THE Ml TATED GE \ I 7 



their phenocopies, according to the way he treated hia animals 

 with abnormal conditions. This, however, has never been 

 verified. Heche's (1907) experiments changing the feathering 

 of the dove ScardafeUa inca hy action of moisture also belong 

 here. They resemble Standfuss' experiments in so far as the 

 type was shitted toward the type of a known but unanalyzed 

 geographic race 



The first controlled experiments with genetically known char- 

 acters were made hy TimofeefT (1926), who reported that some 



[nutations found in Drosophila funebris were paralleled by 

 identical mollifications. In one case, a change of temperature 



COllld produce one of these phenocopies. Contorted wings. 



A systematic study of this problem was inaugurated by 

 Goldschmidt, who reported in 1929a that it is possible to repro- 

 duce the phenotypic likeness of many mutants of Drosophila by 

 treating larvae of a definite stage with temperature shocks, and 

 this in a quite regular manner. The main results as published 

 later (1935a) are: 



If Drosophila larvae of a definite age are treated with tempera- 

 ture shocks. of 35 to 37°C, a considerable number of the flies 

 hatching from such larvae are more or less modified. Among 

 these modifications we find many types of abnormalities and 

 simultaneously the phenocopies of numerous known mutations. 

 Among these are most frequently the changes of bristles, copying 

 practically all known types of bristle mutations, e.g., forked, 

 bobbed, stubble; further, the copies of all types of wing mutations 

 like miniature, arc, (airly, pointed, dumpy, scalloped; further, the 

 copies of many types of mutation of eye form like small eye, lozenge, 

 star; furthermore, rare but sufficiently frequent copies of differ- 

 ent mutants of body form and appendages like aristaless, daehs. 

 abnormal abdomen, and also such types as benign (tumor); 

 more rarely copies of changes in color and design like sooty, 

 t rident. All of these phenocopies appear not only in the complete 

 form represented by the well-known mutants but in a series 

 of different grades resembling all types of known multiple allelo- 

 morph- or such as are to be expected without having been found. 

 Generally speaking, one might say that all groups of known 

 mutants are represented and a considerable number of individual 

 members of each group. As only one method with not too many 

 variations was used, one might expect that a proper variation of 



