COMPARISON OF TURFAN STRATA 379 



It will readily be seen that the essential features of the Mesozoic coal 

 measures are similar to those of the Quaternary vegetal layers found 

 among the lake deposits. The Mesozoic coal appears to have been depos- 

 ited in swamps resembling those in which the lignite of recent times was 

 deposited. The ancient swamps were probably like the muddy plain of 

 today, except that water was more abundant and vegetation more luxu- 

 riant. The green shales, in the midst of which the coal measures lie, 

 were evidently deposited in a large body of standing water, either the 

 sea or a lake. No fossils have been found by which the nature of the 

 body of water can be determined. The very fact, however, of the absence 

 or rarity of fossils in shales so well adapted to their -preservation argues 

 against marine conditions. ' Continental conditions certainly prevailed 

 immediately before the deposition of the main body of green shales of 

 formation number 8, as the vegetal character of the coal measures proves. 

 They also prevailed immediately after the deposition of the green shales, 

 for the latter pass by gradual and almost imperceptible stages into the 

 typical subaerial strata of the formation numbered 10, full of thin lenses, 

 mud-cracks, ripple-marks, and other signs of deposition in very shallow 

 water, which covered the country at certain seasons only, leaving it dry 

 the rest of the time (plate 36, figure 1, and plate 39, figure 1). In view 

 of these facts, it is not improbable that the green shales of Turfan were 

 deposited in a large lake. If this is so, they, together with the inter- 

 bedded layers of coal, probably preserve the record of a strophe during 

 which the lake contracted during thesial epochs, permitting the growth 

 of vegetation. Similarly the alternating red and green strata at the top 

 of the green shales suggest a strophic period at the end of which there had 

 been such a change in the continental form of central Asia or in the 

 climate that the lake of earlier times disappeared. 



Characteristics and Origin of Eed Strata 



1. featvres indicating subaerial origin 



Eed beds have been spoken of above as indicating aridity and as indi- 

 cating subaerial conditions of deposition. It is not meant by this to 

 imply that there are no red beds which are of marine origin or which 

 were deposited under humid conditions. It appears probable, however, 

 that red beds of the kind here described are subaerial, and that the color 

 indicates aridity during at least part of the year. Although reasons for 

 this view have already been given, it seems advisable to restate them more 

 fully. Many red strata are characterized by frequent changes of text- 

 ure or structure. Such has been the case wherever I have examined 



