1907.] Wheeler, Fungus-growing Ants of North America. 783 



perfect condition. On raising the bell-jar a peculiar odor is noticeable, 

 which Doflein believes to be a gas fatal to the insects. In the wild nests this 

 gas must be carried off by the chimneys which thus act as ventilating shafts. 



All of the foregoing observations relate to Old World Termites. One 

 is naturally led to inquire whether any of the American species raise mush- 

 rooms. Haviland was of the opinion that certain of the South American 

 forms such as T. dirus are "almost certainly fungus growers." The only 

 observations I have found on the habits of this species are contained in 

 Silvestri's work (1903). He says: ''I have seen in the galleries (Fig. 298) 

 pieces of grass 10 mm. long, of leaves 6-10 mm. long and twigs 30 mm. long 

 and 2 mm. in diameter. I have found such materials accumulated in small 

 quantities at various points in the galleries, but I believe that they are not 

 utilized in this form but are brought together in some more subterranean 

 portion of the nest for the development of a fungus on the mycelium of which 

 the termites feed." He found similar vegetable fragments in the nests of 

 T. (jrandis and moJesius. Of the latter species he says: "I was unable to 

 reach the center of the nest, but I succeeded in finding small masses of grass 

 with the mycelium already developed." From these, which he figures, he 

 concludes that the species grows, fungi. But these observations are by no 

 means conclusive as is evident from a comparison with the above cited 

 observations on the Old World species. These do not raise fungi on pieces 

 of dead leaves, twigs, etc., but on finely comminuted particles voided from 

 the alimentary canal and built up In the form of a sponge. Moreover the 

 temporary stores of leaves, etc. which are brought into the nests as food may 

 easily mould when left in the moist galleries. We may conclude therefore 

 that there is really nothing in Silvestri's observations to prove that any of 

 the South American termites eat and grow fungi. 



The most important study of the fungus-growing termites has been 

 recently contributed by Fetch (1906). Unfortunately I could not consult 

 this work till after the present article had gone to press, so that I am unable 

 to review it at length. Fetch carefully investigated the habits and fungus 

 gardens of the Ceylonese Termes obscurtceps Wasm. and T. redemanni 

 Wasm. In several particulars his account differs from those of Holtermann 

 and Doflein. I quote from the summary of his beautifully illustrated paper 

 the passages relating to the fungi for the purpose of showing how complex 

 and difficult are the problems with which the mycologist is confronted in 

 any critical study of the fungus-growing insects. After describing the 

 sponge-like combs in the chambers of the nest, he says:^ 



"The mycelium on the comb bears small white, stalked or almost sessile 



J In the quotation I have omitted the numerals belonging to the paragraphs and have 

 run the latter together. 



