THE ANATOMY OF THE MADREPORAEIA. 9 



can be gathered from this form studied merely by itself ; it can only 

 be resolved by a comparative study of allied species. Differentiation 

 of function appears to be incomplete ; both forms are reproductive, 

 both apparently digestive. The most that can be said is that A is, 

 perhaps, more digestive and less reproductive than B, for the fila- 

 ments are more developed than in the latter form, and I have only 

 once observed an ovum on a modified mesentery. Should the 

 modification be digestive in function, as is probably the case, A might 

 certainly be termed a " gastrozooid." 



But at present any explanation of the function of the structure 

 above described, cannot be other than a mere speculation. It cannot 

 be regarded as a necessary result of the colonial habit, since nothing 

 similar occurs in the next species to be described— M. aspera. It 

 can hardly be connected with reproduction, as ova are of rarer 

 occurrence in the modified than in the unmodified polyps ; and an 

 excretory apparatus is not required by an organism whose cells are 

 capable of amoeboid activity, egestion as well ingestion. 



The only evidence on the point is derived from the distribution of 

 the zooxanthellas. These are most plentiful, firstly, in the external 

 canals just under the body wall ; and secondly, among the elongated 

 cells of the mesentery. Assuming, as we may fairly do, that nutri- 

 ment and aeration were the determining factors of such distribution, 

 it would seem that, in the first case, there must be a strong current 

 of nutritive " chyleaqueous fluid " (to use a word of the older 

 zoologists) in these external canals, and that aeration was effected by 

 diffusion of oxygen through the body wall from the surrounding 

 medium; and in the second place, that the elongated vacuolated 

 cells of the mesentery were in some way assimilative, while oxygena- 

 tion of the tissues for these special digestive processes (and therefore 

 secondarily and accidentally to the benefit of these symbiotic algae) 

 resulted from a constant stream of water flowing through the central 

 ectodermal canal of the mesentery. 



That such a stream does pass through this canal is extremely 

 probable, for the longer ectodermic cells are all morphologically on 

 the same side of the canal ; a wave of ciliary action must therefore 

 result in a current through the canal from one of the apertures into 

 the 8tomatoH;<:um towards the other. A comparison of Fig. 8 7 with 

 Fig. 6 M will explain this arrangement of the cells. 



