BEES, WAXl'S, ANTS, ETC. 



509 



till September ; and it is only in this last month that the gall begins to form. As 

 soon as the larva has attacked some cells the increase is effected. Mr. Adler (of 

 Schleswig) has even proved that whilst the young larva has the hind part of its body 

 still enclosed in the membrane of the egg, a large proliferation of cells is formed in 

 front of it round the slightly wounded tissue." 



The reproduction of the Cynipidse is attended with some strange phenomena, which 

 have not been understood till quite recently. It has been found that a true alternation 

 of generations exists in many species of the family. This is of the kind deiined by 

 Balfour as heterogeny ; that is, there is an alternation of sexual and parthenogenetic 

 generations. In many cases the two generations of a species differ so greatly that they 

 have been placed by systematists in different genera. 



As observed by Adler, Neurot&rus lenticularis produces, on the under surface of 

 the leaves of the oak, galls which appear in July and fall to the ground in September 

 or October. The larva is at this period very small, and the perfect insect does not 

 come out till April or the beginning of May. Scarcely has it escaped from the gall 



FIG. 631. Galls of Cynips quercus-spongifica and C. quercus-inana. 



in which it was developed, when it deposits its eggs upon the buds of the oak. Around 

 these eggs are formed, upon the leaves and peduncles of the male flowers, galls differ- 

 ing from those which had nourished the Neuroterus. The insect which emerges from 

 them is no Neuroterus at all, but has been classed in another genus under the name 

 of Spatheyaster baccarum. This, in its turn, will deposit eggs which will produce 

 Neuroteri. The same alternation has been observed in three other species of Neuro- 

 terus corresponding to three distinct species of Spatlieg aster. " Not only do the two 

 generations live in galls differing in form, color, and situation, and the insects exhibit 

 among themselves differences of size, proportions, and structure, but what renders the 

 contrast more striking is that the Neuroterus generation is only represented by females, 

 ichilst the Spathegaster generation presents individuals of both sexes.''' 1 



Although alternation of generations occurs in many species of the Cynipidae, it is 

 believed that there are other species in which the parthenogenetic form exists alone, 

 that is, the species reproduced continuously without any males appearing. There can 

 be no doubt that these parthenogenetic species have descended from sexual forms. 

 Still it is said that no case is now known of a sexual form existing alone. "All the 

 sexual species are only known to us as a link in a cycle containing an organic genera- 



