544 C. R. KEYES THE GEOGRAPHIC CYCLE IN AN AJIID CLIMATE 



peneplain. Singularly enough, no large peneplain is known that still 

 remains near the baselevel with respect to which it was worn down. 

 On the hypothesis of regional lowering and leveling by stream action, 

 great and even unsurmountable difficulties are at once met with in at- 

 tempting to explain satisfactorily the larger relief features of deserts. 



Under conditions of aridity and with wind-scour as the chief denuding 

 power, there need be no recent regional uplifting in order to initiate the 

 arid cycle of erosion. The affected area may be an old plain of pene- 

 plain-like aspect, or it may be a vast plains surface frequently inter- 

 rupted by mountain ridges. Whether it could be ever an area occupied 

 entirely by lofty mountains is very questionable. Altitude, however, is 

 practically a negligible factor. The initial heights of some deserts were 

 doubtless several thousands of feet above sealevel. In view of the possi- 

 bility of desert-leveling, the flat-topped Bural-bas-tau and the associated 

 plateau-like highlands in the Tian Shan range in Turkestan need recon- 

 sideration, as Davis well observes. Because of the strong possibility of 

 its formation above baselevel in a region of inland drainage, Friederich- 

 sen^^ expresses objection to regarding it as a once low-lying peneplain, 

 as urged both by Davis^^ and Huntington.^^ In southwestern United 

 States, as represented by the remnantal plateau of the Mesa de Maya, 

 the initial surface of the desert of that region must have been at least 

 8,000 to 10,000 feet above sealevel. The high South African deserts 

 offer other examples. 



On the other hand, desert lowering by the wind goes on below the 

 level of normal peneplanation, not perhaps indefinitely below normal 

 baselevel, provided the sea be kept out, as urged by Penck,^^ but some 

 little distance below sealevel, until stopped by ground-water level, as ap- 

 parently in the cases of the Death and Imperial valleys, in California. 

 By deflation an arid cycle could be initiated on an old peneplain without 

 any change in elevation. 



The objections to ascribing a possible initiation of an arid cycle in a 

 mountainous region I have already pointed out. 



34 



ANTECEDENT RIVERS OF DESERT REGIONS 



On the basis of the normal cycle, the drainage features, or rather the 

 lack of them, in arid regions appear utterly inexplicable. Elevation of 

 surface, which is so all important in the introduction of a new cycle in 



^ Petermann's Mitteilungen, vol. xlix, 1003. p. 136. 

 3^ Appalachia, vol. x, 1904, p. 277. 



32 Carnegie Institution Publications, No. 26, 1005. p. 157. 



33 American .Journal of Science (4), vol. xix, 1905, p. 167, 

 3* Bull. Geol. Soc. America, vol. 21, 1910, p. 589, 



