ORIGIN OF RED STRATA 381 



cross-bedding. Among the modern sand dunes of Persia, Transcaspia, 

 and Chinese Turkestan these features are frequently seen, as appears in 

 the photographs of plates 36, figure 2; 37, figure 1, and 38. Among an- 

 cient deposits, they are finely exemplified in Utah in the massive red 

 sandstone of the Kanab formation and the white sandstone of the Colob 

 (plate 37, figure 2). They appear occasionally in the red sandstone of 

 various ages found in Turfan, the Lop basin, and Russian Turkestan. 

 Cross-bedding of the type here considered seems to be strong evidence of 

 both subaerial and arid conditions. 



Another indication of aridity in red beds is their relation to fossils. In 

 general, fossils are absent, and where they occur they usually appear to be 

 of terrestrial types. Barrell has shown that in many cases the few plants 

 of which traces can be found in red beds are adapted by their long 

 roots and other characteristics to arid conditions. He has also shown 

 that in subaerial strata redness is not consistent with the continued pres- 

 ence of ground water. Where the ground is always saturated the plants 

 leach out the oxidized iron from the soil as fast as it is formed ; hence the 

 soil can not become very red. Where there are no plants, on the contrary, 

 the oxidized iron remains in the soil, which, if long exposed to the air, 

 becomes red, unless, indeed, it happens to be highly quartzose or for some 

 other reason contains little iron. 



A good example of the relation of redness to the presence of water 

 impregnated more or less with carbonic acid and other vegetal products 

 has been called to my attention by Mr E. W. Sayles. In the Medina 

 formation, below Niagara Falls, the small faults which characterize the 

 region are brought into prominence by the fact that on either side of 

 them the red sandstone for an inch or so has been leached of its iron, and 

 is of a greenish tint. In other words, the presence of water which has 

 reached the fault fissures after passing through the atmosphere and 

 through a layer of vegetation, serves to nullify the conditions which 

 originally produced the red color. The obvious inference is that when 

 the red strata were formed no such water was present. Unless the cli- 

 mate was arid at the time of the deposition of the Medina, such water 

 must have been present, whether the strata are of estuarine, or, as seems 

 more probable, of subaerial origin. 



A final reason for believing that redness is in many cases the result of 

 aridity is found in certain observations which I made in the deserts of 

 Transcaspia in Russian Turkestan, and of Takla-makan, from 1,200 to 

 1,500 miles farther east, in Chinese Turkestan." These deserts, espe- 

 cially the latter, are far more arid than the deserts of America. They are 



