266 INTERNAL FACTORS IV. i 



qualitative but always a quantitative process. The combination 

 is handed on in its entirety to every cell of the body. The 

 chromatin has in this sense an individuality, for however much 

 the chromosomes may disappear while the nucleus is at rest, the 

 set of qualitatively distinct units always re-emerges (though not 

 necessarily in the same order) to be quantitatively divided between 

 the two daughter-cells. We reach, therefore, a conception of the 

 chromatin akin to Roux's idea of the ' Reserve-idioplasson ', 

 the germ-plasm which is passed on entire to all the tissues of 

 the body. 



And now Boveri proceeds to develop a theory already suggested 

 by Driesch of the part played by the chromatin in differentia- 

 tion. The cytoplasm is heterogeneous, anisotropic; and these 

 differences in the cytoplasm are sufficient to account for the 

 earliest differentiations, the pattern of segmentation, the deter- 

 mination of the embryonic axes and possibly some few others. 

 But then the chromatic elements come into play not the 

 chromosomes, which are too large and too few to be regarded 

 as determinants but far smaller subdivisions of these. Under 

 the influence of the cytoplasm, different in the various cells, 

 some of the units become patent and active, while others 

 remain latent, and by the reaction of the former upon the 

 cytoplasm new differentiations arise which will in their turn 

 call forth into activity new nuclear elements. 



Whatever value may be attached to these speculations, there 

 can be no doubt of the importance of the experiments on which 

 they are founded. Provisionally we have good ground for believing 

 in the dissimilarity of the chromatic elements and of their 

 necessity for differentiation. It only remains to be seen whether 

 further investigations on other forms will bear out the interpre- 

 tation which Boveri has placed upon his observations. 



There is one more point worthy of notice. An ordinary 

 sexually produced individual possesses in its nuclei 2 n chromo- 

 somes, that is, two complete sets. In maturation one division of 

 the chromosomes is transverse, and Boveri suggests that in this 

 transverse division homologous units are separated from one 

 another, or, to use the expression made familiar by current 

 Mendelian theory, members of pairs of allelomorphs. 



Before bringing this section to a close a brief reference must 



