of the Sponges to the Ohoanoflagellata. 375 



terata from very early times, and led to the development of 

 urticating capsules. 



In general Marshall is inclined to conceive of the Sponges 

 as retrograded Coelenterata, the coelenteric apparatus of which 

 originally resembled that of the higher Coelenterata, and like 

 that possessed aquiferous pores, but at first did not employ 

 these for the reception of nourishment. It was only later, 

 according to Marshall's view, that a change of function took 

 place in the Sponges, the water, and with it the food, being 

 inhaled through the external pores. During this change 

 any tentacles that may have existed disappeared, together 

 with the urticating capsules, and the afferent canal-system 

 underwent a special development. 



Undoubtedly the decision in favour of one or the other of 

 the two opposite views can only be arrived at with any 

 degree of certainty when a thorough knowledge of the 

 ontogeny of numerous Sponges and Cnidaria justifies definite 

 conclusions as to the phylogenetic development of the two 

 groups. 



What we at present know of the ontogeny of the Sponges is 

 not, to my mind, in favour of the correctness of Biitschli's 

 hypothesis. For if the Sponges had really originated from 

 colonies of Ohoanoflagellata, and been indebted to this cir- 

 cumstance for their collared cells, we should expect that in 

 the ontogenetic development of the Sponges the collared cells 

 would make their appearance in that phase which corresponds 

 to the phylogenetic stage of a Protozoan colony, namely in 

 the blastula. This would indeed really be the case if Saville 

 Kent's representation were correct, according to which the 

 blastula or " swarm-gemmule " (at least of Sycandra com- 

 pressor) consists of a layer of collared cells. 



But as the sponge-larvai with which we are acquainted do 

 not possess these collared cells, but, like the larvaa of the 

 Cnidaria, simple flagellate cells, and, also like the Cnidarian 

 larvas, have already attained the Metazoan stage by the 

 differentiation of two distinct cell masses before the meta- 

 morphosis into the typical sponge, and consequently the 

 formation of the collared cells, commences, it follows that the 

 constitution of the sponge-larvre favours, not the independent 

 origin of the Sponges from Choanoflagellata, but rather a 

 close relationship of the Sponges to other Metazoa, such as 

 the Cnidaria. It is true that in this way the agreement 

 between the collared cells of the Sponges and the Choano- 

 flagellata becomes more difficult to understand. But still the 

 possibility of a spontaneous production of the collar in the 



26* 



