50 William Morton Wheeler 



recurved in the amber form. Moreover, the worker E. antiqua bears 

 the same relation in size to Mayr's type specimen and the male de- 

 scribed above, as does the worker E. longi to its female and male. 

 I believe, therefore, that I cannot be mistaken in my generic diagnosis. 



E. antiqua, as we must now call the species, acquires a peculiar 

 interest from the fact that hitherto only two species of the genus 

 have been described, one (E. longi) from Texas, and the other (E. peru- 

 viana Emery), from Peru. Only the female of the latter has been 

 seen. Since no species of the genus has been described from the Old 

 World, it would be easy to jump to the conclusion that we have here 

 a striking resemblance between amber and neotropical forms, but this, 

 in my opinion, would be premature, for the recent Erebomyrmce seem 

 to be very rare ants, and it is not at all improbable that living 

 species may yet be discovered in the tropics of the Old World. The 

 discovery of species m such widely separated localities as Texas, Peru 

 and the Baltic region proves, nevertheless, that the genus was once 

 cosmopolitan. 



E. antiqua is interesting also from an ethological standpoint. 

 Its pale, diminitive workers, with their vestigial eyes, show very 

 clearly that it was a hypogseic ant, the large, pigmented sexual forms 

 of which appeared above the surface of the ground only for their 

 nuptial flight. Emery, Forel and others have shown that several 

 species of the allied genera Aeromyrma, Carebara and Diplomorium 

 in the Old World, of Tranopelta in the New World, and of Solenoims 

 in both hemispheres live as thief-ants in the nests of other Formicidse 

 and of termites. As the type specimens of E. longi were taken near 

 termite nests by Mr. W. H. Long, and as termites are known from 

 the amber, we may safely infer that in its habits E. antiqua closely 

 resembled its living Texan congener. Social symbiosis and parasitism 

 in ants, therefore, are not necessarily recent acquisitions but may date 

 from the early Tertiary or even from the Mesozoic. 



The very small size of the worker of E. antiqua as compared 

 with the male and female would seem to indicate that the species 

 once possessed polymorphic workers like Pheidologeton, Oligomyrmex, 

 Aeromyrma and a few species of Solenopsis^ but had already lost all 

 but the smallest caste of these sterile forms during Oligocene times. 

 If this view of the origin of the discrepancy in size between the 

 workers and queens, which is held by Emery, is correct, the hypothesis 

 which I advanced in a former paper (1908) to the effect that the 

 polymorphism of the worker caste is of more recent origin than the 

 Oligocene, will have to be abandoned. 



