6 



The Plant World. 



by both plants in equal numbers, and the question then may be 

 why this colonization does not go on at the same rate on both 

 slopes. 



An answer in general terms may be given, that natural 

 selection prevents this colonization at equal rates. Indeed, to 

 use a now well-worn simile, it appears to the observer on the 

 ground, that the sieve-like action of natural selection is at work 

 at the mouth of the arroyo, the sieve meshes being so adjusted 

 that only the one species may enter the one slope, while the other 

 only may gain footing on the opposite slope. The figure, how- 

 ever, is but a figure, and it remains to offer a definite explanation 

 of the condition described. 



It is obvious that the two opposing slopes are exposed to 

 widely different degrees of insolation, and this has been called 

 upon in explanation of the different vegetations there found; 

 and it has been stated or inferred that the effect is not direct, 

 but rather indirect, in the effect upon the moisture conditions. 

 But the soil moisture on both slopes is, in the case before us, 

 sufficient for both the species in mature state. It would there- 

 fore seem necessary to push the analysis one step farther and sup- 

 pose that selection is effective only on the seedlings. The same 

 suggestion has been made by Ganong * a propos of certain 

 peculiarities of distribution of trees upon the Miscou sandplains. 

 And it is now understood, since the work of Livingston {Pub. No. 

 .50, Carnegie Institution of Washington), that the seeds of desert 

 plants do not show any indifference to water supply, and obser- 

 vation has shown that periods of germination in the desert 

 are those during which moisture is abundant. It seems likely, 

 however, that germinating seedlings show considerable differ- 

 ences in the rate at which resistance to drought is acquired, 

 some becoming quickly succulent, or otherwise resistant, while 

 others show little or no capacity in this regard above that of 

 mesophytes. We can easily see, therefore, that the length of 

 the season in the desert during which mesophytic conditions 

 prevail will have a selective effect upon the vegetation, a spor- 

 adically short period serving to eliminate the kinds with a long 

 term of non-resistance for the time being, and, in deserts in 



*Bot. Gaz. 43: 341-344, 1907. 



