6 BIOLOGICAL LECTURES. 



really contradicts the mosaic hypothesis ; for the course of 

 events in the uninjured blastomere, or its products, is radically 

 altered by changes on the other side of the embryo. 



A more decisive result was reached in 1891 by Driesch, who 

 succeeded, in the case of Echinus y in effecting a complete sepa- 

 ration of the blastomeres by shaking them apart. A blastomere 

 of the 2-celled stage, thus isolated, gave rise to a perfect but 

 half-sized blastula, gastrula, and Pluteus larva ; an isolated 

 blastomere of the 4-celled stage produced a perfect dwarf gas- 

 trula one-fourth the normal size. Even in this case, however, 

 the earliest stages of development (cleavage) showed traces of 

 the normal development, the isolated blastomere segmenting, 

 as if it were a half-embryo, and only becoming a perfect whole 

 in the blastula stage. In the following year, however, the 

 writer repeated Driesch's experiments in the case of Amphi- 

 oxus (the egg of which is extremely favorable for experiment), 

 and found that in this case there is, as a rule, no preliminary 

 half-development whatever. The isolated blastomere behaves 

 from the beginning like an entire ovum of one-half or one- 

 fourth the normal size. 



It is quite clear that in AmpJiioxus the first two divisions of 

 the ovum are not qualitative, as the mosaic theory assumes, 

 but purely quantitative ; for the fact that each of the two or 

 four blastomeres may give rise to a perfect gastrula proves that 

 all contain the same materials. Nevertheless, in the normal 

 development, these cells give rise to different structures i. e., 

 they have a different prospective value from which it follows 

 that, in this case at least, differentiation is not caused by quali- 

 tative cell-division, but by the conditions under which the cell 

 develops. 



These facts are obviously a serious blow to the mosaic 

 theory, and the efforts of Roux and Weismann to sustain their 

 hypothesis in the face of such evidence only serve to emphasize 

 the weakness of their case. In order to explain the facts of 

 post-generation i.e., the capability of isolated blastomeres 

 to produce complete embryos both Roux and Weismann are 

 compelled to set up a subsidiary hypothesis, assuming that 

 during cell-division each cell may receive, in addition to its 



