THE MOSAIC THEORY OF DEVELOPMENT. 5 



In the hands of Weismann this theory attains truly colossal 

 proportions. The primary germs or units (which he calls 

 " biophores " ) are aggregated to form "determinants," the 

 determinants to form "ids," and the ids to form " idants," 

 which are identified with the chromosomes of the ordinary 

 karyokinetic figure. Upon this basis is reared a stately group 

 of theories relating to reproduction, variation, inheritance and 

 regeneration, which are boldly pushed to their utmost logical 

 limit. These theories await the judgment of the future. 

 Brilliantly elaborated and persuasively presented as they are, 

 they do not at present, I believe, carry conviction to the minds 

 of most naturalists, but arouse a feeling of scepticism and 

 uncertainty; for the fine-spun thread of theory leads us little 

 by little into an unknown region, so remote from the terra fi mm 

 of observed fact that verification and disproof are alike impos- 

 sible. 



In its original form the mosaic theory has, I believe, received 

 its death-blow from the facts of experimental embryology, 

 though both Roux and Weismann still endeavor to maintain 

 their position. It is rather curious that the very line of 

 research struck out by Roux, by which he was led to the 

 mosaic theory, should in later years have ended in a view dia- 

 metrically opposed to his own. In 1888 Roux succeeded in 

 killing (by puncture with a heated needle) one of the first two 

 blastomeres of the segmenting frog's ^^^. The uninjured 

 blastomere continued its development as if still forming a part 

 of an entire embryo, giving rise successively to a half-blastula, 

 half-gastrula, and half-tadpole embryo, with a single medullary 

 fold. Analogous results were reached by operation upon four- 

 celled stages. It was this result that led Roux to compare the 

 development to a mosaic-work, asserting that " the development 

 of the frog-gastrula, and of the embryo immediately derived 

 from it is, from the second cleavage onward, a mosaic-work, 

 consisting of at least four vertical independently developing 

 pieces." Roux himself, however, showed that in later stages 

 the missing half (or fourth) is perfectly restored by a process 

 of "post-generation," which begins about the time of the 

 formation of the medullary folds — a result which, in itself, 



