THEORY OF THE EMBRYONIC STAGE OF ONTOGENY. 839 



sphere, which is in some cases a free-swimming larva. 

 This blastospere, however, becomes filled up by cells which 

 wander in from the external layer ; in most cases this 

 seems to take place from one end of the somewhat elongated 

 blastosphere, and in Aurelia this process is replaced by in- 

 vagination. Now the important point to notice in these 

 larvee is that during their free-swimming life the gut is func- 

 tionless; and this accounts for the fact that it is represented 

 by a solid rudiment. A precise parallel to the difference 

 between the endoderm of the Coelenterate and Echinoderm 

 larvse may be found amongst the larvae of the Ectoproct 

 Polyzoa. In the pelagic larva of Membranipora (Cyphonautes), 

 which has a long free-swimming life, we find a perfectly well- 

 developed gut with mouth and anus ; on the other hand, in 

 the larva of Alcyonidium we have a stomach of yolky cells, 

 an almost occluded oesophagus, and no intestine, whilst in 

 that of Bugula the whole mesoderm and endoderm is repre- 

 sented by a solid mass of cells. These larvae are developed 

 from yolky eggs and take in no nutriment during their free 

 life. I hold, therefore, that Heider 1 is perfectly justified in 

 his statement that the ancestor of the Coelenterata was " a 

 ciliated, oval, free-swimming form, in which by invagination 

 at the posterior end an archeuteron was formed." 



Lankester's view finds its chief support in the development 

 of Geryonia. In fact, this form is the only known one in 

 which such a process as he supposes to have taken place in 

 the blastula ancestor appears in the ontogeny. Are there any 

 reasons for regarding the development of Geryonia as specially 

 primitive ? I think we may fairly say none ; but that on the 

 contrary it shows manifest signs of its secondary character ; 

 the egg is yolky, and the development proceeds directly to the 

 medusa form, the hydra form being suppressed. The most con- 

 clusive argument, however, against Lankester's hypothesis is 

 that on his assumption the cavity of the blastosphere is identical 

 with the cavity of the future gut. Now all recent investigations 

 have gone to show that the blastocoele is the rudiment of the 

 ' ' Lehrbuch der Ver. Emb.,' vol. i, p. 81. 



