R. K. Nabours 3 



cannot further be broken up, tliat the pattern is allelomorphic to each 

 of the others (all of them having been tried with several, and some with 

 all) ; in other words each letter represents one of a series of multiple 

 allelomorphs, fourteen being used. The Greek letter @ represents a 

 melanic pattern which can exist either in the single or double dose only 

 with, and in addition to, the true breeding patterns or their hybrids. 

 This pattern is allelomorphic only to its a&se?ice, if indeed it may be con- 

 sidered allelomorphic at all. It shows well with any of the multiple 

 allelomorphs, but perhaps best with BB, BG and CO where it is sliown 

 both in single and double doses (heterozygous and homozygous). At a 

 given time soon after moulting, it can be determined whether most 

 of the forms are heterozygous or homozygous for (8), but after a'few days 

 those carrying only a single dose of @ become melanic to the extent 

 that they are not readily distinguishable from those carrying a double 

 dose. The distinctions were not attempted except in a few instances, 

 because it would have required practically constant watch to make 

 the records at the proper time subsequent to moulting which occurs 

 irregularly after the third or fourth instar. However, the groups were 

 separated in matings (242), (244), (845), (246), (247), (2.54) and (255) 

 (see table at end of summaries on p. 40), and this is being continued with 

 extant ciiltures. 



In most of the crossings between different pui'e types the resulting 

 hybrid is readily discernible from either parent, usually intermediate, 

 though in some cases the one pattern is more apparent than the other. 

 In all crosses between any of the forms and A A, the hybrid superficially 

 resembles AA less than the other parent form, and in some instances 

 cannot be distinguished by casual examination ; e.g., the crosses AA x II, 

 A A X FP, and AA x SS produce progeny that can scarcely be discerned 

 from II, PP, and iSS, respectively. However, this is only an apparent 

 dominance of /, P, and /S' and others over A ; for when careful exami- 

 nations are made it is found that the pigmental elements of each pattern 

 are present in the hybrid in about equal proportions (5). 



In most instances where A and single and double doses of © were 

 involved I grouped together in the records the members of a progeny 

 supei-ficially similar in apjiearance. On this account some groupings 

 were made that should give the apparent 9:3:3:1 ratio (matings 

 (175), (176), (177) (AB® x AB®), and (185) {AG% x ACB)). Also on 

 this account there were from two to nine possibilities, including reci- 

 procals, in the parents, and from one to four in the progenies (in 

 matings (315), (446), (220), (221) and others), where the four kinds 



1—2 



