ANIMAL PARTHENOGENESIS 51 



genetic generations found, for example, in the aphids. These 

 present a somewhat puzzling problem. It is frequently 

 assumed that differentiation takes place at the reduction 

 division during the maturation of the germ-cells ; for example, 

 the segregation of characters in a Mendelian hybrid (hetero- 

 zygote) is commonly regarded as taking place at this stage. 

 But in Chermes or Phylloxera a differentiation is found among 

 the offspring of a single parent quite as striking as that in 

 Mendelian segregation, and yet no reducing nuclear division 

 takes place. The two cases are of course not quite comparable, 

 for the Mendelian segregation separates characters (allelo- 

 morphs) permanently, and in the aphid certain characters 

 become latent, and reappear in later generations. But clearly 

 the cause of these differences between children of one parent 

 must in this case be sought in something other than chromosome 

 reduction. 



The problems of sex are intimately connected with partheno- 

 genesis, but although in some cases we seem to approach a 

 solution, others offer little help, or may even appear 

 contradictory. 



In the bee and a number of other Hymenoptera sex appears 

 to be determined by fertilisation. Virgin eggs yield males, 

 fertilised eggs females. Hence it has been assumed by some 

 that the spermatozoon bears the sex-determinant. But other 

 parthenogenetic species produce both sexes from unfertilised 

 eggs — for example, gallflies, aphids, and daphnids. In a few 

 cases the parthenogenetic eggs which will yield females produce 

 one polar body, those giving males two ; hence it was suggested l 

 that the female determinant might be separated from the eggs 

 at the second polar division. But, again, in the aphids and 

 daphnids eggs producing both sexes have only one polar 

 division, and among the Hymenoptera females arise in many 

 species from eggs which have two. 



But one conclusion on this matter can be made with some 

 confidence from the study of parthenogenesis, viz. that the 

 sex is determined from the beginning of development, and is 

 not as a rule altered by circumstances affecting the animal 

 afterwards. A bee-larva may be made into a queen or a worker 

 by circumstances, if it is female to begin with, but if it arises 

 from an unfertilised egg no treatment will turn it from a 



1 Castle, Bull. Mus. Harvard, xl. 1903, No. 4. 



