42 DISTRIBUTION AND MOVEMENTS OF DESERT PLANTS. 
slowly from old positions and advance into new ones, as apparently they 
have done from an indefinite past. If this view is correct, the study of 
present local distribution and movements is the surest guide to an under- 
standing of those of earlier times. 
Observations carried on in more widely separated regions, particularly 
in the desert near Alamogordo, New Mexico, and along the western limits 
of the Salton Basin, have confirmed the results that have been indicated 
as regards the importance of physiographic features and soil characters 
in determining local distribution of plants. As to the latter factor, how- 
ever, perhaps no more striking case can be found than that reported 
by Mr. J. C. Blumer in his recent studies of the ecology of the Fort Bayard 
(New Mexico) watershed,' in which, within a comparatively limited area, 
no less than seven distinct societies of plants, each referable to its special 
rock-formation, are clearly recognized and described. ‘Thus, giving them 
the name of the rock on which they occur, there are: (1) limestone, (2) 
quartzite, (3) granite, (4) andesite, (5) rhyolite, (6) conglomerate, and 
(7) basalt societies. Very remarkable differences of composition are 
noted in societies close together, or even adjoining. Of two adjacent 
areas, each including 0.05 acre, meeting on the contour line that marked 
the contact between basalt and conglomerate, that on the basaltic slope 
showed 41 individuals of Agave applanata and one of Quercus arizonica, 
while that on the conglomerate had 36 of Quercus artzonica and none at 
all of the Agave. 
Aspect in the same region is recognized as a powerful, and in fact the 
most powerful factor; and the analysis of the flora, based on this, presents 
very striking evidence as to its importance as a factor in local distribution. 
As an example, ‘the east side of Cameron Canyon changes repeatedly 
from west to northwest aspect, and on the latter not more than 30° in 
aspect is sufficient to give the entire slope, from top to bottom, over to 
bull pine.’’ Similar cases have been observed everywhere, but they 
are particularly well marked in the semi-arid regions of the southwestern 
United States. An extremely interesting locality in New Mexico was 
visited by the writer in July, 1906, where at an altitude of 6,000 feet 
the flora of the pifion association occupies the left bank of a narrow can- 
yon, while on the right bank, hardly a stone’s throw distant, are the 
plants of the desert below—mesquite, creosote-bush, ocotillo, Yucca, 
Agave, and various cacti. 
On the Laboratory domain, the floristic differences presented by neigh- 
boring areas, due to differences of aspect, though generally less striking 
than those just referred to, are sufficiently marked to afford a favorable 
opportunity for an investigation of the differences of environment to 

1Unpublished notes communicated to the writer. 
