INCREASE OF SHRUBS. 23 



Jong associated with the conditions or interested in this phase of 

 desert vegetation to form from memory accurate judgments. 



It is true that the grassy belt is noticeably not continuous in these 

 mountains or elsewhere, but on the other hand, the desert mesquite 

 forest often extends quite to the beginning of the real mountain 

 forest. Such a condition is easily accounted for by local influences. 

 In the canyons the growth of timber is continuous from the moun- 

 tains to the desert, and in places on the ridges and gently sloping mesas 

 as well (PI. VI, fig. 1). In both cases the writer believes this condi- 

 tion to be due mainly to the comparatively low fertility of the soil 

 in these situations, assisted in the canyons by the excess of water. 

 The bottoms of the canyons are less productive of grass here to-day 

 than the mesas between, owing to the constant shifting of the sands and 

 the washing off of the finer particles of soil by the frequently changing 

 courses of the flood waters. Action of fires was therefore less pro- 

 nounced in the canyons, allowing trees and shrubbery to get started, as 

 well as when they did get started, furnishing greater moisture in the 

 subsoil. 



The main factor, though, in the opinion of the writer, has been that 

 of fire. It is firmly believed that were it not for the influence of this 

 factor the grassy mesas would to-day be covered with brush and 

 trees, the same as the canyons, except that the growth would be 

 smaller, owing to a more limited supply of moisture. In short, the 

 same laws apply here that govern in our great prairie States (see PI. 

 VI, fig. 2), where the treeless plains were kept so by frequent fires. 

 It is a very conspicuous fact that the continuance of the desert forest, 

 up to the mountains upon the mesas, occurs where the soil is poorest; 

 in other words, upon the lands which produced least grass, and, con- 

 sequently, the smallest amount of food for fires. This fact is illus- 

 trated in one of the pastures, where there is an area of mesquite upon 

 the gently sloping mesa. But this area is naturally poor soil, and its 

 poverty, on account of location close to the terminus of a temporary 

 stream, has been aggregated for many years by excessive grazing. 

 The same remarks apply to the region just north of Helvetia. 



The spread of the seed of the mesquite by cattle and horses eating 

 the beans, thus furnishing a good culture for their development, 

 may have some influence, but it is so seldom that seedlings are met 

 with that it is questionable whether cattle or horses have as much 

 influence as has been suggested. They probably do cause a more 

 thorough and widespread distribution, but probably assist the spread 

 of shrubbery more by eating off the grass, so that there is nothing to 

 burn, than they do by distributing the seeds. 



177 



