256 Dr. G. C. J. Vosmaer on the 



often (much oftener than at first one is inclined to believe) 

 upon the fact that the large aperture faces downwards, and 

 upon the relatively powerful current, the so-called gastral 

 cavity being the common canal, the cloaca, into which all the 

 other canals open. And even in cases where spicules project 

 into tlie '' stomach," which might eventually retain nutriment, 

 these spicules are constantly curved towards the osculum so as 

 to prevent entrance, but in no way preventing exit. 



Their developmental history teaches us that the Porifera 

 and Coelenterata separate from one another at a very early 

 date. As Heider correctly and expressly insists, the Sponge- 

 gastrula attaches itself by the mouth, while the Ccelenterate- 

 gastrula attaches itself by the aboral pole. Thus the two 

 types proceed together as far as the gastrula-stage, but then 

 each goes its own way. Finally, Balfour * has already 

 pointed out the early appearance and great development of 

 the mesoblast as a striking diiFerence between Porifera and 

 Coelenterata. Thus if I cannot agree with those who would 

 regard the Sponges as Coelenterates, I also do not agree that 

 they have descended from Coelenterates. 



In considering the question whether the Sponges have de- 

 scended from Protozoa we must^ in order to avoid misunder- 

 standing, distinguish between a direct descent {i. e. regarding 

 the question as Saville Kent does, and then, as a necessary 

 consequence, viewing the Porifera as a progressively deve- 

 loping group) and an indirect descent (supposing Sponges in 

 general or sponge-ancestors to have been derived as Metazoa 

 from Protozoon colonies) . The latter view appears to me the 

 most plausible. We can hardly imagine a direct descent. I 

 will not further urge the conclusion that Sponges are not 

 colonies of Monads or Choanoflagellata ; but the differences 

 between the Sponges of the present day and the Protozoa are 

 also so great that we can only properly discuss the question 

 whether the ancestors of Sponges descended from Protozoa ; 

 and in this sense I can only answer the question in the affir- 

 mative, it being still left quite uncertain in what manner the 

 transition was brought about. 



It is well known that Balfour started with the Amphi- 

 blastula larva, and saw therein the ontogenetic recapitulation 

 of a parent-form which stood between Protozoa and Metazoa. 

 He assumes that the cells of the two halves differentiated 

 themselves functionally into nutritive (the amoeboid cells) and 

 respiratory -locomotive (the flagellate cells). When the sponge 

 became attached these (locomotive) flagellate cells must for the 

 most part have become functionless, while the amoeboid cells, 

 * ' Comparative Embryology,' ii. p. 285. 



