38 Psyche (April 



one of them a stinging form, were occupying the nest. On exami- 

 nation, the fragments thrown into the can were seen to contain 

 large numbers of both species, together with many larvae and pupae 

 of each. The larger ant proved to be a Ponerine, of the genus 

 Odontomachus. Males and females of Dolichodenis were also 

 found, but there was none of these phases of the other. But the 

 presence of many larvae in all stages showed that the sexual forms 

 of both species must have been present. During the two days 

 following I revisited the nest frequently, and stirred it up. Each 

 time both species of ant sallied out. Touching the nest lightly 

 would bring out Dolichodenis, but it required a more vigorous 

 prod to excite the other. My first examination had so disarranged 

 the nest that I could make no close study of its structure, but was 

 able to ascertain definitely that the Odontomachus were gathered 

 together in its deeper recesses, while the smaller species occupied 

 the peripheral chambers and galleries. No other nest was found, 

 nor did I again find either of the ants. This may be readily ex- 

 plained by the usual inaccessability of tree tops to the collector, 

 so failure to find more in no wise proves that the species are un- 

 common. 



It is unfortunate that only one nest was found as this leaves 

 some doubt as to whether the two species are normally parobiotic. 

 Odontomachus is, as Wheeler's observations have shown, ^ entomo- 

 phagus in habit. Mr. C. T. Brues tells me that in Grenada, B. W. 

 I., he found numerous colonies of a closely related ant, Anochetus 

 emarginatus Fabr., living in the nests of termites which form 

 their chief food. In this case the ant lives in a small part of the 

 nest constructed by the termites. This association is, of course, 

 not in any sense parabiotic, and there is a possibility that the 

 Brazilian Odontomachus may have been feeding on the Dolicho- 

 denis larvae. But it seems hardly probable that the latter would 

 remain in the same nest with a colony of ants which were destroy- 

 ing its young. That they had been associated for some time is 

 shown by the presence of all stages of the immature forms of both 

 species. Odontomachus is normally ground-inhabiting, nesting 

 beneath stones or in rotten logs, but it is not surprising that a 



'Biol. Bull. II, 1900. 



