334 Bulletin American Museum of Natural History. [Vol. XXII, 



While it is equally clear that the ants of the rim can readily descend 

 to the lower limit of the coniferous trees, it is not so easy to account 

 for the origin of the fauna of the three lower zones. Spots like the 

 Indian Garden, which are veritable oases, confined to springs and 

 damp stream beds, probably derive their fauna from similar but 

 more extensive areas in New Mexico, Arizona and Texas. I have 

 taken all of the species of ants of the Indian Garden in similar loca- 

 tions in central Texas and the valley of the Rio Grande, with the 

 exception of Nylandcria imparis. This ant, so common in our northern 

 woods, occurs also in the mountains of Colorado and California and 

 may have been overlooked among the species occurring on the 

 Kohonino Plateau. One of the most abundant ants of the Indian 

 Garden, Formica gnava, is very characteristic of the banks of streams 

 in central and western Texas and parts of Colorado. 



Two sources suggest themselves for the fauna of the Angel Plateau 

 and the lower reaches of the canon, namely, the Desert of the Little 

 Colorado to the east and the more remote deserts about the Lower 

 Colorado and Gila Rivers in southwestern Arizona. The ant-fauna 

 of the latter region, as I can state from personal observation, com- 

 prises the very species found in the two arid zones of the canon, 

 in addition to several species which future collectors, working along 

 other trails, will probably bring to light. The alternative sources 

 above suggested are really reducible to one if we accept the view 

 advanced by Merriam in regard to the origin of the fauna of the 

 Desert of the Little Colorado, or Painted Desert, as it is sometimes 

 called. He says: "The Desert of the Little Colorado, it will be 

 remembered, is a deep basin on top of the Great Colorado Plateau. 

 It is wholly disconnected from the desert region of Southern Arizona 

 by the elevated and timber-covered highlands occupying the crest 

 of the plateau escarpment. In fact the highest part of Arizona south 

 of the Grand Canon, except a few isolated mountains, is the edge of 

 this plateau, which is nowhere below 2,130 meters (7,000 feet) and 

 in places rises to the height of 2,740 meters (9,000 feet), as at the 

 MogoUon Mesa. On the east the desert is separated from the valley 

 of the Upper Rio Grande by a broad area covered with cedar and pinon, 

 through which the continental divide passes, at an elevation of 

 upwards of 2,130 meters (7,000 feet). Therefore the only possible 

 channel through which the fauna and flora of the Painted Desert 

 could have reached this desert during existing climatic conditions is 

 by way of the Grand Canon of the Colorado. At first thought it 

 seems incredible that a fauna and flora should extend several hundred 



