16 Cliroiiiosomen and Sex in Abnixas 



strong that in this figuiv there arc in f.ict 56 chromosomes in place 

 of the normal 55. I should not lay any great emphasis on a single 

 exceptional case in which only one equatorial plate is available, if it had 

 nc)t appeared, after the record of the exceptional count was published, 

 that family '12.25 was peculiar in another respect. The chroinosomc^s 

 were counted in the winter of 1912-13 ; when the moths hatched 

 in the following May and June, two females of this family, out of a 

 total of 16 reared, were ijrossulariata instead of lacticulor. The ])arents 

 were gruns. $ x lact. ,} ; the normal result of such a mating, owing to 

 the sex-limited transmission of the (jvu^s. character by the female, is 

 that all male offspring should be tjross., all females lact. The males 

 were all f/ro.ss. (21 were reared), but the presence of two gross, females 

 shows that in this case there was an exception to the normal sex-limited 

 transmission. 



I have very rarely had such excej)tions in previous years, but 

 since the moth is very common, and the possibility of the accidental 

 introduction of a wild larva with the food-plant difficult to guard 

 against with absolute certainty, I have always hitherto regarded such 

 exceptions as at least possibly due to accident rather than to failure 

 of the normal sex-limited transmission. In tiie present case, however, 

 such an explanation is impossible. The two exceptional females are 

 very closely alike (Figs. B, C), and show peculiarities in their markings 

 which make it certain, not only that they are sisters, but also that 

 they are daughters of the mother of brood '12.25 (Fig. A). Their 

 peculiar features are quite exceptional in wild (/rutmalariata (cf Fig. E), 

 but exist in the mother and other grossulariata ancestors of the family. 

 Some of these features, it is true, are found in collaterally related 

 families (Fig. D), and it might be argued that two larvae had been 

 accidentally transferred from another brood. Apart from the great 

 improbability of such an accident in carefully controlled experiments, 

 I have no other specimen which combines the peculiarities in the way 

 seen in the females in question, and it seems certain that they are 

 genuine exceptions. Similar exceptions, of course, are well known 

 in other species, e.g. Canaries (Miss Durham) and Pigeons (Staples 

 Browne). 



Now it is a very remarkable coincidence, if it be coincidence, that 

 the only exception to the number 55 in the oogonial chromosomes 

 should occur in the same brood as these two very rare exceptions to the 

 normal sex-limited transmission. If, however, it is not coincidence, the 

 facts fall quite simply into line. In the strain with 55 chromosomes, 



