Social Insects. 55 



I have offered for the facts, in discussions with friends and be- 

 fore the society, limiting the action of natural selection to colonies 

 as a whole. Few persons who have not had large experience in 

 rearing insects can appreciate the full influence of larval environ- 

 ment and food on the ultimate imago, or the power of larval 

 accomodation to various conditions. All insects in the larva 

 state possess this power, within varying limits, and it is nowhere 

 more marked than in the Aculeate Hymenoptera. I have called 

 attention to it on numerous occasions* when treating of parasitic 

 species, and it is particularly noticeable in the fossorial Hymen- 

 optera and the Melouht;. Size, especially, may easily be dimin- 



where they state that " such conditions as deficient or abnormal food " and 

 others " causintjc preponderances of waste over repair * * * tend to re- 

 sult in tiie production of males," while " abundant and rich nutrition " and 

 other conditions which "favor constructive processes * * * result in 

 the production of females." He then cites J. H. Fabre's statement that in 

 the nests of Osmia tricorniH the eggs at the bottom of the cell which are 

 first laid and accompanied by much food, produce females, while those at 

 the top, laid last and accompanied by one-half or one-third the (juantity of 

 food, produce males. (Souvenirs Entomologiques, Seme serie, page 328). 

 He further refers to Iliiber's observations, that the queen bee only lays 

 eggs of drones when declining nutrition or exhaustion has set in, and that 

 when the workers in bees and wasps lay eggs, these produce drones. 



These statements are not entirely justified. I cannot speak ]>ositively of 

 Fabre's observations, though I suspect something back of the larval food- 

 supply which has fixed the sex and determined the treatment of the larva. 

 But the queen bee produces drones at any age by the egg passing into the 

 drone cell and not being impregnated in passing the spermotheca. She pro- 

 duces drones only when she is superannuated, because the spermatozoa have 

 become exhausted. In wasps it is just the contrary, the unimpregnated egg 

 producing ordinarily, not a drone or a male, but a female. I have already 

 called attention to the ease with which erroneous conclusions are drawn in 

 this matter of regulating sex by food of larv;e, ex ovo (Am. Naturalist, Vol. 

 VH, pp. 513-531, September, 1873) and the evidence would seem to 

 how that the influence is confined to arrestation or modification of the sex 

 without changing it. The subject is, however, most intricate, and further 

 experimental facts are needed. Spencer's conclusion is, nevertheless, gen- 

 erally true, namely: ^^ that one set of dijf'erences in dnidure and instincts is 

 determined l)ji nutrition hejore the egg is laid, and a further set of differences in 

 structure and instincts is determined by nutrition after the egg is had. ' ' 



*See notes on Tiphia inornata, Sixth Report on the Insects of Missouri, p. 

 123, and upon Blister-beetles, Fu-st Report U. S. Entomological Commission, 

 pp. 295-302, 



