328 SCIENCE PROGRESS. 



taken as the morphological unit, but this idea does not 

 apply to morphological conceptions of structures which are 

 themselves intra-cellular in nature. The only units of 

 which we can avail ourselves in such cases are constituted 

 by the microsomes and the more or less definite fluid reticular 

 in which they are suspended. In the individual cell, such 

 units are always arranged in an organised complexity, of 

 which the apparent differentiation into nuclear and cyto- 

 plasmic structures is the visible expression. And we have 

 seen further that, beyond this structural differentiation, any 

 superadded complexities of parts are always evolved during 

 the course of cellular ontogeny. Nevertheless, these super- 

 added structures are often purposive, and consequently 

 ought to be regarded as organs of the cell ; and it is one 

 of their most remarkable features that such similar and 

 obviously homologous cellular organs are apparently not 

 always evolved from antecedent structures of a similar kind. 



We have seen that the spheres in the metazoa are 

 apparently modified in a limited number of ways, and 

 although each type of sphere seems to be restricted to 

 similar kinds of cells in different kinds of animals, con- 

 cerning the origin or significance of these modifications 

 hardly anything is at present understood. 



The ascertained existence of attraction-spheres in re- 

 lation to the division of the Protozoa is in itself a fact of 

 great importance in comparative morphology, because it 

 banishes once for all any doubt as to the identity of the 

 protozoan body and the metazoan cell ; while the curious 

 similarity in the modifications of the proto- and meta-zoan 

 spheres, to which I have herein drawn attention [supra), is 

 a fact of which it is at present impossible to estimate the 

 full significance. It is, however, certain that the dis- 

 membered condition of the sphere in diatoms, and its 

 correspondence with a similar condition of the sphere in 

 mammals, are facts of more than merely passing interest ; and 

 yet they are in no way more remarkable than the corre- 

 spondence between the other types of proto- and meta-zoan 

 spheres. Few things, one would suppose, could be further 

 apart in the organic series than a Noctiluca and the 



