ON THE NATURE AND CAUSES OF DIFFERENTIATION 317 



between Amphioxns on the one hand, and the snail or ctenophore 

 on the other, simply means, I think, that the initial differentiation is 

 less extensive or less firmly established in the one than in the other. 



We thus arrive at the central point of my own conception of devel- 

 opment, and of Driesch's later views; which were developed in a most 

 able and suggestive though somewhat abstruse manner in his Analy- 

 tische TJieorie der organischen Entwicklung(^g^\ and slightly modified 

 in a later paper published jointly with Morgan ('95, 2). The gist of 

 Driesch's theory is as follows. All the nuclei are equivalent, and all 

 contain the same idioplasm equally distributed to them by mitotic 

 division. Through the influence of this idioplasm the cytoplasm of 

 the egg, or of the blastomeres derived from it, undergoes specific and 

 progressive changes, each change reacting upon the nucleus and thus 

 inciting a new change. These changes differ in different regions of 

 the egg because of pre-existing differences, chemical and physical, 

 in the cytoplasmic structure ; and these form the conditions ("Form- 

 bildungsfaktoren ") under which the idioplasm operates. Some of 

 these conditions are purely mechanical, such as the shape of the 

 ovum, the distribution of deutoplasm, and the like. Others, and 

 probably the more important, are far more subtle, such as the distri- 

 bution of different chemical substances in the cytoplasm, and the 

 unknown polarities of the cytoplasmic molecules. 



A nearly related conception was developed with admirable clear- 

 ness by Oscar Hertwig ('94) nearly at the same time. Both 

 Driesch and Hertwig thus retreated in a measure towards the 

 theory of germinal localization in the cytoplasm, which both had at 

 first rejected; but only to a middle ground which lies between the 

 two extremes of the strict predestination theory and the theory of 

 cytoplasmic isotropy. For these writers now maintain that the initial 

 cytoplasmic localization of the formative conditions is of limited extent 

 and determines only the earlier steps of development. With each 

 forward step new conditions (chemical differentiations and the like) 

 are established which form the basis for the ensuing change, and so 

 on in ever-increasing complexity. This view is substantially the same 

 as that which I have myself urged in several earlier works, and I have 

 pointed out how it enables us to reconcile the apparent contradiction 

 between the partial development of isolated blastomeres of such 

 forms as the ctenophore, on the one hand, with the total development 

 of such forms as Amphioxns or the echinoderm, on the other. In the 

 latter case we may suppose the cytoplasmic differentiation to be but 

 ; feebly established at the beginning, and the blastomeres remain for a 

 time in a plastic state, which enables them on isolation to revert to 

 the condition of the original entire ovum. In the former case the 

 initial differentiation is more extensive or more rigidly fixed, so that 



