PLANTS IN ARID REGIONS—SPALDING. A457 
to the plant itself. The advantage of a thorough knowledge of soils 
is too obvious to call for comment, but it must be remembered that we 
are as yet only at the threshold of a greater and more promising work; 
namely, the investigation of the physiological requirements and capa- 
bilities of plants that can grow in a true desert habitat as compared 
with those that can not. In such comparative study lies, as it seems, 
the hope of real progress. It is impracticable for any investigator 
at the present time to mark out a straight path for others to pursue, 
and it would very properly be regarded as an impertinence were he 
to attempt this; yet there are certain obvious suggestions that may be 
offered. 
In the first place, important results have already followed the 
simplest experiments and observations when these have been con- 
ducted with exactness and with a definite end in view. To refer to 
a specific case, Professor Thornber, of the University of Arizona, 
undertook a few years ago to compare the habits of certain desert 
plants as regards germination. It was found that while the seeds of 
some species germinated at a given temperature, others could not be 
made to do so until they had been subjected to temperatures approach- 
ing the freezing point. These latter were seeds of winter annuals, 
and by this method a fundamental physiological difference between 
them and the summer annuals was established. Doubtless an in- 
definite amount of instructive and necessary work remains to be done 
in this direction, but the key to the situation was found in carrying 
out the simple experiments described. Again, partly as a relief from 
severer work, Doctor Cannon undertook, in the midst of his investi- 
gations at the Desert Laboratory, to map the distribution in the soil 
of the roots of some of the plants growing in the vicinity. Hardly 
was the work well in hand, and the root topography of less than half 
a dozen species mapped, when it was found that the clue to certain 
facts of distribution, blindly observed up to that time, had been dis- 
covered. I have spoken of this in more detail in another connection. 
Obviously it is indispensable that determination of physiological 
data and of those belonging to the physical environment should pro- 
ceed step by step together; and nowhere is this more strikingly true 
than in the investigation of soil relations. To refer to one more case 
of recent experience, within the past year Doctor Livingston has 
determined the percentage of soil moisture present in soils obtained 
from each of the topographic areas of the Desert Laboratory domain 
and the adjacent flood plain of the Santa Cruz River. His studies 
were conducted independently, though naturally not in ignorance of 
ecological studies which were being carried out at the same time on 
the same ground. It now appears that a well-nigh perfect cor- 
respondence exists between the two sets of facts obtained by inde- 
pendent workers, so perfect, in truth, that a causal relation offers 
