NEW FORMICID/E FROM BARRO COLORADO ISLAND. 165 



own colony, her diminutive stature, and the small size of her 

 gaster, which scarcely exceeds that of the worker. The fact that 

 the workers show only a beginning in the reduction of the eyes 

 would seem to indicate that the hypogaeic and inquilinous habit is 

 of rather recent phylogenetic origin. This supposition is also 

 supported by the consideration that the Attini themselves consti- 

 tute a young, or recent tribe of Myrmicine ants. 



It is evident that Cepobroticus is merely a single aberrant 

 species of Megalomyrmex which has abandoned an independent 

 life, has associated itself permanently with Sericomyrmex and has 

 taken to feeding on the fungus which it cultivates. The associ- 

 ation thus established is a type of "compound nest," as defined 

 by Wasmann, but differs from all the known types in certain 

 important particulars. The relationship between the two species 

 is somewhat like that obtaining between the xenobiotic Lepto- 

 thorax emersoni and Myrmica canadensis in the mountains of our 

 northern states and British America, but is in certain respects 

 much less intimate. Although the Cepobrotici look after their 

 own brood, they do not, like the Leptothorax, construct special 

 chambers communicating with those of the host. Mutual feeding 

 by regurgitation has not been developed, because both species 

 feed on a delicate plant which is carefully provided and cultivated 

 by one of them. We may, therefore, regard the relations of the 

 Cepobroticus to the Sericomyrmex as a case of what the Germans 

 call "Futterparasitismus," a case to which we might, perhaps, 

 apply the term "mycetometochy." With the possible exception 

 of the Pseudoatta described by Gallardo, we know of no other 

 example of this relation among ants, but further investigation 

 may reveal its occurrence among the termites of Africa and 

 Southern Asia. 



In the soil immediately surrounding the fungus chambers of 

 some of the Sericomyrmex nests I found five minute species of ants 

 which are described below as Pheidole (Hendecapheidole) mendi- 

 cula sp. nov; Oligomyrmex panamensis sp. nov. ; Solenopsis 

 conjurata sp. nov.; Tranopelta gilva Mayr var. columbica Forel 

 and Rhizomyrma sp. With the exception of the last these seem 

 all to be "thief," or lestobiotic ants, but further observations will 

 be required to establish their precise relations to the fungus 



