of the above-mentioned methods of obtaining recovery from typical fibrillation , induced by means that did not permanently damage the heart , has been such* that in recent years of experimentation there has not been failure in any instance .
For valued assistance in some of the experiments of this investigation , I have to record my thanks to Drs. G. Spencer Melvin and J. R. Murray .
A portion of the costs was defrayed by a grant from the Carnegie Trust .
The Artificial Production of Echinoderm Larvae with Two Water-Vascular Systems , and also of Larvae Devoid of a Water-Vasdular System .
By E. W. MacBride , F.R.S. ( Received January 18 , 1918 .
) [ Plates 4-10 .
] The development of Echinoderms has been characterised , and with justice , as the most remarkable ontogenetic change in the animal kingdom .
For the larva is an almost perfect example of a simple , bilaterally symmetrical Metazoon , and the amazing thing is , not that the radially symmetrical adult should develop out of a bilaterally symmetrical larva , but that the axis of symmetry of the radial adult should cut the principal axis of the bilateral larva at an angle which approaches 90 ' .
In the three orders Asteroidea , Ophiuroidea , and Echinoidea , the general vol. xo.\#151 ; b. 2d 324 Prof. E. W. MacBride .
The Artificial Production of anatomy of the early larva is of the same type .
In all three groups the larva possesses a simple alimentary canal , consisting of a conical oesophagus opening by a wide mouth , a globular stomach , and a sac-like intestine opening by a narrow anus and directed forwards , so that the whole alimentary canal has the form of a U- On each side of the oesophagus a flattened coelomic sac is situated ; of these , the left sends up a vertical outgrowth termed the pore-canal , which fuses with the dorsal ectoderm , and opens to the exterior by a pore called the madreporic pore .
Each coelomic sac subsequently grows backwards , so that its posterior portion lies beside the stomach , and this portion later becomes separated by a constriction from the rest .
Consequently , each sac becomes divided into an anterior and a posterior coelom .
In all three orders , when this stage has been reached , a vesicle is budded off from the posterior end of the anterior coelom on the left side , which is termed the liydroccele .
This vesicle dominates the whole later development of the larva , and eventually becomes transformed into the water-vascular system of the adult .
It gives off lobes , or primary tentacles , which become converted into the radial vessels of that system , and it becomes bent into the form of a hoop , which is then converted into a ring by the meeting of the two ends , and this ring is the water-vascular ring-canal .
The axis of adult symmetry , as all know , passes through the centre of this ring .
At metamorphosis , portions of the larval body are cast off , consisting mainly of the ciliated band , with its various outgrowths or " arms , " which constitute the organs of locomotion of the larva .
The left side of the larva becomes converted into the ventral or actinal side of the adult which bears the tube-feet , whereas the right side of the larva becomes changed into the dorsal or abactinal side of the adult , developing spines and pedicellarise as the case may be .
It is obvious that , if not only the lbut also the right , anterior coelom , were to bud off a hydrocoele , the bilateral symmetry of the larva would be retained , and the whole subsequent course of the development would be profoundly modified .
Johannes Muller figured , without describing it , an Ophiurid larva , with a five-lobed hydrocoele on both sides of the oesophagus ( 16 ) .
Later , Metschnikoff ( 15 ) , who investigated the development of the viviparous Ophiurid Amphiura squamata , asserted that the rudiment of a hydrocoele was normally formed on each side of the embryo , but that only the left rudiment developed .
In 1895 , when engaged in the investigation of the development of the Asterid Asterina gibbosa , I encountered several specimens with two Echinoderm Larva with Two Water-Vascular Systems .
325 hydrocceles , and these were described in my paper on the development of that species ( 11 ) .
In 1910 my friend and colleague , Mr. Fuchs , whilst making cultures of the larvae of Echinus esculentus , discovered one with two hydrocceles .
This abnormal specimen he was good enough to hand over to me for examination , and in 1911 I published an account of it ( 14 ) , describing at the same time a similar specimen of the larva of Echinus miliaris , which I had found in my own cultures .
In 1911 a larva with two hydrocceles was observed by Caswell Grave in cultures of the larvae of the Clypeastrid Mellita testudinata , and a similar larva was found by von Ubisch in 1913 in cultures of the larvae of Strongylocentrotus lividus ( 18 ) .
Finally , in 1914 , Gemmill published a description of a number of larvae with two hydrocceles , which he discovered in his cultures of the larvae of the Asterid Asterias ruhens ( 6 ) .
In this paper Gemmill suggests that the doubling of the hydroccele was due to some factor in the environment , because practically all the larvae in one jar showed signs of two hydrocceles .
In 1915 he also recorded the occurrence of larvae with two hydrocceles in cultures of the larvae of Porania pulvillus ( 7 ) .
In this list of recorded instances of the occurrence of two hydrocoeles amongst Echinoderm larvae , it will be noticed that only one such specimen was found amongst larvae fished from the open sea , although hundreds of such larvae have been examined .
This unique larva was that figured by Johannes Muller .
All the others were found amongst larvae reared in confinement , and this circumstance led me to agree with Gemmill in ascribing the formation of the double hydroccele to some factor in the environment .
Gemmill put forward the tentative suggestion that overfeeding caused the formation of two hydrocceles amongst Asterid larvae , for he noticed that , in the jar in which the largest number of larvae with two hydrocceles was found , the diatom which he used as food for them had undergone great multiplication .
In the process of metamorphosis amongst Asterid larvae , the partitions between the various ccelomic cavities become to a large extent absorbed , and , in particular , the left posterior coelom sweeps to the right under the stomach , and opens into the right anterior coelom .
In overfed larvae , according to Gemmill , the swollen stomach prevents this process from taking place , and the right anterior coelom , freed from the disturbing contact with the left posterior coelom , develops in the same way as its left antimere , i.e. it buds off a hydroccele from its hinder end .
It is worthy of note that Gemmill had previously observed ( 5 ) that in about 10 per cent , of the'young larvae contained in a normal culture of the larvae of Asterias ruhens the right anterior coelom formed a pore-canal , so 326 Prof. E. W. MacBride .
Th Artificial Production of that there were two pore-canals and two madreporic pores , a light and a left , but there was no relation to be found between the occurrence of two pore-canals and two hydrocceles .
Most of the larvae with two pore-canals possessed only one hydroccele , and some of the larvae with two hydrocceles had only one pore-canal .
The cause assigned by G-emmill for the production of two hydrocceles , even if correct , could only be operative in the case of the larvae of Asteroidea , for at the stage of development at which two hydrocoeles have been observed in the larvae , both of Ophiuroidea and of Echinoidea , no such extension of the left posterior coelom normally takes place ; indeed , at no time in the normal development of the larvae belonging to these two orders does this cavity come into relation with , much less open into , the right anterior coelom .
As the larvae of the common green shore urchin Echinus miliaris had for several years been successfully reared through their entire development in the tank-room of the Zoological Department of the Imperial College of Science , it seemed to me that this species would form a very suitable subject on which to investigate the causes of the production of a double hydroccele .
These investigations were begun in the summer of 1914 , and carried to a successful conclusion in the summer of 1917 .
Material and Methods .
The specimens of Echinus miliaris used in these experiments were procured from Plymouth in 1914 , from the Essex coast in 1915 and 1916 , and in 1917 from Looe , in Cornwall .
All four lots gave concordant results , and so the conclusions at which I finally arrived have a considerable body of evidence behind them .
The best results were obtained from the Looe sea-urchins , and I have to thank Dr. Allen , the Director of the Plymouth Marine Biological Station , for his kindness in sending a special expedition to Looe Island in order to secure these specimens .
The urchins were sent to London by the swiftest route available , packed in damp seaweed .
A laboratory assistant awaited their arrival at the railway terminus , so that the delay which would have been occasioned by the slowness of delivery in London was entirely obviated .
They were opened and the ova and sperm were extracted from them , within an hour of their arrival at the laboratory .
