1907.] Wheeler, Fungus-groiving Ants of North America. 765 



texture. Exposure of only a fev\' moments to the air causes the dehcate 

 myceUum to wither and contract. The garden of the chamber represented 

 in Fig. 28 was thus dried, but the one in Fig. 29 was photographed imme- 

 diately after its removal from the nest. The ants appear to be crepuscular 

 or nocturnal. I have not seen them at work after ten o'clock in the morning 

 except on very cloudy days. 



On June 5, when I paid a second visit to the sandy country at iNIonto- 

 polis and Delvalle, all the nests were closed and the craters revealed no signs 

 of recent excavation. They had merely crumbled, marking the sites of the 

 nests as obscure little piles of sand. I opened several of the nests and found 

 the workers moving diligently about in their gardens, which were in fine 

 condition. On June 26, Avhen, just before leaving Texas, I paid a final 

 visit to the dry post-oak woods, not a trace of the nests could be found. 

 The wind and rain had completely obliterated the fragile turrets and fused 

 their sandgrains with the surrounding surface, so that even the closest 

 observer would never have suspected the existence of innumerable colonies 

 of little ants diligently cultivating their hanging gardens in the dark bosom 

 of the yellow sands. 



The foregoing description of the nests of Mycetosoritis shows that this 

 ant is closely related to Trachijmiirmex. The members of the genus Cijpho- 

 myrmex, as will be seen from the following accounts of two species have 

 very different habits. 



7. Cyphomyrmex wheeleri ForeJ. 



This species appears to be more widely distributed than most of the 

 preceding, since it ranges from Central Texas to California and probably 

 also over a large portion of northern ]\Iexico. In Texas it is rather rare 

 and, according to my observations, occurs only in arid regions, especially 

 on the Edwards Plateau and Grand Prairie and in the stony deserts of the 

 Trans Pecos country about Langtry and Fort Davis. Although several 

 of the preceding Attii prefer to live in dry localities among plant associations 

 of a more or less xeroplniic habitus, the abode of C. wJirrleri is characterized 

 by even greater aridity. Most of my ob-servations on the habits of this ant 

 were made among the lime-stone hills of the plateau escarpment just west 

 of Austin. Some of these hills, which are often beautifully stratified and 

 terraced and belong to lower cretaceous formations, are shown in Fig. 30, 

 from a photograph taken in the early morning when the long shadows 

 accentuate their j)eculiar structure. The terraced slo})es are strewn with 

 blocks of limestone of different sizes. Among these hills, from early spring 

 to late autumn, the heat and the glare of the sun reflected from the white 



