Social Insects. 53 



of bees show gradations in these two kinds of females, and that 

 some species permit more than one queen or fertile female in the 

 colony and would refer for further details, both as to present 

 gradations and variations to the Notes, especially numbers 1, 2, 3, 

 and 4. Natural selection, if it has played any part at all, must 

 have done so chiefly in the manner ingeniously suggested by 

 Darwin himself, namely, not as between individuals, but as be 

 tween colonies. The tendency to produce arrested females or 

 neuters doubtless became fixed in some ancestral form through 

 social selection, and is kept up by this and colony selection. 



In the wasps we have a very different state of things, involv 

 ing the parthenogenetic production of arrested females and the 

 seasonal production of fully developed forms of both sexes. 

 Here again, the evidence all goes to show that the differences de 

 pend for each generation on the environment, food and method 

 of nurture of the larva, the tendency having become fixed in vary 

 ing degrees in the different species, and only so fixed by being 

 transmitted through the queen or sexually perfect females. So 

 far as natural selection has acted at all, it has acted on the poten 

 tiality or inherited tendencies of these females. Very exact in 

 formation is not yet at hand as to how far the neuters are vari 

 able, whether as to condition of the reproductive organs or as to 

 size. But judging merely by mounted specimens which I have 

 examined in various species, it is probable that there is some 

 variation in these respects, though the three classes are quite 

 neatly differentiated, much as in the bees. 



When it comes to the ants, the problem is more compli 

 cated ; but we may safely assume that the different forms 

 have been brought about by the same influences. In a large 

 colony of individuals, where size and character are not fixed 

 by a definite cradle, but where the young larvae are free and 

 are carried about, nursed and fed by the workers, there would 

 naturally arise greater variations between individuals, and while 

 the kind of nourishment, or the kind of nurture, or the age 

 of the female at the time the ova are produced, or the season of 

 the year, have doubtless all contributed to the variation, and may 

 still independently contribute to it at the present time ; yet, what 

 ever the causes of this variation, it has become fixed in certain 

 definite lines that are more or less useful to the species. Whether 

 or not the proportion of the different individuals is under the 



