152 C. F. TOLMAN 
effect of the swing from one kind of climate to another received more 
than introductory treatment at the hands of Barrell. 
Climate is the product of complex causes, a few of which may be 
considered the critical factors, and the rest quite subordinate. Bar- 
rell’s classification of climate into warm and cold humid, and warm 
and cold arid, or again into constantly rainy, intermittently rainy, 
subarid, and arid, group together certain more or less decided expres- 
sions of climate, the fundamental factors of which may vary consider- 
ably, and, moreover, under the present practice of compiling weather 
data, these differences, some of major importance, could not be 
detected. The present unsatisfactory state of climatic analysis may 
be realized by comparing the conclusions of Barrell and Huntington 
in regard to the climate oscillations in regions outside the ice-sheets 
during the glacial epochs. Barrell suggests that in certain regions 
increasing cold may increase rock disintegration, without a corre- 
sponding increase in the transporting power of the streams, and there- 
fore the outwash deposits may have been built up steeper as a response 
to the increasing cold of the glacia! period, and on swing to a warmer 
climate there was erosion." 
Huntington explains that in the arid regions of Asia the moist 
(glacial) epoch accelerated the weathering of rock and growth of 
vegetation. Therefore waste was stored in the upper valleys of the 
mountains. Aridity followed with torrential precipitation and the 
destruction of vegetation, therefore the stored material was washed 
down on to the bajadas, and later, the supply from above giving out, 
these outwash deposits were dissected. ‘Therefore each swing from 
glacial to interglacial is represented by a terrace.’ 
In reaching these different results in regard to the glacial-inter- 
glacial swing in elevated subarid regions, different sets of critical 
factors are assumed. It is not unlikely that both assumptions are 
correct and that the variations in different factors occurred at different 
times. It surely therefore is profitable to find out if possible what 
factors are of quantitative importance, and what range of play is 
‘possible between them. 
t Op. cit., pp. 172, 173. Barrell applies this to the dissection of the Gila Conglom- 
erate. The author has reached a different conclusion which he hopes to present in a 
later contribution. 
2 Op. cit., pp. 357, 358. 
