THE GEOGRAPHICAL CYCLE IN AN ARID CLIMATE 393 



and thus a larger and larger part of the central waste will be redis- 

 tributed and exported. As there is no relation of parts in the winds 

 analogous to that of small branch and large trunk streams in river 

 systems, the surface eroded by the winds need not slope toward 

 any central area, but may everywhere be worn down essentially 

 to the same level. The surface ever wearing down, the waste ever 

 washed irregularly about by the variable disintegration of the drainage 

 system and continually exported by the winds, a nearly level rock- 

 floor, nowhere heavily covered with waste, and everywhere slowly 

 lowering at the rate of sand and dust exportation, is developed over 

 a larger and larger area ; and such is the condition of quasi-equilibrium 

 for old age. At last, as the waste is more completely exported, 

 the desert plain may be reduced to a lower level than that of the 

 deepest initial basin; and then a rock-floor, thinly veneered with 

 waste, unrelated to normal baselevel, will prevail throughout — 

 except where monadnocks still survive. This is the generalization 

 that we owe to Passarge; it seems to me secondary in value only 

 to Powell's generalization concerning the general baselevel of erosion. 

 So long as the sea is held out, it would seem that a desert surface 

 might be worn even below sea-level, as certain writers have pointed 

 out in a general way (Penck, b, p. 167) ; but that such a desert should 

 persistently maintain a plain surface while it is slowly worn lower 

 and lower is a surprising result of deduction. Little wonder that 

 an understanding of the possible development of rock-floored deserts 

 of this kind, independent of baselevel, was not reached inductively 

 in western America; for there has been so much disturbance in the 

 way of fracture and uplift in that region during Mesozoic, Tertiary, 

 and Quaternary time that the attainment of arid old age has not 

 been permitted; but that the problem was not solved deductively 

 by the present generation of American physiographers before it was 

 encountered and solved by others in Africa serves to show how 

 insufficient still is the use of the deductive method among us. 



Passarge writes that his attention has been called to the difficulty 

 of explaining the vast plain surfaces of South Africa by wind-action, 

 because the wind has no baselevel of erosion, and it therefore can 

 and must excavate considerable hollows in rock areas whose waste 

 it can easily remove. He adds that this difficulty disappears as soon 



