Sex- Determination in the Gall- Fly, 1ST. lenticular is. 185 



supply of galls was more than enough for the sleeves on the branches which 

 had been covered. About four to eight galls were put in each sleeve, and 

 since some of the parent flies had produced 50-60 galls, in some eases 

 a dozen or more sleeves were used for the galls produced by a single sexual 

 female in the previous summer. In all, over 80 lots of galls were sleeved 

 out. It would probabh r have been better, in some cases at least, to put only 

 one gall in each sleeve, but as my previous experience had shown that flies 

 when sleeved singly frequently fail to produce any galls, I judged it better 

 to put several galls in each sleeve in order to be more certain of getting 

 a result. As it happened, the season was extraordinarily favourable, and 

 when the leaves appeared those inside the sleeves were crowded with galls. 

 Not quite all were collected, but I hatched and noted the sex of about 

 9000 flies, as shown in the Table on pp. 186 and 187.* 



Inspection of this Table shows at once that the vast majority of flies reared 

 in the sleeves containing galls produced by any one sexual female are of one 

 sex, so that the sexual females A, B, F. H, J, and L had male-producing 

 daughters, while the others had female-producing. In most families, however, 

 there is a small percentage of exceptions, and the question arises, what is the 

 explanation of these ? If they occurred exclusively in the sleeves on branches 

 which had not been covered before the galls were sleeved out, it might be 

 assumed that they were produced by very early wild flies which had laid eggs 

 in the buds before the sleeves were put on the tree. They are, however, 

 about as numerous in sleeves on branches which had been covered as on 

 uncovered branches. Further it is to be noticed that the exceptional males 

 in female families are about as numerous as females in male families, and 

 that the number of exceptions varies largely from sleeve to sleeve in the 

 same family. For example, among the flies reared from sexual female C, 

 sleeves 22a, 23, 24, 2Aa, 25, 27a had no exceptions, with a total of 873 

 females, while sleeves 21, 22, 27, 28, 29 had a total of 501 females and 

 33 males, or over 6 per cent, of exceptions. The other families show the 

 same sort of thing. In all, of 80 sleeves (35 male-producing, 45 female- 

 producing), 18 of the male sleeves, containing 1,543 males, and 24 female 

 sleeves, containing 2,561 females, had no exceptions. Thus almost exactly half 

 the sleeves were without exceptions, and these sleeves were evenly divided 

 between the male and female producing, and contained nearly half the total 

 flies reared (4,014 out of 9,574). Frequently, also, the exceptions appeared 

 among the later flies to hatch ; this was only definitely recorded in the case 



* I take this opportunity of recording my indebtedness to Mr. F. Balfour Browne for 

 his kindness in preserving some of the flies as they hatched, while I was prevented by 

 indisposition from attending to them. 



