140 THE SECOND-CHROMOSOME GROUP 



In most of these cases the character is also "specifically" truncate, 



and usually due to the original truncate gene rather than to a fresh 



mutation. In the absence of its usual intensifiers truncate may lurk 



unsuspected in a stock or an experiment for many generations and is 



difficult to eliminate. For this reason it is practically impossible to be 



certain in any unexpected case of truncate appearance that there has 



been a fresh mutation. It is not ordinarily profitable to pay any 



attention to these appearances of truncate, but in two instances in 



which, because of the pedigree, there was less than the usual likefihood 



that the truncate was due to the original gene, tests were made, and in 



both cases the character was found to be either truncate or else a very 



simikir allelomorph, though these tests could throw no new fight on the 



question of whether the gene were the original or a fresh truncate 



mutation. 



SNUB. 



The first of these tests was made by MuUer (unpubfished). The 

 second case appeared in the ninth generation of some experiments on 

 "duplication." A cross had been made between a female with the 

 new sex-linked recessive wing-character "cut" and a not-cut male 

 (February 17, 1916, culture 3338, Bridges). AU the daughters were 

 expected to be wild-type and all the sons cut; but 9 of the 77 daughters 

 were seen to be sfightly truncated, the character being called "snub," 

 while some 18 of the 67 sons were cut with shortened and blunted wing 

 ends (cut snub). 



One of these cut-snub males outcrossed to a wild fenoale gave about 

 a quarter of the Fi flies with the snub or truncate character. This 

 result showed that the character was a dominant, though a "poor" 

 one ; that is, not all the flies genetically alike (heterozygotes) showed 

 the character somatically. The snub appeared among the Fi males as 

 well as among the females, and this fact showed that the character was 

 non-sex-linked, for had it been sex-linked it could have appeared only 

 among the daughters of the above cross. 



When snub flies were bred together, the result was usually an approxi- 

 mation to a 2 snub : 1 not-snub ratio, often with the snubs below expec- 

 tation because of the above-noted occasional failure of heterozygous 

 snubs to show the character somatically. This ratio and the fact that 

 it proved impossible to obtain a pure-breeding snub stock suggested 

 that the mutant was lethal when homozygous, as are most of our 

 dominants. In cut-snub stock the approximation to the 2 : 1 ratios 

 was much closer, and it seems certain that the character cut favors 

 the differentiation of snub (see cultures 1 to 10, table 9). We are well 

 acquainted with such intensifiers or modifiers in other cases, and 

 tnmcate itself was known to be very susceptible to intensification. 

 A few of the cut-snub pairs gave nearly all of the flies snub (see espe- 

 cially 10 and 12, table 9, Morgan), and it seems probable that in these 