The sea-water used in order to carry out the fertilisation of the eggs and the rearing of the larvae was sterilised by the method introduced by Dr. Allen in Plymouth .
This method is as follows : \#166 ; The sea-water is first shaken up with animal charcoal and then allowed to settle and the clear fluid is siphoned off .
This fluid is in this way freed from soluble toxins which might interfere with the growth of the larvae .
lm Echinoderm Larvce with Two Water-Vascular Systems .
327 is then passed through a Berkfeld filter in order to remove from it all bacteria .
The ripe eggs obtained from the females were fertilised in finger bowls , and the next morning , when the free-swimming blastulae had risen to the top these were decanted off into " Breffit jars " in which a considerable portion of , and in some cases all , their further development took place .
These " Breffit jars " \#151 ; the use of which was introduced by Dr. Allen and the workers in the Plymouth Marine Biological Laboratory\#151 ; deserve some words of comment .
They are made of thick green glass , and each contains , when full , about 2\ litres of sea-water .
The colour of the glass , which admits approximately only the same rays of light as can penetrate the surface-layers of the sea , seems to be of some importance in keeping the larvae healthy .
The jars were placed in a sink which was kept moist by a trickle of water from the tap .
The larva is capable of developing for four or five days without food , indeed the ectodermal oesophagus only joins the endodermal stomach on the fourth day ( fig. 1 ) .
By the fifth or sixth day the initial impetus of development has been exhausted , and although the larva can live for several days longer , degenerative processes set in , and sooner or later it falls to the bottom and dies .
The degeneration shows itself especially in the absorption of the ciliated arms .
It is therefore necessary to supply the larva with food , and the food chosen was a pure culture of the diatom Nitzschia .
These pure cultures are prepared in the Plymouth Biological Station , and directions for sub-culturing are given in Dr. Allen 's paper on the subject ( 1 ) .
The successful rearing of the larvae depends principally on the establishment of a healthy growth of Nitzschia in the culture jar ; but according to Dr. Allen 's experience , not more than about thirty larvae will grow to an advanced stage of development in a Breffit jar , and of these ordinarily only a few will complete their metamorphosis .
It is , however , possible to rear hundreds of larvae in such a jar until they are about a fortnight old .
But for a successful issue of a rearing experiment , it is necessary as the larvae grow older to thin out the culture at frequent intervals by transferring portions to fresh Breffit jars , and my experience coincides with that of Dr. Allen that one is lucky if one succeeds in rearing six or seven larvae in each jar through the metamorphosis .
In 1914 , 1915 , and 1916 I reared the larvae in Breffit jars , but I was dissatisfied with the numerical results obtained , and in 1917 I adopted the plan of transferring the larvae when they were about ten days old , into much larger receptacles , viz. , the plunger jars .
328 Prof. E. W. MacBride .
The Artificial Production of A plunger jar , such as is installed in the tank-room in the Zoological Department of the Imperial College of Science , is a large glass bell-jar which holds when inverted about 30 litres of water .
Each jar is provided with a glass cover , the upper side of which is covered by black paper , and in this way the water contained in the jar is shielded from the direct action of sunlight , as the tank-room is lit from above .
The plunger was invented by Mr. E. T. Browne , and is described by him in a paper on the rearing of Medusae ( 2 ) .
It keeps the water in constant slow movement .
' If one compares advanced larvae reared in plunger jars with those reared in Breffit jars , one is struck with the fact that the former have far longer arms than the latter , and in this respect are certainly more normal , for they resemble the larvae that are found in the open sea .
But the production of more normal larvae is by no means the only advantage which accrues from the use of plunger jars .
A much more important one is the far greater number of larvae which can be successfully reared to an advanced stage of development .
In one of these jars at the Imperial College , on a previous occasion , no less than 200 larvae completed their metamorphosis , and in 1917 , when the experiment was only carried on until the larvae were six weeks old , one plunger jar alone must have contained at least 1000 larvae .
If one contrasts this result with that which is obtained by rearing the larvae in Breffit jars , the advantage of the plunger jar becomes obvious .
Even if it were possible\#151 ; which it is not\#151 ; to rear to an advanced stage 50 larvae in each Breffit jar , one would require 20 such jars containing in all 50 litres of water in order to accommodate 1000 larvae ; but , as every experimenter knows , it is practically impossible to secure exactly the same conditions in any two Breffit jars ; in some the Nitzschia , for some unexplained reason , fails to flourish , and as a consequence few larvae survive .
For this cause , whilst I was able to obtain qualitative results in 1914 , 1915 , and 1916 , quantitative results were only obtained in 1917 .
The larvae of Echinus miliaris are beautifully transparent , and were in many cases drawn living .
Very serviceable whole mounts were obtained by preservation in carefully neutralised 40-per-cent , formalin , followed by absolute alcohol .
Alcoholic eosin or light green was used for staining .
In order to elucidate the relationship between the right hydrocoele and the pore-canal it was necessary to cut sections .
These were taken parallel to a plane cutting the larval arms at right angles .
Larvae of which it was desired to cut sections were not preserved in formalin , as this fluid does not give sufficiently good fixation of histological details .
They were preserved in Bouin 's fluid , which gives a very good preservation , and which makes the larvae far less brittle than osmic acid followed by Muller s fluid , which I Echinoderm Larvce with Two Water-Vascular Systems .
329 employed in previous work on Echinoderm larvae .
They were embedded in celloidin , and subsequently in paraffin , by the method which I described in detail when giving an account of the development of Echinus ( 12 ) .
The Normal Developmemt of Echinus miliaris .
In a former communication to this Society ( 12 ) , I have described in detail the development of Echinus esculentus , and Echinus miliaris , the larva of which was first described by me in 1898(11 ) , %develops in essentially the same way .
Nevertheless , it seems desirable to recapitulate briefly the principal points in my paper on the development of Echinus esculentus , in order that the reader may more easily grasp the results of the experiments to be described in this paper .
The result of the segmentation of the egg is to form a spherical blastula , one side of which then becomes flattened , and an invagination of the centre of this flattened surface forms the primordial gut or archenteron , and so the blastula is changed into a gastrula .
From the apex or blind end of the archenteron a bilobed vesicle is cut off ; this is the rudiment of the secondary body-cavity or coelom ; the wide space which intervenes between the archenteron and the ectoderm , and which contains a few connective tissue cells , is the blastocoele or primary body-cavity ( fig. 1 ) .
The larva now grows in length , and becomes obliquely flattened on its anterior surface ; it is as if it were bevelled .
This new flattened surface is the oral surface , and in the midst of it the mouth is formed as the opening of a wide shallow invagination , which is termed the stomodseum .
The archenteron having given off " the coelomic rudiment now becomes converted into the larval gut .
It was originally straight , but it now becomes arched , so that its apex is bent towards the bottom of the stomodseum , and it becomes divided by constrictions into three portions , viz. , oesophagus , stomach , and intestine .
The oral surface of the larva becomes surrounded by a thickened ridge of ectoderm , which carries very long and powerful cilia , and so the ciliated band , which is the organ of locomotion of the young larva , is formed .
Soon after , the bottom of the stomodseum breaks through into the oesophagus , and so the formation of the larval gut is completed .
The anus is , of course , the persistent blastopore or opening of the invagination which made the archenteron .
The ccelomic rudiment becomes divided into right and left ccelomic sacs at the same time , and the ciliated band grows out into four lobes .
These lobes are the first of the larval ciliated arms to be formed ; two are situated in front at the sides of the mouth and are termed antero-lateral , and two are placed behind the mouth in front 330 Prof. E. W. MacBride .
The Artificial Production of of the anus .
These latter are termed post-oral arms .
The two arms on each side of the larva are supported by delicate calcareous rods which are outgrowths of a calcareous star , the beginnings of which can be traced back to the blastula stage .
