— 104 — 



burrowing rodents. In the terms used by Cowles this met- 

 amophosis is a topographic succession, due to topographical 

 changes of the surface. Korshinsky regards the matter from 

 a different point of view. His opinion (1896 pp. 6, 8) is that 

 the "normal type" of sand-desert is the Desert-plain "level 

 or undulating areas of sand" covered with Haloxylon, Salsola 

 Arbuscula, Calligonum, Ephedra, Ammodendron and Eremospar- 

 ton, and during spring with a number of herbaceous spring- 

 plants; this desert- tree vegetation Korshinsky regards not only 

 as the most characteristic, but also as the original type (comp. 

 above p. 36). He is of opinion that in former times a similar 

 vegetation covered the whole area of sand, and that man is the 

 agent of destruction to whose devastations the naked and shifting 

 sand-desert is due. "As soon as we get away from roads 

 and human habitations and reach more solitary places, we 

 always find that the sand-surface becomes more closely covered 

 with vegetation of trees or shrubs. These trees arrest the sand, 

 not through their sand-binding roots, but through their size 

 which modifies the force of the wind and screens the soil from its 

 attacks. Under their protection a richer herbaceous vegetation 

 is also developed." This development from stable to unstable 

 desert through the agency of man must be termed a biotic 

 succession (Cowles). This and the topographic succession 

 mentioned above do not exclude each other. 



Even if we allow that Korshinsky is correct, that man 

 through felling trees for firewood and through his herds of 

 cattle roaming about, has in many places laid bare the arrest- 

 ed sand and thus brought about the appearance of the naked 

 desert, — it is still probable that the Desert-plains and the 

 Hummock-desert have originated from the shifting-desert. In 

 what other way could the hilly or undulating sand-desert 

 originate? With Obrutschew and Radde I consider it most 

 natural to assume that the normal and natural development 

 (the regional succession, (Cowles) not the topographic or biotic 

 one) has taken place from Barchans to Hummock-desert and 

 from the latter to Desert-plains. It is not correct as Korshinsky 

 states that the stable desert is always found far from human 

 habitations, for in such places I have seen wild sand-desert 

 almost devoid of vegetation. 



