Ants from British Guiana. 373 



forms, wliicli swarm out to the attack. Tliey speedily ascend 

 the legs of the intruder, and iiaving secured a good grip with 

 their jaws, retain it with a bull-dog tenacity. They are 

 specially fond of all kinds of cultivated plants, and it is 

 almost impossible to cultivate any kind of plant in some 

 districts owing to their depredations. The leaves of the 

 Para Kubber Tree {Hevea braziiiensis) are readily attacked 

 despite the exudations of the sticky sap which often proves 

 fatal to other insects. No reliable method for their extinction 

 has yet been devised. I have frequently observed, though 

 have never been able to capture, a small species of fly 

 (apparently a Muscid) which hovers over the ants while 

 working in the daytime. From time to time the ants excrete 

 from the tip of their abdomens a tiny globule of liquid, and 

 as soon as this appears the fly darts down and rapidly absorbs 

 it ; the ants, though apparently uneasy, make no attempt to 

 drive away the intruder. 



Atta Acromyrmex octospinosa, Reich., ^ . 

 Issororo, N.W. District, 1915. 



A. [A.~) molleri, For., subsp. meinerti, For., 

 var. globocuUsy For. (in litt.). 



^. "This species appears to inhabit the interior; it has 

 never been met with on the coastlands. The communities 

 are never large, but frequently a number of communities are 

 met with in a small area. Formicarium with fungus- 

 chamber is invariably found within a decaying log of wood, 

 either just under or slightly above the soil-surface, and easily 

 accessible. I have always observed this species to utilise 

 freshly cut pieces of leaves and they are daylight workers. 

 This species also enters human habitations and will carry off 

 particles of food-stuffs. On one occasion the greater part of a 

 half-pound packet of dried raisins were carried off by these 

 ants. They have a distinct partiality for the foliage of Para 

 Kubber (^H. braziiiensis).*^ 



N.W. District, 1913. 



It would be interesting to know whether the number of 

 communities in a small area, spoken of by Mr. Bodkin, all 

 belong to the same colony, or are separate colonies. As far 

 as is known, the colonies of Acromyrmex are much less 

 populous than those of Atta, sensu stricto. 



