120 Ha^edoorn. 



fact that unexpected "novelties"' in Drnsophila generally crop up in 

 stockbottles with masscultures excludes this t^st. In the second i)lace, 

 if a novelt}^ arises as an aabb novum from the mating of two pheno- 

 typically identical, AAbb and aaBB grandparents, it will, if compared 

 with either the AAbb or aaBB strain be a single recessive. 



How strong in the cii'cumstautial evidence that these novel re- 

 cessives, and for that matter the novel dominants in this material, can 

 have originated at least in part of the cases by a natural process, re- 

 combination of genes? In the first place it should be recalled that in 

 the beginning, when the strain was pure, no mutations were noted for 

 a great many generations. In the second place we know that the wild 

 strain, which is used by these workers as a standard to compare their 

 novelties to, has been repeatedly changed. If all the Drnsophila material 

 had originated from one original pair, or even from one original strain, 

 origin of novelties by recombination of genes would be out of the (luestion 

 as an explanation. 



It is cle;ir that the recombination-hypothesis would be, to these 

 workers especially, a veiy much more welcome hypothesis than any un- 

 explainable, spontaneous mutation. How can they have missed it, if 

 it should be the true one? The answer is clearly that they did not 

 hope to obtain a causative explanation, that their material was very 

 unfavorable and that their method of comparing every novelty to a wild 

 strain and of naming each gene after the change it brings about in a 

 wild strain, severely handicaps them. It is clear that all the wild strains 

 are phenotj-pically practically identical. Each wild fly is homozygous 

 for a gene which differentiates it from a whiteeyed one. If we call 

 such a gene "white" or "red'" we are apt to forget that it is only one 

 link in the chain of processes resulting in a red eye, and that two wild 

 strains, each homozygous for a gene differentiating it from a white eyed 

 type may possess different "redmaking"' genes, and therefore will, when 

 crossed, produce one in sixteen whiteeyed grandchildren. 



The objection may be raised that, if a novel character originated 

 as aabb from AAbb < aaBB, the gene studied by comparing it to a 

 wild type should not always be one, but sometimes one and sometimes 

 the other, so that we would expect to find "the gene" for the character 

 localized now in one, than in the other chromosome. But it is evident 

 that this objection does not hold true, if the most common wild type 

 is AAbb always. An occasional cross with an aaBB fly would produce 

 the aabb novelty, but in this case, only the gene A would be studied. 



