REPRODUCTION AND HEREDITY 323 



The veil which has so long covered the problem of heredity 

 begins to lift and at least for a short distance we are able clearly 

 to perceive the tangled paths leading into this labyrinth of the 

 numerous cases of the plant and animal kingdom which follow 

 Mendel's law and can be explained rationally, thanks to the results 

 of cell-investigation. 



We must, however, never forget tha't Mendel's law applies 

 only in the case of a proportionally small number of ' characters,' 

 now that there are more instances which do not follow the rule, 

 and in which the result of crossing is essentially different from what 

 the law would lead us to expect. Externally it is impossible to 

 observe in a character whether it will ' mendel ' or not ; breed- 

 ing experiments alone can supply the proof. But, as Julius Gross 

 points out, all cases of Mendelism have this in common, in spite 

 of their heterogeneousness, that the ' splitting ' characters 

 originated suddenly, by abrupt variation. Wherever we re- 

 investigate a case it will always be found that t the Mendelian 

 race represented by one or more was suddenly observed on a 

 flower-bed or in a field of the stem-form, without the presence 

 of the transition-forms between the two types. Indeed it is re- 

 garded as a rule for experiments with 'prevalence' and 'splitting* 

 that it is necessary to select two forms which in the first place 

 breed pure, and secondly form sharp contrasts to each other with- 

 out transition forms. But as for this reason old and good species 

 are excluded there remain on this ground alone only varieties that 

 have suddenly come into being.' ' This empirical result obtained 

 from the facts could also be obtained per deductionem. For varie- 

 ties that originated gradually and are still connected with the 

 type of the species by transition-forms must in crossing with the 

 species undoubtedly produce intermediate hybrids simply on 

 account of their own variability and their tendency towards the 

 type. Therefore, only forms that have suddenly come into 

 existence can exhibit the phenomena of Mendel's law.' In other 

 words, cases of Mendelism are mutants in the sense of De Vries. 

 Even in animals or plants in which one or the other character 

 follows the rule of prevalence and splitting we can observe that 

 the other characters of both are found in the hybrids in any 

 proportion. In order to put forward an explanation for this 

 phenomenon we are compelled to leave experiments and venture 

 on the ground of pure speculation ; the result will therefore only 