On the 5th day the left ccelomic sac sends out a tubular protrusion , which ascends to the dorsal surface , fuses with the ectoderm , and opens to the outside .
This outgrowth is the pore-canal , and its opening to the exterior is the primary madreporic pore .
The development up to the point which I have just described is exceedingly familar to all workers in experimental embryology , for Echinus is closely allied to the famous Echinus microtuberculatus of the Bay of Naples , if indeed it be not merely a northern race of that species .
As all know , countless experiments have been made on the larva of Echinus microtuberculatus , but few experimenters have carried their cultures of the larvae beyond the point of development which has just been described , which is just the point at which the larvae begin to require food .
As mentioned above , the appearance of the older larva was first described by me ( 12 ) and the only series of experiments in which the older larva has been used are those carried out by Debaisieux ( 3 ) in the laboratory of the Imperial College under my supervision and those made in the marine biological laboratory at Plymouth by Shearer , De Morgan and Fuchs ( 16 ) .
When the larva has attained the age of about eight days the left ecelomic sac becomes divided into anterior and posterior portions , and a little later the right coelomic sac begins to be constricted into anterior and posterior portions ( fig. 2 ) , but these parts are not finally divided from one another until the 12tli day .
On the 10th day the hydrocoele begins to be formed as a swelling of the hinder end of the left anterior ccelemic sac .
This bud-like swelling is never completely separated from the rest of the sac ; the neck of union between the two portions becomes drawn out so as to form a narrow tube , which is the stone canal ( st.c .
, fig. 4 ) .
It follows that the connexion of the hydrocoele with the exterior is an indirect one , for the left anterior axiom persists into the adult stage as a sac-like space termed the axial sinus , and the stone canal leads from the hydrocoele or water vascular ring into this sinus , and from this sinus the pore-canal ( which in the adult becomes multiplied into numerous pore-canals ) leads to the exterior .
The ciliated band grows out on each side into an extra arm termed the postero-dorsal arm , which is supported by a calcareous rod growing out from a new calcareous star on each side .
In the bay surrounded by the loop of the ciliated band leading from the base of this arm to the base of the postoral arm , an invagination of the ectoderm takes place which is termed the amniotic invagination { am .
, fig. 3 ) .
The ectoderm at the bottom of Echinoderm Larvcp , with Two Water-Vascular Systems .
331 this pit becomes thickened and pressed against the hydroccele bud , whilst the opening of the pit becomes narrowed and eventually closed : the pit is then termed the amniotic sac and its roof the amnion .
The complex structure consisting of the adpressed hydrocoele and tioor of the amniotic sac is known as the Echinus-rudiment , and the subsequent development of the larva consists mainly in the growth and elaboration of the Echinus-rudiment ( Teh .
, fig. 5 ) .
Before , however , we deal with the further development of this structure , we may glance at the changes which take place in other parts of the larva .
A fourth pair of arms which are termed the prse-oral arms grow out from that portion of the ciliated band which extends between the bases of the two antero-lateral arms . .
This pair of arms is extremely apt to be imperfectly developed in larvae which have been artificially reared ; most frequently one of the pair is longer than the other , but sometimes one and sometimes even both remain entirely undeveloped .
They are supported by calcareous rods which are outgrowths of a median Y-shaped ossicle the dorsal arch ( .
, fig. 3 ) , which lies above the oesophagus and which makes its appearance about the 18th day , before indeed any trace of the arms themselves has appeared .
The stem of the Y extends backwards and develops into a calcareous network which surrounds the madreporic pore and is the rudiment of the madreporic plate or madreporite of the adult .
From this plate about the 30th day a knob sprouts out which becomes one of the adult spines and is termed by me the madreporic spine ( m. sp. , fig. 22 ) .
About the 15th day the ciliated epaulettes make their appearance .
These rapidly develop into great crescentic lobes , horizontally placed , which together almost encircle the larva .
They bear numerous long cilia and in later larval life constitute the most important locomotor organ of the larva .
They were originally loops of the ciliated band , the dorsal ciliated epaulettes being formed from the loops between the antero-lateral and the postero-dorsal arms , whilst the ventral epaulettes are formed from that part of the ciliated band which stretches between the bases of the post-oral arms .
These loops become entirely separated from the ciliated band , but the breaches of continuity in the band so caused quickly heal up ( .
, fig. 5 ) .
So far as is yet known , ciliated epaulettes are confined to comparatively few genera and are not by any means characteristic of all Echinoid larvae .
We have mentioned above that the separation of the right anterior coelom from the right posterior coelom is not completed before the 12tli day .
After it is complete the right anterior coelom sends out from its posterior end a tongue of cells which arches over the oesophagus towards the mid-dorsal line .
From the apex of this tongue a cellular bud is detached , which becomes hollow and 332 Prof. E. W. MacBride .
The Artificial Production of constitutes the madreporic vesicle , so termed because it lies just under the madreporite close to the madreporic pore .
In my paper on the development of Echinus esculentus ( 12 ) I put forward the view that the vesicle represented a vestigial right hydroccele .
At that time I had never seen an Echinus larva showing a genuine right hydrocoele .
It will be made plain in this paper that that interpretation was incorrect , for the madreporic vesicle may be well developed in larvae which show an unmistakeable right hydroccele ( figs. 18 , 21 ) .
On the right side of the larva , about opposite the spot where the amniotic invagination is formed on the left side , just above and below the loop of the ciliated band which extends between postero-dorsal and post-oral arms , two knobs begin to appear about the 20th day .
About the 25th day it becomes obvious that each is developing into a pedicellaria .
Each is based on a calcareous plate which becomes one of the five genital plates of the young sea-urchin .
As the larva grows older an additional pedicellaria and several square-topped spines appear on each plate .
These plates are just as characteristic of the right side of the larva as the Echinus-rudiment is of the left side ( fig. 6 ) ; midway between them another calcareous plate is developed { calc .
, fig. 6 ) .
We may now consider the development of the Echinus-rudiment .
This structure , originally small , grows until it occupies the whole side of the larva .
The hydroccele gives rise to five lobes which protrude into the amniotic space , each covered , of course , with a layer of ectoderm derived from the floor of the amniotic sac .
These lobes are the five primary tentacles which constitute the radial water-vascular canals of the adult .
It is characteristic of Echinus miliaris that each primary tentacle soon gives rise to two lateral buds which constitute the first paired tentacles , whereas in Echinus esculentus this event does not take place until after metamorphosis .
Between the bases of each two adjacent primary tentacles there are developed as outgrowths of the floor of the amniotic sac four pointed spines , each of which soon develops at its base the neuro-muscular collar characteristic of the adult spine .
This area of the amniotic floor , between two tentacles , which gives rise to these spines may be termed the interradius .
Erom the outer wall of the left posterior coelom lying beneath each interradius , a pocket-like evagination is formed .
Each pocket becomes one of the dental sacs { d.s. , fig. 19 ) of the adult and in its walls calcareous ossicles are formed which give rise to a tooth and the pair of jaws embracing it .
The five dental sacs and the ossicles to which they give rise constitute the complex structure known as Aristotle 's lantern in the adult .
After metamorphosis an outgrowth from the / Echinoderm Larvce with Two Water-Vascular Systems .
333 larval stomach meets a central depression of the amniotic floor and in this way the adult mouth and oesophagus are formed .
It will be obvious , therefore , that in the formation of the Echinus-rudiment no less than three distinct tissues are involved .
These are ( 1 ) the hydrocoele , ( 2 ) the amniotic floor , ana ( 3 ) the wall of the left posterior coelom .
One of the main objects of the experiments described in this paper was to determine , 334 Prof. E. W. MacBride .
The Artificial Production of if possible , how far these tissues were independent of one another in their modifications , or whether the modification of one caused the modification of the rest .
Description of the Experiments .
A. The Production of Larvae with Two Hydrocoeles .
On thinking over the possible changes in the environment which might be likely to be encountered by larvae reared in artificial cultures it occurred to me that the salinity of the water in the culture-jar might be raised by accidental evaporation and that this stimulus might be the factor which caused the formation of a second hydrocoele .
