THE POSITION OF SPONGES, ETC. 441 



vacuoles at the base of the ceil and the nucleus in the 

 upper portion, but as a matter of fact in this species the 

 nucleus is at the base of the cell and any vacuoles that 

 occur are lodged in the upper portion. Hence the alleged 

 occurrence of contractile vacuoles in collar cells requires 

 further confirmation before such vacuoles can be reearded 

 as constant features of the cells. A second point with 

 reference to the collar cells is their multiplication. We 

 have already noticed Kent's statement to the effect that 

 they multiply by means of spores. He also asserts that 

 the ciliated chambers develop from "the segmentation of a 

 primitive amceboid body produced ... by the coalescence 

 ot more or fewer metamorphosed collared zooids ". These 

 assertions, so far as they have been noticed by subsequent 

 observers, have generally been regarded as erroneous 

 and without value. The fact remains, however, that both 

 the multiplication of collar cells, and the formation of new 

 ciliated chambers, in a growing sponge, are scarcely known, 

 and no satisfactory observations have been recorded with 

 reference to this point. Nevertheless the similarity of 

 collar cells and Choanoflagellata remains a very striking 

 fact, and furnishes the strongest argument, not only for 

 those who seek to class sponges as Protozoa, but also for 

 the authors who look upon them as descended from Pro- 

 tozoa independently of the Metazoa. Those, on the other 

 hand, who regard sponges as Metazoa are obliged to 

 explain the collar cells as secondarily acquired. Schulze 

 (1885, p. 190) considers that the independent origin of 

 collars in the cells of sponges is by no means impossible, 

 since protoplasmic processes and membranes of various 

 kind are of common occurrence in Protozoa. It would, 

 however, have been more to the point to prove that similar 

 structures occur among Metazoa. 



While the collar cells have furnished the strongest argu- 

 ment for separating sponges from Metazoa, the peculiar 

 canal system of sponges is a feature which, as may be seen 

 from the opinions quoted above, has very often been used 

 on the other side, especially by the advocates of the coelen- 

 terate theory. Not only was the central gastral cavity of 



