THE GEOGRAPHICAL CYCLE IN AN ARID CLIMATE 391 



in relatively early and prevent the attainment of mature drainage 

 integration. In any case, as soon as the process of drainage disinte- 

 gration begins to predominate, maturity may be said to pass into old 

 age. 



This feature of the arid cycle has no close analogy with the features 

 recognized in the normal cycle. In the latter case, the drainage 

 systems of maturity tend on the whole to persist, even though the 

 streams weaken and wander somewhat — and according to theory lose 

 some of their adjustments — in very advanced old age; in the former 

 case, as old age advances, the integrated and enlarged drainage systems 

 of maturity are broken up into all manner of new and local, small 

 and variable, systems. The further results of drainage disintegra- 

 tion in the later stages of the cycle are even more peculiar. 



Leveling without baseleveling. — The later consequences of ero- 

 sion in an extensive arid region have been, as far as my reading goes, 

 first and recently stated by Passarge, in connection with his studies 

 of the arid regions of South Africa, as is more fully indicated below. 



As the dissected highlands of maturity are worn down, the rainfall 

 decreases, and the running streams are weakened and extinguished; 

 thus, as has been suggested above, the winds in time would appear to 

 gain the upper hand as agents of erosion and transportation. If such 

 were the case, it would seem that great inequalities of level might be 

 produced by the excavation of wide and deep hollows in areas of weak 

 rocks. As long as the exportation of wind-sw r ept sand and of wind- 

 borne dust continued, no easily defined limit would be found for the 

 depth of the hollows that might thus be developed in the surface, for 

 the sweeping and lifting action of the wind is not cotrolled by any 

 general baselevel. In an absolutely rainless region there appears 

 to be no reason for doubting that these abnormal inequalities of sur- 

 face might eventually produce a strong relief in a still-standing land 

 of unchanging climate; but in the actual deserts of the world there 

 appears to be no absolutely rainless region; and even small and occa- 

 sional rainfalls will suffice, especially when they occur suddenly and 

 cause floods, as is habitual in deserts, to introduce an altogether 

 different regime in the development of surface forms from the rock 

 hills and hollows which would prevail under the control of the winds 

 alone. The prevailing absence of such hill-and-hollow forms, and the 



