18 MISC. PUBLICATION 303, U. S. DEPT. CF AGRICULTURE 
Climate, which is an expression of temperature and moisture con- 
ditions, is without question of first importance. It is, in fact, a 
summary of factors pertaining to water and to heat. Such factors 
may be further subdivided into amount and distribution of water 
or heat; or where the two interact the terms of the analysis are 
evaporation, kind of precipitation, and humidity. 
Soil, which involves both chemical and physical structure, is un- 
doubtedly of second importance. Plant distribution within areas of 
uniform climate may be modified within each area by soil factors, 
but it seems obvious that the climate plays the major role in pre- 
venting vegetation from spreading. ‘The interaction of soil and 
climate (the latter is often a contributing factor in the formation 
or modification of the former) may be of great importance. With 
sufficient rainfall, for instance, the leaching out of certain bases may 
produce an acid or neutral soil; with insufficient rainfall the soil may 
be alkaline. 
Both the preceding factors may operate provided sufficient time has 
elapsed. A newigatenned species must have time to spread before 
its range can be delimited by changes in climate or soil. If a species 
1s young, its climatic and edaphic boundaries may not yet be operat- 
ing. In its distribution, time is still a very important factor. If a 
species is old enough so that its further distribution is limited by 
factors of climate or soil, time has ceased to be an important factor 
in its present distribution. ; 
Tt is true that climatic and soil factors may be uniform or may 
change very slightly over a given area of country, as in any of Mul- 
ford’s plant-growth regions. Knowing empiricaily that a number of 
cultivated plants succeeded about equally well throughout such a 
growth region, we are reasonably safe in assuming that in general 
the range of a native species can coincide with boundaries of such 
growth regions provided it has had the necessary time to spread. 
it is Aan itent that where there is a large number of endemics and 
where great variation in characters is evident the growth-region 
boundaries do not coincide at ali well with limits of distribution, but 
where there are few endemics and variation is much less, the bounda- 
ries coincide much better. 
The correlation of Mulford’s plant-growth regions with natural 
vegetation areas has as yet not appeared in print. The present 
author has been interested in the idea for the last 5 or 6 years, has 
checked a few species with care, and a great many in a general man- 
ner. The subject cannot be dealt with adequately here, but it can 
be said that in a general way correspondence of the hnes bounding 
the range of a given woody species with the lines bounding Mulford’s 
plant-growth regions is very close. Correspondence of the growth- 
region lines with climatic and soil boundary lines is also close, where 
either the temperature, precipitation, or soil condition, or any com- 
bination of them, acts as a limiting factor in plant distribution. 
A line bounding a growth region may be considered to consist of 
series of attached shorter lines, each representing a change in soil 
condition, in average precipitation, in average temperature, or some 
other change displayed by plant distribution. On the map (fig. 1), 
the lines must be considered not as sharply delimiting one region, 
but as being drawn to separate two regions differing in one or more 
