PRIMARY SUCCESSIONS 109 



but they are also relatively inaccessible to investigators 

 They have been studied in a few cases, for example, those 

 of Krakatoa by Treub, but this study has been confined to 

 the general features of revegetation. Ash fields and lava 

 beds are widely different in compactness, but they agree in 

 having a low water and nutrition content. The pioneer 

 plants in both will be intense xerophytes, but the soil 

 differences will determine that these shall be sandbinders 

 in the former, and rock-weathering plants in the latter. 



Practically all primary successions start on soils produced 

 by weathering : this is also true of coral or volcanic islets 

 and of lava beds, for no terrestrial vegetation can secure a 

 foothold upon them until the surface of the rock has been 

 to some extent decomposed or disintegrated. Weathering, 

 as is well known, consists of two processes, disintegration 

 and decomposition, which usually operate successively, 

 though they are sometimes concomitant. Disintegration 

 usually precedes, especially in rock masses, and, unless it 

 is soon followed by decomposition, results in dysgeogenous 

 soils. Decomposition often goes hand in hand with disinte- 

 gration, or it takes place so ratiidly and perfectly that it 

 alone seems to be present. In either case, the resulting 

 soil is eugeogenous. The relation of decomposition to dis- 

 integration determines the size and compactness of the soil 

 particles, and upon the latter depend the porosity, capillarity 

 and hygroscopicity of the soil. These control in large degree 

 the character of the first vegetation to appear on the soil. 



Another point of fundamental value in determining re- 

 vegetation is the disposition of the weathered rock. If it 

 remains in situ, it will evidently differ in respect to compact- 

 ness, homogeneity, nutrition content, water content, dissem- 

 inules, etc., from weathered material which has been trans- 

 ported. An essential difference also arises from the fact that 

 a rock may be weathered a long distance from the place 

 where the decomposed particles are finally deposited, and 

 in the midst of a vegetation very different from that found 

 in the region of deposit. The disposition of the weathered 

 material affords in consequence a satisfactory basis for the 



