THE MOSAIC THEORY OF DEVELOPMENT. 9 



and which is accepted by all opponents of the mosaic theory, 

 namely, that the cell cannot be regarded as an isolated and 

 independent unit. The only real unity is that of the entire 

 organism, and as long as its ce;lls remain in continuity they 

 are to be regarded, not as morphological individuals, but as 

 specialized centres of action into which the living body resolves 

 itself, and by means of which the physiological division of 

 labor is effected. This view, at which a number of embryolo- 

 gists have independently arrived, has been most ably urged by 

 Whitman, in one of the lectures of this volume, though in 

 connection with a general conception of development peculiarly 

 his own. 



It is important not to lose sight of the fact that Hertwig, 

 no less than Roux and Weismann, conceives the idioplasm 

 (which he would locate in the cell-nucleus) as an aggregate of 

 units ( " idioblasts " ) which severally correspond to the heredi- 

 tary qualities of the organism ; and since cell-division is not 

 qualitative, every cell must contain the sum total of the heredi- 

 tary character of the species. Differentiation is conceived by 

 Hertwig (following de Vries) as the result of physiological 

 changes in the idioblasts, some of which remain latent, while 

 others become active, and thus determine the specific character 

 of the cell, according to the nature of the active idioblasts. In 

 regeneration such of the latent idioblasts are called into action 

 as are necessary to carry out the regenerative process. 



We have found good reason for the conclusion that the 

 mosaic theory cannot, in its extreme form, be maintained. It 

 remains to inquire whether the extreme anti-mosaic conception 

 rests upon a more secure foundation, and whether the mosaic 

 hypothesis may not contain certain elements of truth. I have 

 elsewhere more than once pointed out that the views of Hert- 

 wig and Driesch have received a strong bias, from the circum- 

 stance that the discussion has hitherto been confined mainly 

 to the echinoderm ^%^, which shows no visible differentiation 

 in the cells until a relatively late period (i6-celled stage). 



The whole question assumes a somewhat different aspect 

 when we regard such highly differentiated types of cleavage as 



