ANIMAL PARTHENOGENESIS 43 



and others on the larch, and all arising parthenogenetically 

 from individuals derived from a fertilised egg. Even more 

 complicated instances probably exist, for it is said that in one 

 species of Phylloxera twenty-one distinct forms occur between 

 one sexual generation and the next. 



In these cases which have just been described, we find 

 a more or less regular cycle of alternation between bisexual 

 and parthenogenetic generations ; in some cases the number of 

 the latter appears to be variable, and to some extent dependent 

 upon environment, in others it is fixed and regular. Perhaps 

 the best cases of regular alternation are provided by the gallflies 

 (Cynipidae), which have two generations in the year, one of 

 which consists wholly of females and reproduces parthenogenetic- 

 ally, the other contains both sexes, and lays eggs requiring 

 fertilisation. These flies lay their eggs in the leaves or buds of 

 trees (generally oak), and as the egg develops, a gall grows 

 round it, and is used as nourishment by the larva. Commonly 

 the parthenogenetic generation appears in the season when the 

 trees are without leaves, and the eggs are laid in buds ; from 

 the galls thus produced a bisexual generation arises in the 

 summer, the females of which lay fertilised eggs in the leaves. 

 Not only are the galls produced by the two generations very 

 different, as might be expected from their time and mode of 

 origin, but the flies themselves differ so widely that they were 

 commonly placed in separate genera before the relation between 

 them was known. The fly which has to lay its eggs deep down 

 in the buds of an oak has an ovipositor some ten times the 

 length of its daughter's, which has only to prick a soft leaf, 

 and many other characters show a correlated difference. 



It is interesting that a number of species of gallfly seem 

 to have dropped the sexual generation entirely; they now 

 produce only one generation in the year, which consists entirely 

 of females, and these, as far as is known, go on reproducing 

 by parthenogenesis year after year indefinitely without a male 

 ever appearing. 



Among the sawflies (Tenthredinidae) the conditions are some- 

 what different. In them, in some cases, we appear to get a 

 mixture of parthenogenetic and fertilisable eggs produced by the 

 same female, and in these it is usual, but not universal, that 

 the parthenogenetic eggs should give origin to males. But in 

 other species males are excessively rare or unknown, and as in 



