GENEOLOGICAL. 383 



tributions to biology have been momentous, pub- 

 lished the results of an extensive series of experi- 

 ments * on the same subject, and these were far 

 from harmonising with the conclusion reached by 

 Roux. 



According to Hertwig, if one of the first two seg- 

 mentation-cells (or blastomeres) be completely des- 

 troyed, the surviving half forms a fairly normal 

 embryo, with structural defects of slight importance. 

 If the destruction be partial, division may occur in 

 the injured half, either in its own strength or with 

 help from the intact half. But a destroyed half 

 cannot be revitalised, nor does Koux's post-genera- 

 tion occur. The development of the uninjured half 

 is quite normal. No half-gastrula or half-embryo 

 is ever formed, when one of the first two blastomeres 

 is destroyed. Therefore, as Hertwig concluded, the 

 mosaic theory of development is contradicted by fact 



We wish to dwell upon this particular case be- 

 cause it is so vividly illustrative of scientific method. 

 Here we have observers of equal competence reach- 

 ing discrepant conclusions from similar experiments 

 on the same material ! 



The puzzle was solved (in great part at least) by 

 the very careful research of Prof. T. H. Morgan, | 

 who showed that either a half -embryo or a whole half- 

 sized dwarf may result from the experiment, accord- 

 ing to the position of the blastomere. If, after one 

 of the first two cells has been destroyed, the other 

 be left in its normal position, then a half-embryo 

 results (11 cases) as Roux described. But if the 



* Archiv fur mikroskopische Anatomic, XLII. (1893), 

 pp. 662-807, 6 plates (with bibliography of 52 papers). 



t Half-embryos or whole-embryos from one of the first 

 two blastomeres of the frog's egg. Anat Anzeig., 1895. 