I may say in anticipation that this surmise proved to be partly right and partly wrong .
It is true that increased salinity , acting at a certain critical period of development , will lead to the production of a right hydroccele , but other causes must also be competent to produce the same result , for increased salinity will not account for every instance of this phenomenon .
It transpired , from preliminary experiments in 1914 , that larvae would grow fairly well in water of which the salinity had been raised by evaporation to 37 parts per thousand , and with difficulty in water of which the salinity was 39 parts per thousand .
Accordingly , for all further experiments , water of a salinity of 37 parts per thousand was employed , although in 1914 some trials were made of transferring larvae which had been reared up to a certain stage of development in water of a salinity of 37 parts per thousand to water of a salinity of 39 parts per thousand .
Water of enhanced salinity is usually denoted as " hypertonic , " and , for the sake of brevity , I shall adopt this term in further discussion of the subject .
I was unsuccessful in fertilising the eggs in hypertonic water , although I do not assert dogmatically that this difficulty could not be overcome .
However , I abandoned further attempts to do this , and all the eggs used in these experiments were fertilised in normal sea-water , and the larvae were transferred to hypertonic water when they had reached an age of three days ( 1915 , 1916 ) or four days ( 1917 ) .
The exact stage of development attained at these ages varies with the time of year and with the vigour of the culture , but , broadly speaking , the stage reached at four days may be taken to be characterised by the completion of the alimentary canal , by the division of the ccelomic rudiment into two equal coelomic sacs , and by the appearance of the antero-lateral and post-oral ciliated arms , but by the absence of a madreporic pore .
In a word , the stage may be described as one in which no departure from strict bilateral symmetry has as yet occurred ( fig. l)t .
In 1914 , when the larvae were about a fortnight old , and had lived for Echinoderm Lawce with Two Water-Vascular Systems .
335 10 or 11 days in the hypertonic water , the hydroccele was formed on the left side as in normal cultures , but in some there appeared on the right side an outgrowth from the right anterior coelom in the form of a little vesicle , which was not present in normal larvae ( .
, fig. 3 ) .
I hailed this structure as a rudiment of a right hydroccele , but no amniotic invagination was formed over it , and , in spite of all my endeavours , the larvae refused to develop further , and so this experiment led to an indecisive result .
In 1915 the experiment was repeated , and the same result was obtained .
Some of the most promising larvae were picked out and placed each in normal sea-water in a separate finger-bowl , with a rich supply of In one of these larvae the bud on the right side developed five lobes , showing that it really was a right hydroccele ; the other specimens died .
In 1916 the experiments were again repeated , and , during that year , some larvae were kept living in Breffit jars filled with the hypertonic water , whilst others , after having been for periods varying from a week to three weeks in this water , were transferred to Breffit jars filled with normal sea-water .
At the end of six weeks I had the immeasurable satisfaction of seeing some larvae belonging to both categories develop an unmistakable right hydrocoele , provided with five tentacles , which projected into a well-formed amniotic cavity ( fig. 7 ) .
In these experiments , the tendency of the hypertonic water to inhibit the growth of the Nitzschia was overcome by the constant addition of fresh doses of Nitzchia .
I was still , however , far from satisfied with the results which I had obtained .
The larvae grown in the Breffit jars were in some respects unhealthy , as fig. 7 shows , and the number which survived until reaching the stage when the hydrocoele develops lobes was small , and so I made arrangements to repeat the experiment on a large scale during the summer of 1917 .
Owing to difficulties connected with the war , I was unable to secure a supply of sea-urchins until July 7 .
The experiment was closed on August 14 , and so it lasted only between five and six weeks\#151 ; an insufficient amount of time to allow of the larvae completing their metamorphoses , but long , enough to allow of them producing large Echinus-rudiments with all their characteristic features .
As I have already mentioned , the urchins sent me in this year ( 1917 ) were collected at Looe Island , and I have reason to believe that , owing to the experience of the assistants of the Plymouth Laboratory , who packed and forwarded them , they reached me in better condition than the urchins which I received in 1915 and 1916 .
I decided to increase the salinity of the water , not by evaporation , but by the addition of 5 grm. of common salt to each Breffit jar , which raised the 336 Prof. E. W. MacBride .
The Artificial Production of amount of all the salts present in the water to 37 parts per thousand .
When the method of evaporation is used , the concentration of all the salts present in sea-water is raised in equal proportions , and , in addition , the acidity of water is reduced by the expulsion of CO2 .
In the method employed in 1917 , the acidity of the water was undisturbed , and the concentration of NaCl alone was raised .
The eggs were fertilised on July 7 , and on July 11 some of the larvae were transferred to Breffit jars with hypertonic water ; other larvae were retained in normal sea-water to serve as controls .
On July 17 , when the larvae had lived for six days in the hypertonic water , a large number of them were transferred to plunger jars .
At this time they were showing signs of the division of the ccelomic sacs into anterior and posterior portions ( fig. 2 ) .
I had altogether eight plunger jars at my disposal\#151 ; four being situated on each side of the tank-room .
To show how slight are the variations in conditions of light , etc. , which make for success or failure , I may mention that only the cultures on one side of the tank-room were successful , and we shall therefore confine our attention to this group of four jars .
Of these , one was filled with hypertonic water and the Nitzschia would not flourish in it , and as a consequence all the larvae perished .
The remaining three jars were filled with normal sea-water , and each received the contents of a Breffit jar which must have contained over a thousand larvae ; two of the plunger jars received the larvae from " salted " Breffit jars , and one those from a Breffit jar which had contained normal sea-water .
The culture in this jar , therefore , served as a control with which to compare the results of the cultures in the other two .
It is necessary to add that all the larvae which were added to the plunger jars were the offspring of the same cross .
This method of starting a culture of Echinoderm larvae in a plunger jar differs from the method which I adopted in previous attempts to rear the larvae of Echinus esculentus and Echinus miliaris .
My former method was to pick out about 300 of the most healthy looking larvae from a Breffit jar and transfer them to a plunger jar .
This method sometimes succeeded , but often failed\#151 ; in fact one may say that it always failed unless the larvae were taken from the Breffit jar when they had attained a relatively advanced state of development , when , in fact , all their weaker brethren had perished .
The reason for this seems to be that the eye of the experimenter is unable to select the really healthy larvae , and that the larvae must therefore be subjected to a period of struggle\#151 ; in a word of natural selection\#151 ; in order that those individuals may be found which are sufficiently vigorous to be able to survive until they complete their metamorphosis .
Echinoderm Larva with Two Water- Vascular Systems .
337 We may now describe the progress of the cultures in the two plunger jars which contained larvae which had been subjected to the action of hypertonic sea-water .
When the larvae had attained the age of about 18 days* many of them showed unmistakeable signs of the formation of the rudiment of a right hydrocoele , such as I had observed in the larvae in my cultures of 1914 and 1915 .
But then the development of this rudiment stopped , ancf it looked as if I were about to obtain a result as indecisive as that which I obtained in 1914 and 1915 .
This period was characterised by dull rainy weather which did not favour the growth of the Nitzschia in the jars .
However , about a week later , sunshine returned , and by means of a persistent addition of Nitzschia culture to the jars the growth of the diatom was maintained in a healthy condition , and at last on August 9 larvae showing an indubitable right hydrocoele were found in the jars .
On August 10 , 13 and 14 a thorough examination of the contents of the jars was made .
The larvae were taken out in groups of 50 , and each group was inspected under the simple microscope .
In the beautifully transparent larvae any trace of a right hydrocoele was at once clearly visible .
In one of the two jars about 2 per cent , of the larvae possessed a right hydrocoele ; in the , at least 5 per cent , showed this structure .
In the plunger jar which served as control only one solitary specimen provided with a right hydrocoele was discovered amongst hundreds which were examined .
