( xliii ) 



to begin with that there are exceptional features about the 

 Nigerian fungus-farmers. They belong, you say, to the 

 genus Cremastogaster, and it woiild appear that those de- 

 scribed by "VNTieeler belong exclusiyely to the great New World 

 assemblage of the Attii. Further, so far as I can discover, 

 most of the Attiines make their cultures under ground, inside 

 their nests. 



" In the case of the Cremastogaster, the fungus is formed 

 outside and quite apart from the main carton nest, and the 

 culture is an aerial one, growing on a shelter of chewed wood 

 which the ants have made over a species of Coccid — surely 

 a imique case of mixed farming. The ants with their nests 

 occur on a laticiferous tree, a Ficus most probably, which 

 has grown, for support, round the large tree on which the 

 carton nest occurs, and, as you pointed out, they appear to 

 be extremely fond of the latex. 



" A further exceptional feature is the presence of the fimgus 

 in the form of a definite fructification, which I think I may 

 without hesitation refer to the Xylariaceae, a sub-family of 

 Sphaeriales, a sub-group of the Ascomycetes. The small 

 black growths are the so-called stroma of the fungus, which 

 may be simple or branched, and on this stroma two kinds 

 of spores may be produced, first asexual (conidia), produced 

 superficially over the upper part, and later a sexual ascigerous 

 stage. It is from this latter stage that the fimgus would be 

 determined specifically, but so far only the asexual stage has 

 been found. It most probably belongs to the genus Xylaria, 

 of which there are one or two British representatives. If the 

 ascigerous stage turns up it will be possible to get the fimgus 

 named. 



" In the case of the Attiines the identity of the fungi 

 appears to be largely a matter of conjecture. The ' kohl- 

 rabi ' growths or ' bromatia,' to use A\Tieeler's term, are 

 nearly always sterile. He mentions that Moeller foimd on 

 old nests of a species of Acromyrmex, an Agaric which he 

 named Rhozites gongylophora, but, for reasons which he goes 

 into in his book, Wlieeler doubts the validity of Moeller's 

 conclusion. 



*" To the fimgus cultivated by a primitive Attiine, CypJio- 



