290 The Plant World. 



found a relatively dense vegetation in which occur examples of 

 nearly all, if not all, 3f the types which I have mentioned, grow- 

 ing intermingled under identical environmental conditions. 

 The larger differences of soil texture and topographic site deter- 

 mine the flora of a given hectre of ground in such situations as 

 these, but the relative abundance of this or that type or species 

 of plant is not, I wish to maintain, so much a matter of physical 

 conditions as they affect the adult plants, but rathei an outcome 

 of the vicissitudes of germination and establishment for the 

 species of the flora of that area. After the larger features of the 

 habitat have determined its flora, I am convinced that the ac- 

 tual make up of the vegetation, the relative abundance of the 

 different types or species, and even, to a large extent, the density 

 of the stand itself are products of the conditions which con- 

 trol germination and the activities of the seedling during its 

 first twelve months. This is indicated by the extreme irregu- 

 larity of admixture of vegetation types in small areas. The work 

 of Cannon f has shown that great diversity of root habits exists 

 among plants of the different types, which may have much to 

 do with enabling a mixed population to survive on an area 

 where a uniform population of the same number of individuals 

 could not do so, owing to the fact that the different types secure 

 their supplies of soil water either at different levels of the soil cr 

 at different seasons of the year. Serving merely as an extreme 

 case of the irregularity of the allotment of desert plants are the 

 not infrequent dense colonies of the same species, near which, 

 and under identical conditions, may be colonies of other species. 

 These cases are nearly conflned to the cacti, however, which ob- 

 tain their annual store of water during the summer rains and 

 do not therefore invalidate the explanation of the usual diver- 

 sity of stand which is afforded by the work of Cannon. These 

 pure stands must also have their explanation in the chances of 

 seeding, geimination and establishment (or of vegetative multi- 

 plication) rather than in any physical differences of habitat. 



I am still further convinced that the chances of geimina- 

 tion have a very important part in determining the character 

 and density of desert vegetation, by the extreme slowness with 



tCannon, W. A. — The Root Habits of Desert Plants. Pub. Cam Inst.. No. 131 1911. 



