414 TKANSACTIONS OP SECTION D. 



continued to live in its new situation and has caused the skin covering it to 

 become modified into a lens-like structure. Hence we must conclude not only 

 that the optic vesicle secretes a substance which acts on the skin covering it and 

 compels this skin to become modified into a lens, but that all the skin of the 

 body is capable of undergoing this modification if acted on by the appropriate 

 stimulus. 



A third instance of the same kind has come under my own notice. During the 

 past few years I have been engaged in rearing large numbers of the pluteus of 

 the Echinus miliaris in the tanks of the laboratory at the Imperial College. This 

 pluteus is exceptionally favourable for observation because of its extreme trans- 

 parency. Since the development of Echinoderms is a somewhat specialised 

 branch of embryology, with which it is sufficient for moet zoologists to cultivate 

 only a bowing acquaintance, I may perhaps be forgiven for recalling to your 

 minds the sa.lient features of the development of this species. The plutei with 

 which Driesch experimented were reared up till the stage when they possessed 

 only four arms and a single pair of ccelomic sacs lying at the sides of the 

 oesophagus. In their subsequent development, however, the number of arms is 

 increased to eight, symmetrically arranged. Each ccelomic sac becomes divided 

 into anterior and posterior portions, and from the anterior portion of the left 

 side a small rounded vesicle, termed the hydroccele, is nipped off, which is the 

 rudiment of the adult water-vascular system of the ring, the radial canals, and 

 the canals of the tube feet. After its formation an invagination of the over- 

 lying ectoderm can be observed — this is the amniotic invagination. Its open- 

 ing becomes constricted, so that the invagination becomes flask-shaped and 

 finally closed, thus cutting off the sac from all connection with the exterior, 

 so that we have an ectodermic sac overlying a ccElomic one. From the floor of 

 this ectodermic sac are developed a series of pointed spines each with the 

 characteristic neuro-muscular ring surrounding its base and also the sensory 

 nervous ectoderm clothing the tube feet and from which the adult nervous 

 system is developed. The posterior coelomic sac extends forwards and inter- 

 venes between the stomach and the hydrocoele. From this sac are formed five 

 outgrowths surrounding the hydrocoele, which form the pockets of Aristotle's 

 lantern in the adult, from whose walls the teeth are developed. The stomach 

 develops an outgrowth in the centre of this circle which is the rudiment of the 

 oesophagus of the adult. On the right side of the larva there are normally 

 developed two pedicellarise each supported by a little calcareous plate on which 

 later little square-topped spines make their appearance. Now, it occasionally 

 happens, for reasons which I am investigating but have only succeeded in 

 partially elucidating, that on the right side of the larva a second hydrocoele is 

 developed from the right anterior coelomic sac, and in certain circumstances it 

 continues to develop. When this occurs, a second amniotic invagination is 

 formed on the right side of the larva, from whose floor a second series of 

 pointed spines is developed, whilst the pedicellarise and square-topped spines, 

 which should normally be formed, fail to put in an appearance. The right 

 posterior coelomic sac extends forwards between the second hydrocoele and the 

 stomach and develops a series of pockets which give rise to a second Aristotle's 

 lantern ; whilst the stomach gives rise to a second larval oesophagus in the centre 

 of these. We are thus driven to the conclusion that the ectoderm of the right 

 side of the larva is just as capable as that of the left side of forming a nervous 

 system and pointed spines, and that the right posterior coelom can form just as 

 well as the left posterior coelom the complex structure known as Aristotle's 

 lantern. 



When I brought these facts to the notice of Driesch as being very difficult 

 to explain on his theory of entelechy, he replied that he regarded them as an 

 instance of twinning, i.e., the formation of partial wholes, comparable to cases 

 of the formation of Siamese twins. Now, undoubtedly such twinning can occur 

 in EchinodeiTn larvw. Gemmill has published a most interesting account of such 

 twin larvfe of the star-fish Luidia, which he found developing from segmenting 

 eggs which had been fertilised in the West of Ireland and sent to him by post. 

 Gemmill rightly attributes the twinning to the partial separation of the 

 blastomeres due to the shaking which they endured on their journey. But no 