At the same time some of the larvae which had remained in the " salted " Breffit jars developed a right hydrocoele .
These larvae were the offspring of different crosses from the cross the offspring of which was used to form the cultures in the plunger jars .
When one puts together the results of the experiments made during the four years , no reasonable doubt can be left in the mind of anyone that exposure to the action of hypertonic water at an early stage of development really does bring about the formation of a right hydrocoele .
It is probable that by a modification of the methods which I employed , the proportion of larvae with two hydrocoeles could be raised to a much higher figure ; but the difficulty of the experiment lies in this , that to produce a right hydrocoele , or at any rate to produce a fully developed one , two factors are necessary , viz. : ( 1 ) the exposure to hypertonic water , and ( 2 ) the provision of a superabundance of food .
Since the growth of a second hydrocoele carries with it the development of a second set of spines and teeth it is an " expensive " structure and makes large demands on the food supply in the body of the larva .
338 Prof. E. W. MacBride .
The Artificial Production of Since , moreover , the growth of Nitzschia varies immensely with conditions , too much or too little light being equally deleterious to it , it is very difficult to ensure the co-operation of this factor in the right strength at the right time .
The appearance of a single larva with two hydrocoeles in the control plunger jar serves to warn us that other stimuli besides increased salinity may give rise to this phenomenon , and the circumstance that embryos with two fully developed hydrocoeles were observed by Metschnikoff ( 14 ) among the specimens which he extracted from the womb of squamata tends to emphasise this point .
Indeed , when we reflect that most cases recorded of larvae with two hydrocoeles have been represented by the occurrence of a single specimen amongst hundreds of normal larvae , we may feel certain that their development has not been caused by increased salinity .
We may now turn our attention to the structure of the abnormal larvae which were obtained .
In a minority both hydrocoeles were equally developed ( fig. 8 ) , and in these cases the characteristic pedicellariae of the right side were absent .
The right hydrocoele had induced the formation of a right amniotic cavity , into which its outgrowths projected as a second series of primary tentacles , and from the floor of this cavity pointed spines were produced similar to those belonging to the Echinus-rudiment on the left side .
When such larvae were explored by sections , it was found that on the right side the rudiments of a second Aristotle 's lantern were represented by dental pockets just like those on the left side ( .
, fig. 19 ) , and that the right hydroccele was connected with the right anterior coelom by a well formed right stone-canal .
The right anterior ccelom had been enlarged so as to form a right axial sinus which had become fused with the left axial sinus , and thus there arose 'a large median cavity lying above the oesophagus ; and the madreporic vesicle , which , it will be remembered , normally lies above the oesophagus , was sometimes entirely suppressed ( fig. 20 ) , but sometimes persisted as a narrow cavity intervening between the compound axial sinus and the gut ( figs. 18 , 21 ) .
When describing the development of Asterina I stated that the madreporic vesicle was a rudiment of the right hydrocoele , because in the abnormal specimens which possessed a fully developed right hydrocoele I was unable to detect a madreporic vesicle .
No doubt in these specimens the right hydrocoele had completely suppressed the madreporic vesicle as iu the larva represented in fig. 20 .
The compound axial sinus may communicate with the exterior by one ( fig. 20 ) or two pore-canals ( fig. 21 ) .
But in those larvae in which I found two pore-canals both of them belong to the left side , and the second poreEchinoderm Larvae with Two Water-Vase alar Systems .
339 canal is not therefore to be regarded as an attempt on the part of the right side of the larva to form a structure similar to one normally found on the left side , but rather to a precocious beginning of the process of increasing the number of pore-canals such as occurs in all larvae at or after the time of metamorphosis .
On the other hand I did discover one larva with a right as well as a left pore-canal ( fig. 17 ) .
But this larva had been classified by me as a " normal " larva , for in all other respects , as , for instance , in its single hydroccele and its well developed pedicellariae , it was entirely normal .
The two pore-canals fused with one another in the mid-dorsal line and opened by a median dorsal pore .
My results therefore confirm the view of Gemmill that the doubling of the pore-canal and the doubling of the hydroccele are phenomena which are entirely independent of one another .
In the majority of the abnormal larvae the right hydroccele is smaller and less advanced in development than the left one .
In such cases one or even both of the pedicellariae proper to the right side may be present ( fig. 10 ) .
The right hydroccele has thus every appearance of being an after-thought , induced by the intervention of an external factor .
When it is small it may be entirely devoid of lobes and of an amniotic cavity ( fig. 11 ) , and even when an amniotic cavity is developed , a disharmony between it and the hydroccele is frequently apparent , for one or more of the lobes of the latter , instead of projecting into the amniotic cavity , may project into the blastoccele ( ab.t .
, fig. 10 ) .
In some few cases more than one rudiment of a hydroccele may be formed on the right side ; these cases are due to the splitting of the bud which is formed at the hinder end of the right ( or left ) anterior coelom .
One such specimen was sectioned by me : one of the two right hydrocceles was normally developed , and had associated with it an amniotic cavity ; the other was a small structure situated posterior to it , and devoid of an amniotic cavity , but possessing well developed lobes .
Another specimen was discovered in which the stone-canal alone was doubled on the left side , a phenomenon obviously to be ascribed to the splitting of the string connecting the hydroccele bud with the anterior coelom ( .
, st."cf fig. 19 ) .
B. The Production of Larvae Devoid of a Hydroccele .
The converse of a larva developing a hydroccele on both sides would be a larva which developed pedicellariae on both sides .
Such larvae are very , rare ; their significance is discussed later .
I have , however , discovered a means of producing larvae devoid both of hydroccele and of pedicellariae , but which-possess on both sides in the position usually occupied by the hydroccele a group of spines , and also on both sides a spine or spines in the position vol. xc.\#151 ; b. .
2 E 340 Prof. E. W. MacBride .
The Artificial Production of usually occupied by one of the pedicellariae .
Such larvae are represented in figs. 12 , 13 and 14 , and sections of one are shown in figs. 22 and 23 .
In 1915 I possessed a very thick culture of larvae in a Breffit jar ; the larvae were between six and seven days old , and had had as yet no food , and the well-known signs of degeneration were beginning to appear in many of them .
The whole contents of the Breffit jar were then added to a plunger jar in which there was a rich culture of Nitzschia .
A few weeks later the plunger jar was seen to be filled with a vigorous culture of advanced larvae , which , to the naked eye , seemed to be completely normal .
All had the full complement of eight arms and four ciliated epaulettes .
When these larvae were examined with a lens about one-third of them appeared to be laden with a mass of calcareous growth on both sides .
For a brief moment I thought that my hopes of producing a large number of larvae with a double Echinus-rudiment had been fulfilled ; but when these curious larvae were sectioned no trace of hydroccele could be discovered on either side of the stomach ; nor was there any vestige of a madreporic pore or of a stone-canal .
The coelom had been divided into anterior and posterior portions on each side , but on neither side was the anterior division enlarged so as to form an axial sinus .
Dorsal to the oesophagus a median sac , no doubt homologous with the madreporic vesicle , was discovered fig. 22 ) .
Within the loop of the ciliated band ( til .
, fig. 23 ) , where normally the amniotic invagination is formed on the left side , there was situated on both sides a group of pointed spines ( r.sp .
, l.sp .
, fig. 23 ) , and dorsal to this loop , in the position usually occupied by the dorsal pedicellaria on the right side , there is situated on both a pointed spine ( r'.sp ' .
, l'.sp ' .
, fig. 23 ) .
These spines differ from those present in the normal larvae in two respects .
( 1 ) If we compare them with the spines normally developed on the right side by the plates which bear the pedicellariae , then we may notice that whereas the spines in normal larvae have square ends with diverging points ( resembling somewhat in shape old church-towers ) , the spines in these abnormal larvae are pointed .
