MOSAIC THEORY 421 



by Hertwig is likely to be seen. The theory may find support 

 in the experiments of Morgan and Driesch on Ctenophore ova, 

 where a defect in the cytoplasm (not involving the nucleus) 

 is often followed by a modified cleavage and a defective embryo, 

 as if the architecture had been seriously injured ; but it may be 

 opposed by Delage's experiments on merogony, where a small 

 (and non-nucleated) fragment of a sea-urchin's %^'^ may be 

 fertilised and give rise to a complete larva. In some cases like 

 the last it seems impossible to maintain that different parts 

 of the egg are predetermined in relation to particular structures, 

 and the same conclusion is suggested by Wilson's experiments 

 on the lancelet ovum, where an isolated blastomere of the 

 four-cell stage develops into a complete larva. In other cases, 

 however, it seems as if the &gg had a fixed and set architecture, 

 which cannot be damaged without affecting the embryo. The 

 experimental evidence is inconclusive. 



The " Preformation " must be Nuclear. — But the researches 

 of Kolliker, Strasburger, Hertwig, and others led to a transfer- 

 ence of attention from the cytoplasm of the germ-cell to the 

 nucleus. From the importance of the nucleus in metabolism, 

 in the regeneration of Protozoon fragments, in maturation, in 

 fertilisation, and in cleavage, it was argued — most forcibly, per- 

 haps, by Weismann — that the nucleus must be the bearer of the 

 heritable qualities. Meanwhile, many were recognising the value 

 of Nageli's conception (1884) of a specific idioplasm — a complex 

 substance which, in its molecular organisation and in the meta- 

 bolism it induces, is different for each species. Weismann 

 developed this in his theory of the germ-plasm, which he re- 

 garded as wholly resident in the chromosomes of the nucleus. 

 Thus, the locality of the pre-established organisation was shifted 

 from the cytoplasm to the nucleus, though it is not inconsistent 

 with this to suppose that in certain cases at least a cytoplasmic 

 predetermination may arise as a secondary result of idioplasmic 

 influence from the nucleus. In plainer words, perhaps, the 



