ON PARTHENOGENESIS IN THE HYMENOPTERA. 195- 



produced from unfertilised eggs in the hive-bee ; 

 or that if one or the other sex were yielded by 

 unfertilised eggs, that the sex so produced must 

 of necessity be male. As a matter of fact, the 

 phenomena of parthenogenesis in the Hymenoptera 

 (not to go beyond the order to which the bee 

 belongs) are very complicated. There are, in truth,, 

 four kinds : — 



(First) — Species which only produce males partheno- 



genetically ; 

 (Second) — Those which yield males and females ; 

 (Third) — Those which have no known males, and 



consequently are only known in the female sex;, 



and 

 (Fourth) — Species in which a sexual generation 



alternates with an unisexual generation. 



With facts like these before us, we need not be 

 surprised if it be found that the drones may fre- 

 quently show signs of having had a father. I have 

 just given an extract from Mr. Cheshire's book to 

 the effect that, if (say) an Italian queen be mated 

 with a black English drone, the drones born from 

 her will be pure Ligurian. It is not clear, however, 

 that this is always the case. In 1867, Mr. John 

 Lowe read a paper before the Entomological 

 Society of London,* in which he gave the results of 

 numerous experiments conducted specially with the 

 object of testing the truth of Dzerzon's theory. 

 These consisted of mating Apis mellifica, A. ligustica, 

 and A. fasciata, in order to find out if drone hybrids 

 would be produced. He found that not only did 

 the workers exhibit traces of their mixed parentage,, 

 but that the males were equally affected, bearing 

 unmistakable evidence of having had a father. Mr. 

 Frederick Smith,f who had made a special study of 

 the species of honey-bees, testified that Mr. Lowe's 

 drones bore unmistakable hybrid characters. 



Since then, various other workers have recorded 



* Irani . Ent. Soc. (Third Series), vol. v., pp. 547-560. 

 jFroc. Ent. Soc. 1867, p. xcii. 