( 2 ) If we compare them with the spines which in the normal larva are borne by the Echinus-rudiment alternating with the primary tentacles , then it is to be noticed that the spines of the normal larva are pointed , but are encircled at their bases by a collar consisting of nervous tissue above and muscular tissue beneath , whereas the spines of these abnormal larvae are totally devoid of such a collar .
In 1916 and 1917 the experiment which I have just described was repeated and the same result obtained .
Amongst all the larvae which were sectioned one was discovered which possessed a madreporic pore .
Echinoderm Larcce ivith Two Water-Vascular Systems .
341 Discussion of the Eesults of the Experiments .
We may say , then , that , as the result of experiments carried on for four years , two points have been determined , viz. : ( 1 ) that if the lame of Echinus miliaris be exposed to the action of hypertonic sea-water for a week commencing with the fourth day of development , many of them will develop on the right side as well as on the left a hydrocoele or water-vascular rudiment , and that in connection with this second hydrocoele all the structures which normally develop in connection with the left hydrocoele may make their appearance , i.e. , spines , tentacles and dental sacs ; ( 2 ) that if the larvie of Echinus miliaris be starved during the first week of their existence , and then placed under favourable conditions both as to food and space , they will continue their development , but many of them will be devoid of both pedicellarise and hydrocoele , but will have in place of both a group of pointed spines on each side .
Such larvae will , in the majority of cases , be devoid of madreporic pore and axial sinus , but will possess a well-developed madreporie .
vesicle .
If we confine ourselves for the moment to the first of these results , , we cannot help recalling that the result of using hypertonic water on the unfertilised eggs of other Echinoidea is to cause some of them to develop into larvae without the aid of spermatozoa .
Although Loeb , in his improved method , employed butyric acid to stimulate unfertilised eggs to development , and only used hypertonic water as a subsequent treatment , in order , as he supposed , to correct the cytolytic tendency which had been produced by the acid , yet , in his earlier experiments ( 10 ) , he found that hypertonic water alone was able to produce parthenogenetic development .
If , however , we make this comparison , another thought is inevitably suggested .
Just as the process of development of an unfertilised egg is not something foreign to its nature , which has been forced on it by the hypertonic water , but consists merely in rendering active a latent potentiality , so we cannot assume that the formation of a right hydroccele is a new accomplishment forced on the larva ; it must be merely the rendering actual of a potentiality which exists in the normal larva , but which usually remains in abeyance .
That this is so is supported by the occurrence of a single specimen showing the double hydrocoele in a culture reared in normal sea-water which contained hundreds of larvrn .
The most obvious explanation of this latent potentiality is to assume that the Echinoderms are descended from some type of animal which had a hydrocoele-like structure on each side .
Such a hypothetical ancestor is 342 Prof. E. W. MacBride .
The Artificial Production of very nearly realised in the existing genus which belongs to the group of the Hemichorda , and is thus related to the vertebrate phylum .
Cephalodiscus possesses two groups of long ciliated tentacles , and each group spring from a semicircular canal situated at one side of the mouth .
Gilchrist ( 8 ) has shown that the living Cephalodiscus crawls about on its prseoral lobe , and occasionally fixes itself to the substratum by the surface of this lobe , exactly as I have shown that the larva of Asterina gibbosa does ( 11 ) .
The theory that Asteroidea , and inferentially all other Echinodermata , are descended from an ancestor like Cephalodiscus , was fully worked out by me in my paper on the development of Asterina gibbosa , and the results of the experiments recorded in this paper confirm my confidence in the soundness of that theory .
But , when I described the first two larvae of Echinus with two hydrocceles which came into my hands ( 14 ) , I pointed out that the probable descent of Echinodermata from a Cephalodiscus-like ancestor affords at best only a partial explanation of the phenomena with which we are concerned For it is to be noted that spines provided with a nervous and muscular eollar are characteristic of the highly developed Echinoidea , and are not found in the more primitive Asteroidea , and the same is true of the highly developed teeth and jaws which constitute Aristotle 's lantern .
It is certain that the bilaterally symmetrical ancestor possessed none of these structures ; but that they were evolved long after the preponderance of growth of the left hydroccele over the right one had led to the conversion of .
the bilateral symmetry into a radial symmetry .
The theory that there was a bilateral ancestor of Echinodermata therefore cannot account for the existence of a double set of highly developed spines and dental sacs in the abnormal larva .
In the paper on these two abnormal larvae , I suggested that the conversion of a portion of the ectoderm of the left side into highly developed spines , and the development of dental pockets from the wall of the left posterior coelom , were due not to the inborn capacities of these tissues , but to the acquisition of powers conferred on them by some chemical substance , in a word , some hormone , which emanated from the hydroccele bud .
I said , further , that where a similar bud was formed on the right side , it emitted a similar substance , which forced the right ectoderm and the wall of the right posterior coelom to undergo corresponding changes , although they had never , in the history of the _race , been modified in this manner .
I assumed that the hydrocoele bud was the prime mover in the wonderful array of changes which led to the formation of an Echinus-rudiment , because in normal development it is the first part of this rudiment to appear , but this assumption is proved to be justified by the results of the experiments EchinodermLarva ) with Tivo Water-Vascular Systems .
343 recorded in the present paper .
For we have learned that when the right hydrocoele is small it may develop lobes , but may have no power to cause the formation of an amniotic invagination ( fig. 11 ) , and therefore no spines are formed , and the wall of the coelom is also unaffected and no dental pockets are developed .
We can , therefore , have a right hydrocoele without a right amniotic invagination and right dental sacs , but we cannot have these latter structures without a right hydroccele .
I possess one larva which has a right hydrocoele developing long lobes and a vestigial amniotic invagination which is entirely devoid of relation to these lobes .
The fact that a right hydrocoele can co-exist with one or more pedicellariae raises some interesting questions .
These pedicellariae are situated externally to the loop of the ciliated band within which the amniotic invagination is formed , so that when they are not formed , this non-appearance cannot be due to the right hydrocoele taking up the space which is normally occupied by them .
It seems far more probable that just as a hormone emanating from the hydrocoele causes the formation of a right amniotic cavity , so another hormone derived from the same source tends to inhibit the formation of pedicellariae .
When pedicellariae are nevertheless developed on the same side of the larva , this seems to be due to the arrested growth of the hydrocoele bud , so that when it really does begin to grow , the normal development of the right side has proceeded to such lengths as to be irrevocable , every larva ivhich has pedicellarice on both sides a hydrocoele is also present on both sides although one of them is usually vestigial .
In 1917 , owing to a temporary arrest of the growth of the in the plunger jars , the development of the right hydrocoele stood still for about 10 days and in the meantime the ominous knobs which are the first indications of the formation of the pedicellariae made their appearance on the right side of the larvae , and I feared that all chance of the formation of a well-developed right hydrocoele was lost .
But to my intense delight this fear proved groundless , for when a more abundant food supply was available the right hydrocoele resumed its growth in spite of the fact that by that time one or even two pedicellariae had been formed .
But the occurrence of larvae in which there were present a hydrocoele and two pedicellariae on each side enables us to go further in our analysis .
Such a larva can be explained only if we assume that the influences emanating from a hydrocoele not only tend to inhibit the formation of pedicellariae on the same side but to determine their formation on the opposite side of the larva .
If we assume that in these larvae the growth of both hydroeceles had been arrested at an early stage , but after the stage at which the stimulus to form pedicellariae on the opposite side 344 Prof. E. W. MacBride .
Th Artificial Production of had already gone forth from them , and that then , after the formation of these organs on both sides had been determined , further nourishment became available and the left hydrocoele developed further , the structure of such larvae can be explained .
The second main result obtained by my experiments concerns the further development of larvae which have been starved during the first week of their existence .
Such larvae have no hydroccele\#151 ; it is probable that that part of the coelomic rudiment which should have developed into the hydroccele becomes absorbed during the period of hunger .
In the further history of these starved larvae we see the potentialities of development of the Echinus larva when the hydrocoele bud has been removed .
