34 Bulletin American Museum of Natural History, [^'ol. XXIII, 



is easily possible ; and there is no occasion to suppose that this may not occur 

 among larvae intermediate in size between those of the workers and females. 

 The only apparent objection to this view is one which undoubtedly occurred 

 to Wasmann himself and led him to suppose that the nurses actively trans- 

 mute the female larvse into workers, namely, the fact that the stature of the 

 pseudogynes is usually below, and often considerably below, that of the 

 largest workers. This, however, may be due to purely physiological causes 

 resident in the larvae themselves, for these would suffer starvation as a result 

 of the ants' infatuation with the Lomechusa larvae and hence lose much 

 of their substance (water and fat) by oxygenation during the period previous 

 to pupation and subsequent to the inhibited histogenetic changes that have 

 already progressed beyond the worker stage. Under these circumstances 

 pupation would wear the aspect of a regenerative or regulatory process 

 analogous to that which leads the isolated blastomeres of the sea-urchin and 

 other animal ova to develop into complete but more or less abortive and 

 diminutive larvae. 



The chief advantage of such an explanation lies in the elimination of an 

 appeal to special instincts such as would be implied by an endeavor on the 

 part of the worker ants to restore one portion of the colony — the vanishing 

 worker personnel, at the expense of another portion — the annual supply 

 of virgin females. Such an endeavor, though itself of the nature of a co- 

 lonial regeneration or regulation, is not improbable, but on the principle of 

 eyitia nan sunt multiplicanda prceter necessitatem, does not seem to be nec- 

 essary to an explanation of the phenomena. On general biological princi- 

 ples we should expect moribund ant colonies to take the opposite course 

 and hasten the development of the sexual forms as the most appropriate 

 method of insuring the survival of the species and thereby the production 

 or rejuvenation of colonies. The explanation above suggested has, more- 

 over, the advantage of being applicable to other cases besides the pseudo- 

 gynes of Formica. While treating of PJieidole instahilis (p. 3) I called 

 attention to the occurrence in that ant and in a few allied species belonging 

 to the Sonoran province, of a series of intermediates, or desmergates, as 

 they may be called, connecting the typical worker (ergate) and soldier 

 {dinergate) castes. As these desmergates are perfectly normal I could not 

 regard them as the result of Orasema parasitism like the phthisergates, 

 phthisogynes, and phthisaners. Moreover, similar intermediates are of 

 regular occurrence in species of several other genera (Eoiton, Dorylus, 

 Solenopsis, Azteca, Camponotus, etc.) where parasitism as an explanation is 

 out of the question. In all of these cases the desmergates probably arise 

 from larvae that have been neglected by the ants after having been enabled 

 to grow and develop beyond the typical worker stage. That such neglect 



