178 THE SECOND-CHROMOSOME GROUP 



were not used, because they were full of odds and ends of mutants 

 which might lead to confusion. The tests showed that very little 

 prcmaturation or repugnance is inherent in vestigial, the ratios being 

 exceptionally close to Mendelian expectation; wherefore it seems 

 probable that the shortage met with in the purple vestigial experiments 

 was due to some cause peculiar to the stocks used or to the culture 

 methods used in the experiments. Later tests with stocks descended 

 from these original stocks have failed to give such aberrant viability. 

 Another explanation that has been more recently appUed to partic- 

 ular instances in which a character ordinarily of excellent viability has 

 not appeared in the expected proportion, is that of a lethal gene. Thus, 

 an autosomal lethal in the second chromosome quite far to the 

 right of vestigial {i. e., close to speck) would give results roughly- 

 comparable to those observed. The difficulty with such an explana- 

 tion in this case is tliat the uniform results given by all the cultures 

 would require the lethal to be present in nearly all the indi\aduals, 

 a frequency entirely out of the question both from a priori consider- 

 ations and from the results of subsequent tests made with these stocks. 



THE PURPLE "EPIDEMIC"— MUTATING PERIODS. 



Shortly after the discovery of purple, purples or eye-colors closely 

 resembling purple began to be found in stocks and experiments every- 

 where. In the interval of 6 months following the discovery of purple 

 such occurrences numbered 14 and furnished the first as well as the 

 most striking of the "epidemics of mutation" that seemed to sweep 

 over our material at this period. From later and well-authenticated 

 cases (e. g., vermilion, cut, notch, etc.) it appears that certain muta- 

 tions do recur, and in the case of cut, four independent occurrences 

 followed one another so closely that the term ''epidemic" is descrip- 

 tive of the condition observed. How^ever, in the early cases (purple, 

 jaunty, arc, etc.) it is certain that a large majority of the apparent 

 cases were not true recurrences of the mutative change, but were due 

 to several other conditions. Thus, the first, fifth, sixth, and thir- 

 teenth of the apparent purples proved to be maroon, a third-chromo- 

 some ej'e-color practically indistinguishable from purple in appearance. 

 That is, "mimic" mutations were not at first distinguished from 

 the original type, nor were new mutant allelomorphs distinguished from 

 types already known unless the difference was striking. Certain others 

 of the occurrences were proved not to be of independent origin ; thus, 

 purples 8 and 9 were both show^n to have been descended from a 

 certain common stock, and purples 10 and 11 were traced to a second 

 common stock. It is undoubtedly true that in many cases where no 

 connection can be traced such connection really existed, especially in 

 the case of recessives, which might be distributed without giving sign 

 of their presence. The psychological element, too, is important; it is 

 exceedingly difficult to recognize a mutative change, even a striking 