It is , therefore , most instructive to note that the power of forming pedicellariae has likewise disappeared , and this result confirms the conclusion at which we had already arrived , viz. , that the power which determines the formation of pedicellariae emanates from the hydrocoele .
How are the extraordinary potentialities of the right side to be explained , potentialities , it may be remarked , which are never called into activity during the normal life of the species , and so far as we can judge have never been utilized in the past history of the race ?
We seem shut up to some such hypothesis as this .
When the Cephalodiscus-like ancestor still possessed two hydrocoeles , and whilst it was developing into a Proto-echinoderm with one hydrocoele its dominant activities were comprised in the writhing grasping movements of the tentacles which were the outgrowths of these hydrocoeles .
These activities caused modifications to take place in the neighbouring tissues ; and if one believes , as I do , in the inheritability of acquired characters , one may assume that the tendency to produce these modifications was inherited , so that the modifications were eventually evoked by the mere presence of the hydrocoele bud before the activities of its lobes had begun .
We must further assume that the hormones which produced the modifications , tended to be stored in the hydrocoele bud itself-\#151 ; and that both hydrocoeles were ultimately derived from the same small area of the coelomic wall .
All the hormones therefore necessary to produce the modifications which were acquired after the Cephalodiscus-ancestor by the loss of one hydrocoele had become an Echinoderin , were still situated in the same place ; and so , when this area under the influence of stimulation produced two hydrocoeles instead of one , these hormones were shed abroad on both sides of the larva , although they had only been evolved in connection with the left side .
This hypothesis may help us a little way , but no one is more conscious than I am how incomplete and unsatisfactory it is .
We ask in vain what sort of chemical composition we must ascribe to the right ectoderm and to Echinoderm Larvae with Two Water-Vascular Systems .
345 the right ccelomic wall , so that they are able under the influence of a hydrocoele bud to give rise to structures to which in the history of the race they have never given rise .
It is well to remind ourselves that this amazing phenomenon does not stand alone .
The facts that when the eye-stalk of a shrimp , including the underlying optical ganglion , is cut out , the stump regenerates not an eye but an antenna , and that further when the primary optic vesicle of a young tadpole is cut off from the brain and pushed backwards under the skin into the region of the shoulder it will force the local ectoderm to form for it a lens , warn us that living developing tissues do not obey the laws with which we become familiar when we study dead matter .
In many cases , at any rate , the body of an embryo is not , like a picture puzzle , a mosaic of pieces each destined to form a particular organ , but consists of sheets of indifferent material " without form and void " on which a formative " something " works and evokes the beautiful detail of the adult structure .
As Driesch(3 ) has expressed it , " Ein jedes jedes kann .
" BIBLIOGRAPHY .
( 1 ) Allen , E. J. , " On the Artificial Culture of Marine Plankton Organisms , " 'Journ .
Marine Biol. Ass .
, ' vol. 8 ( 1910 ) .
( 2 ) Browne , E. T. , " On Keeping Medusae Alive in an Aquarium , " Ibid. , vol. 5 ( 1898 ) .
( 3 ) Debaisieux , P. , " The Experimental Hybridisation of Echinus miliar is , Echinus esculentus , and Echinus acutus , " 'Quart .
Journ. Mic .
Sci. , ' vol. 58 ( 1912 ) .
( 4 ) Driesch , H. , ' Analytische Theory der organischen Entwicklung , ' Leipzig , 1894 .
( 5 ) Gemmill , J. F , , " The Development and Certain Points in the Adult Structure of the Starfish Asterias rubens " ' Phil. Trans. , ' B , vol. 205 ( 1914 ) .
( 6 ) Gemmill , J. F. , " Double Hydrocoele in the Development and Metamorphosis of Asterias rubens , " 'Quart .
Journ. Mic .
Sci. , ' vol. 61 ( 1915 ) .
( 7 ) Gemmill , J. F. , " The Larva of the Starfish Porania ' Quart .
Journ. Mic .
Sci. , ' vol. 61 ( 1915 ) .
( 8 ) Gilchrist , J. O. F. , " Observations on the Cape Cephalodiscus and some of its Early Stages , " ' Ann. and Mag. Not .
Hist. , ' Series 8 , vol. 16 ( 1915 ) .
( 9 ) Grave , Caswell , " Notes on the Metamerism of the Echinoid Pluteus , " 'Johns Hopkins University Circular , ' 1911 .
( 10 ) Loeb , J. , ' On the Artificial Production of Normal Larvae from the Unfei'tilised Eggs of the Sea-Urchin , " ' Amer .
Journ. Physiol. , ' vol. 3 ( 1900 ) .
( 11 ) MacBride , E. W. , " The Development of Asterina gibbosa , " ' Quart .
Journ. Mic .
Sci. , ' vol. 38 ( 1896 ) .
( 12 ) MacBride , E. W. , " The Development of Echinoids .
Part I.\#151 ; The Larvae of Echinus miliaris and Echinus esculentus , " 'Quart .
Journ. Mic .
Sci. , ' vol. 42 ( 1899 ) .
( 13 ) MacBride , E. W. , " The Development of Echinus , together with Some Points in the Development of E. miliaris and E. , " 'Phil .
Trans. , ' B , vol. 195 ( 1903 ) .
( 14 ) MacBride , E. W. , " Two Abnormal Plutei of Echinus and the Light which they throw on the Factors in the Normal Development of , " 'Quart .
Journ. Mic .
Sci. , ' vol. 57 ( 1911 ) .
346 Prof. E. W. MacBride .
The Artificial Production of ( 15 ) Metschnikoff , E. , " Studien iiber die Entwicklung der Echinodermen und Nemertinen , " * Mem. de l'Academie Imp .
de St. Petersbourg , ' vol. 14 ( 1869^ .
( 16 ) Muller , Joh. , " Ueber die Larvenzustande und die Metamorphose der Ophiuren und Seeigel , " ' Abh .
Konig .
Akad .
Wise .
, Berlin , ' 1846 .
( 17 ) Shearer , de Morgan and Fuchs , " On the Experimental Hybridisation of Echinoids , " ' Phil. Trans. , ' B , vol. 204 ( 1914 ) .
( 18 ) Yon Ubisch , L. , " Die Entwicklung von tuberculatm ' , Arbacia pustulosa ) , " ' Zeit .
Wiss .
Zool. , ' vol. 106 ( 1913 ) .
EXPLANATION OF PLATES .
List of Abbreviations .
ab.t .
, aberrant tentacle , i.e. , lobe of the right hydrocele which projects into the blastocoele and not into the amniotic cavity ; antero-lateral ciliated arm ; amniotic cavity ; an .
, anus ; ax .
, axial sinus ; ax ' .
, axial sinus developed on the right side from the right anterior coelom ; calc .
, calcareous plate ; oil .
, portions of the ciliated band of the larva forming the loop within which the amniotic invagination is formed both on the right and on the left side ; cce .
, coelom ; d.a. , dorsal arch , i.e. , the calcareous rod lying above the oesophagus in the larva ; dorsal ciliated epaulette ; d.s. , dental sac associated with the left hydrocoele ; d.s ' .
, dental sac associated with the right hydroccele ; Each .
, Echinus-rudiment ; int .
, intestine ; l.a.c. , left anterior coelom ; l.hy .
, left hydrocoele ; l.p.c. , left posterior coelom ; l.sp .
, left group of spines within the ciliated loop ; V.sp ' .
, left spine outside the ciliated loop ; mad .
, madreporite ; m.p. , madreporic pore ; m'.p .
, additional madreporic pore ; .
, madreporic spine ; oes .
, oesophagus ; p.c. , pore canal ; p'.c ' .
, additional pore canal ; p.d. , postero-dorsal ciliated arm ; ped .
, pedicellaria ; p.o. , post-oral ciliated arm ; pr.o .
, prse-oral ciliated arm ; r.a.c. , right anterior coelom ; r.hy .
