8 Chromosomes and Sex in Abraxas 



but the chromosomes have begun to divide, so that several appear 

 double, and it is not impossible that two which appear to be distinct 

 are really separating halves of one. Brood '12.25 is not included in 

 the list ; the exception which occurs in it is discussed below. 



It appears from this list there are eighteen really good figures in 

 which the number 55 is perfectly clear, twenty-nine others in which 

 it is scarcely less certain, fourteen in which it is the most probable, and 

 only seven or eight in which it is doubtful ; and it should be noted that 

 of these doubtful figures, five are in ovaries of larvae in which the better 

 figures were counted as 55. I think, therefore, that on this evidence alone 

 it may be taken as certain that females of the unisexual strain, whether 

 they themselves belong to unisexual or bisexual broods, have 55 chro- 

 mosomes. Confirmatory observations from the maturation-figures of 

 the eggs will be described below. 



Comparatively few larvae with mi descent in the direct female line 

 from unisexual families were examined, since the number 56 had been 

 determined with confidence as characteristic of the species in previous 

 work'. Of those from which ovaries were sectioned, two wild fjmales, 

 three females from brood '12.10, and one from brood '12.22 provided 

 figures which could be counted; all showed 50 (Fig. 1), and were dis- 

 tributed in the various grades as follows: 



Excellent Wild ? B, one. 



Total 3 ., ? C, one. 



„ ? D, one. 



Good '12.10 ? A, one. 



Total 4 '12.22 ? C, three. 



Fair '12.10 ? A', one. 



Total 2 '12.22 ? C, one. 



Probable '12.10 t O, two. 

 Total 2 



These seven larvae, therefore, provided seven good or excellent 

 figures with 56, and four others in which 56 was almost certainly 

 the true number. 



One very important point appt'ared in addition. One wild larva 

 (^ A) not included in the list just given, provided four ovarian mitotic 

 figures, three of them of the ' excellent' and one of the ' probable ' class, 

 all with 55 chromosomes (Fig. 2). Two of them had, in addition, one 

 or two minute stained specks on the edge of the chromosome group, 

 in such a position that they might be regai'ded as chromosomes if they 



' Journal of Genetics, Vol. ii. 1912, p. 189. 



