Coelenterate Nature of the Sponges. 93 



Moreover, the gastrula of these Parazoa differs sufficiently 

 from that of the Metazoa in the fact that in it the hypoblast 

 (endoderm) consists of cells provided with collars. 



This is essentially the course of SoUas^'s argumentation. It 

 will be seen from it that he lays the chief stress upon the presence 

 of tlie cells furnished with collars in the sponge-larvae, and 

 according to him they occur particularly in the amphiblastula 

 of Sycon raphanus. I do not know whether Sollas has him- 

 self observed these collars in this case ; except Kent, as 

 already mentioned, no mortal has yet seen them ; and so com- 

 petent a judge as Schulze says * : — " In my investigations of 

 the swarm-larvse of Sycandra raphanus, which can hardly 

 differ essentially in the structure of its larvae from Sycandra 

 compressa, and of many other sponge-larva3, I have endea- 

 voured, always in vain, to discover anything like the collar at 

 the free extremity of the cylindrical flagellate cells." And 

 he then shows very plausibly in what manner Kent's mistake 

 may have originated. 



My observations, which appear to agree with those of most 

 other investigators, have shown me that flagellate cells always 

 make their appearance in the canal-system of the Sponges only 

 when a current of water through the body of the animal is 

 possible, i. e. after the appearance of a gastro-vascular system 

 with a double communication outwards. The flagellate cells 

 are nothing but specially differentiated endodermic cells, 

 which originally have exactly the same appearance as all the 

 rest. This differentiation is due to division of labour : while 

 the flat endodermic cells chiefly effect the inception of nourish- 

 ment, the flagellate cells, by means of their flagella, produce 

 an energetic circulation of water through the body of the 

 sponge, and, chiefly by means of their collar, effect respiration. . 

 As in an Infusorian, an egg-cell, &c., the clear respiratory 

 plasma t collects as much as possible superficially, and turned 

 towards the point of access of the oxygen, so also in the 

 flagellate cells of the Sponges, both in those of the swimming- 

 larvai and in those of the flagellate chambers. But while in 

 the former the surface is more than sufficient for the reception 

 of a sufficient quantity of oxygen, in the flagellate cells the 

 case is different. They are therefore compelled to enlarge 

 their respiratory surface. But where can this be done ? Only 

 where the cells come into contact with the water containing 

 the oxygen. The rest of the body is more or less wedged in 



* Loc. cit. p. 182 (Ann. & Mag. Nat. Hist. loc. cit. p. 368). 

 t See A. Brass, * Biologisclie Studien, Th. I. Organisation der Thier- 

 ischen Zelle,' pp. 64 and 150. 



Ann. & Mag. N. Hist. Ser. 5, Vol. xvi. 8 