, right hydrocoele ; r.p.c. , right posterior coelom ; r.sp .
, right group of spines within the ciliated loop ; r'.sp ' .
, right spine outside the ciliated loop : sept .
, remains of the septum which formerly divided right and left axial sinuses from one another ; st.c .
, stone-canal ; st.'c ' .
, additional stone-canal on the right side ; st".c " .
, additional stone-canal on the left side ; stom .
, stomach ; v.c.ep .
, ventral ciliated epaulette .
N.B.\#151 ; Skeletal structures are indicated in the drawings of whole larvae by heavy black lines .
Plate 4 .
Fig. 1.\#151 ; Young larva between three and four days old viewed from the ventral aspect .
Preserved in osmic acid , followed by Muller 's fluid , which has dissolved the skeleton .
Magnification 330 diameters .
Note the stomodaeum just about to open into the oesophagus ( oes .
) .
Only the post-oral ciliated arms are developed ( p.o. ) .
This larva is approximately at the stage of development at which larvae were transferred to hypertonic water in 1914 , 1915 and 1916 , and a little younger than that at which they were transferred to this water in 1917 .
lig .
2.\#151 ; Larva 12 days old viewed from the dorsal aspect , showing on the left side the formation of the hydroccele and on the right side the almost completed separation of the coelomic sac into anterior and posterior portions .
Drawn from life from the cultures of 1914 .
Magnification 90 diameters .
This larva is about , the stage of development at which , in 1917 , larvae were transferred from hypertonic water to normal water in the plunger jar .
Echinoderm Larvce with Two Water-Vascular Systems .
347 3\#151 ; Larva 18 days old viewed from the dorsal aspect , showing the first appearance of a rudiment of the right hydrocoele .
This larva was transferred to hypertonic water on the third day and remained in it .
Drawn from life from the cultures of 1914 .
Magnification 80 diameters , pig 4.\#151 ; Larva about 32 days old , from the cultures of 1917 , reared entirely in normal water , and viewed from the dorsal aspect .
Preservation 40-per-cent , formalin .
Magnification 80 diameters .
This larva and that represented in the two following figures give an idea of the stage of development attained by the larvse in the control plunger-jar when the experiments of 1917 were closed ; m.p. , madreporic pore ; st.c .
, stone-canal , the reference line points to the opening of this canal into the axial sinus ( ax .
) .
Plate 5 .
Fig. 5.\#151 ; Larva entirely similar to that represented in fig. 4 , viewed from the left side ; mad .
, madreporic plate ; Each .
, Echinus-rudiment .
Fig. 6.\#151 ; The same larva as that represented in fig. 5 , viewed from the right side ; calc .
, calcareous plate within the loop of the ciliated band midway between the two pedicellarise .
Fig. 7.\#151 ; Larva 40 days old , which had been exposed to hypertonic sea-water , viewed from the dorsal aspect .
Drawn from life from the cultures of 1916 .
Magnification about 50 diameters .
Note l.a.c. , the left anterior coelom , i.e. , the axial sinus into which the stone-canal ( st.c .
) opens ; r.a.c. , the right anterior coelom , expanded into a supplementary axial sinus ; r.hy .
, the right hydrocoele .
Figs. 8-11 , inclusive , represent larvse about 32 days old , from the cultures of 1917 , which have been exposed to the action of hypertonic water and have been subsequently transferred to normal sea-water in a plunger jar .
All have been preserved in 40-percent .
formalin , and all are viewed from the dorsal aspect under a magnification of 65 diameters .
Fig. 8.\#151 ; Larva with two hydrocoeles of almost equal size and no pedicellarise .
Note ab.t .
, aberrant tentacle , a lobe of the right hydrocoele projecting into the blastocoele ; m.p. , m'p ' .
, the two madreporic pores .
Fig. 9.\#151 ; Larva with the right hydrocoele smaller than the left but devoid of pedicellarise .
Note a.x. , the single axial sinus into which the two stone-canals st'.c ' .
) open ; m.p. , the single madreporic pore .
Plate 6 .
Fig. 10.\#151 ; Larva with two hydrocoeles of nearly equal size but with a single pedicellaria on the right side ( ped .
) ; m.p. , the single madreporic pore ; v.c.ep .
, the ventral ciliated epaulette seen through the transparent body-wall .
Fig. 11 .
Larva with two hydrocoeles , of which the right is small and without lobes .
No amniotic cavity is developed on the right side ; ax .
, left axial sinus ; ax .
' , right axial sinus ; m.p. , mthe two madreporic pores .
Two pedicellarise are developed ( ped .
) .
Figs. 12\#151 ; 14 represent larvse from the cultures of 1917 which had been starved for the first week and then generously fed .
They were about 32 days old , and were pi eserved in 40-per-cent , formalin .
Magnification 65 diameters .
348 Echinoderm Larvce with Two Water- Vascular Systems .
Fig. 12.\#151 ; Larva viewed from the dorsal aspect : l.sp .
, the left group of spines within the loop of the ciliated band ; r.sp .
, the right group of spines within the loop of the ciliated band ; I'.sp ' .
, spine on the left side outside the loop of the ciliated band ; r'.sp ' .
, spine on the right side outside the loop of the ciliated band .
Fig. 13.\#151 ; Larva viewed from the left side : m. , madreporic spine .
Fig. 14.\#151 ; The same larva as that shown in fig. 13 , viewed from the right side .
Figs. 15-23 represent transverse sections through larvae preserved in Bouin 's fluid .
All are from the cultures of 1917 , and are about 32 days old , except where otherwise stated .
Plate 7 .
Fig. 15.\#151 ; Section of the normal larva showing the mutual relations of axial sinus , madre-j poric vesicle , pore-canal and madreporic pore .
Magnification 120 diameters , i Fig. 16.\#151 ; Section of normal larva to show the mutual relations of the Echinus-rudiment and the pedicellarise .
Magnification 120 diameters .
Fig. 17.\#151 ; Section of larva with right and left pore-canals ( j and p'.c ' .
) .
In another section from the same series these canals are seen to unite , to open by a single median pore \#171 ; to the exterior .
Magnification 105 diameters .
Figs. 18 and 19 represent two sections of the same larva .
The larva possessed two well-developed hydrocoeles and one pedicellaria .
Magnification 105 diameters .
Plate 8 .
Fig. 18 shows the single pore-canal , the single median axial sinus ( representing the combination of right and left axial sinuses ) , and the relation of this compound axial sinus to the madreporic vesicle .
Fig. 19 shows the two hydrocceles : d.s. , dental sac of the left side ; d'.s ' .
, dental sac of the right side .
Plate 9 .
Fig. 20.\#151 ; Section of a larva with two hydrocoeles and one pore-canal .
The median axial sinus has caused the complete suppression of the madreporic vesicle .
Magnification 120 diameters .
Fig. 21.\#151 ; Section of a larva 40 days old , from the cultures of 1916 , which has two hydrocceles .
Two madreporic pores ( .
and m'.p ' .
) are present , but both belong to the left side .
Magnification 120 diameters .
Figs. 22 and 23 represent two sections of the same larva .
This larva belongs to the cultures of 1917 , and is one of those which was starved and then generously fed .
Magnification 120 diameters .
Plate 10 .
Fig. 22 shows the median madreporic vesicle ( m.v. ) .
Fig. 23 shows the groups of spines which replace the Echinus-rudiment and the pedicellariae .
McuzBrijobe , Roy .
Soc. Proc. B. .
90 , PI .
5 .
JfathJL ozudou Soc. J^oc .
J3 .
'p'oZ .
8 .
WacjBrtde ^c^BticLe del .
IfaoBrtd^ Hoy Soc. Proc. B. voL.90^1.9 .
Bride del .
Huth .
, London WacBride , Hoy .
Soc. Proc. P. vol. 70 .
23 .
\gt ; ?
MacJ3ride del .
Hixth London

