376 E. HUNTINGTON- — GLACIAL PERIOD IN NON-GLACIATED REGIONS 



faulting is evident, not only from the steepness of the escarpment and 

 the- narrowness of the gorges which traverse it, but from the fact that it 

 cuts across terraces dating from the three arsial epochs corresponding, 

 apparently, to the first three glacial epochs of Europe. 



ABANDONED LACUSTRINE STRANDS OF TURFAN 



In spite of the complications due to the upheaval of the Fire moun- 

 tains, there remain on the south side of Turfan various evidences that the 

 climatic history of the region has been the same as that of Lop-Nor. It 

 could hardly have been otherwise, for at the time of greatest expansion 

 the borders of the lakes of Lop and Turfan were only a little over a 

 hundred miles apart. South of the playa of Bojanti, in a region where 

 there appear to have been no notable earth-movements during recent 

 times, two sets of old lake bluffs can be recognized. One lies only a few 

 feet above the floor of the playa. It may correspond to either the 20- or 

 35-foot strand of Lop-Nor. The other set of bluffs lies far back from 

 the playa and is almost completely shrouded in piedmont gravel. Its 

 base appears to lie at an elevation of between 100 and 200 feet above the 

 playa, but so much debris has been washed in from the upper parts of 

 the bluffs that the actual base is everywhere deeply buried. Judging by 

 the appearance of the bluffs, they must be at least as old as the 300-foot 

 strand at Lop-Nor. Further study would probably reveal other strands 

 corresponding to the five epochs of which such abundant evidence is 

 found in other parts of central Asia. My stay in the region was too 

 short for more than the most cursory examination. I found, however, 

 that many of the valleys of Turfan are characterized by five fluvial ter- 

 races like those of the Lop basin, Eussian Turkestan, and Persia. 



PLEISTOCENE (AND PLIOCENE) DEPOSITS OF TURFAN 



1. Lacustrine layers. — The upper bluff at Turfan is cut in uncon- 

 solidated clays. The succession of strata is shown in figure 14. The 

 section is far from complete. Upward it is prolonged by reddish de- 

 posits containing more or less sand and deeply shrouded in gravel. 

 Downward it continues indefinitely. The highest deposits lie 250 feet 

 above the playa, and the lowest almost at its level. The section here 

 shown, however, covers only 110 feet. Three unmistakable lacustrine 

 layers are evident. All three are older than the upper bluffs, and tAvo at 

 least appear to be older than the oldest strand at Lop-Nor. Here, as 

 elsewhere, we find some evidence of epochs of climatic change preceding 

 those recognized in glaciated countries. 



2. Vegetal layers. — An examination of the details of the Turfan sec- 

 tion discloses the fact that there are practically no red strata. Their 



