i86 



FORMATIONS AND GUILDS 



[Part I 



a quantity of water and insects that the occurrence of this luxurian 

 and not remarkably xerophilous plant on such a soil did not appea 

 wonderful. 



As the result of their investigations on the sandy and loamy plain of th 

 Camargue, which is 35,000 acres in extent and lies in the Rhone delt; 

 Flahault and Combres have described the gradual conversion of the bar 

 soil within reach of storm-tides first into open, and later into close 

 formations. They show that if a f^at shoreland tract is withdrawn fc 



vi 



•^ 



Fig. 100. Earliest vegetation on a new volcanic soil (pumice, ashes, &c.) in West Java. 



From a pliotograph. 1 



a long time from the influence of the waves, the earliest vegetation 

 produces is composed of tufts of Salicornia macrostachya growing widr 

 apart (Fig. loi). A shoreland thus colonized is frequently flooded ' 

 winter storms and again deprived of all vegetation ; occasionally, howev. 

 the first .settlers become able to maintain themselves and collect amoi; 

 and on their bushy branches a quantity of sand, small indeed, tt 

 sufficient to render possible the appearance of some new plants, such ' 

 Salicornia sarmentosa, Atriplex portulacoides, and Dactylis sarmentc 



