LOCAL DISTRIBUTION OF SPECIES. 47 
high in summer the differences are less marked, but perhaps are no less 
significant as regards plants growing on the north exposure which are 
sensitive to extreme desert conditions. 
It is evident that, starting with temperature as the initial factor, semi- 
tropical desert conditions are distinctly emphasized on the right side 
of the gulch, and that these conditions are greatly mitigated on the left, 
and it can not be doubted that the great differences of their plant covering 
are directly correlated with these differences of temperature and the 
changed conditions to which they have given rise. 
As already stated, observations of light intensity and readings of wet 
and dry bulb thermometers were carried on for some time; but with the 
growing conviction that with our present methods no satisfactory results 
could be looked for, and with increasing evidence that, for such a compar- 
ison of habitats, light intensity at least, is a matter of very subordinate 
importance. This part of the work is accordingly not reported here. 
MEANS AND AGENCIES OF DISPERSAL. 
The various plant species of Tumamoc Hill are provided with means 
of seed dispersal essentially the same as those found in all great conti- 
nental areas of the globe. The pappus-bearing composites, which are 
fairly numerous, are adapted to wind dispersal; many others, including 
species of Boraginaceze, Umbelliferae, etc., have spiny or hooked fruits; 
others, such as Lycium, Celtts, and the cacti, have fleshy fruits likely 
to insure dispersal through the agency of birds and other animals, and 
to these are to be added various capsular fruits, grasses with twisted 
awns, and still other mechanical contrivances of various degrees of effi- 
ciency in local dissemination. 
Of the several agencies of dispersal, the strong winds of the desert rank 
high in efficiency, and there is no doubt that light, winged seeds are 
borne on the wind to great distances. Exact observations are, in the 
nature of the case, difficult to obtain, but the evidence is ample. Taking 
25 or 30 miles as a measure, far within the limits of safety, of the possible 
annual advance of a species with light seeds adapted to wind dispersal, 
it is certain that in a few years such species may have migrated hundreds 
of miles from their point of departure. Indeed, it can not be doubted 
that their geographical limits are imposed by environmental conditions 
rather than by means of dispersal. Given the time element, there is no 
-part of western America, in which suitable conditions prevail, which 
such species may not reach. Their present actual range may, therefore, 
be looked upon as a resultant of a number of different factors, the deter- 
mination of which presents a most intricate biological problem, involving 
determination of capacity for dissemination, often practically unlimited 
except by natural barriers, and the controlling influence of the most 
diverse natural environments. 
