1925] Storer: A Synopsis of the Amphibia of California 213 
local colonies, of this amphibian, although a large series of specimens 
from appropriate localities might be found to exhibit geographic 
variation. The fact that no differences have been pointed out by 
systematic herpetologists suggests that the species must be quite uni- 
form over its vast range, or at least that it presents no variations 
which can be correlated with geography. This would argue that its 
present ‘patchy’ mode of distribution is an event of rather recent 
time, geologically speaking, or else that the species is very conserva- 
tive as to change. 
Geologists believe that the mountain ranges lying east of the Los 
Angeles plains are of very recent origin, that a relatively short 
time ago these ranges were much lower or lacking. When such con- 
ditions obtained there were probably streams, some of which were 
perennial, on a number of the mountain ranges in the deserts of south- 
eastern California where no such creeks exist today. A species of 
creek-dwelling amphibian could pass across a series of such streams 
in adjacent mountain ranges in a comparatively short span of time by 
means of the connections set up when sudden heavy rains occur, 
whereas present conditions would seem to be entirely against such 
dispersal. The desert tongues of the streams which come clown out 
of the southern California mountains each season change markedly 
in extent. In wet years they are long, in dry years short, and there 
is a progressive seasonal shortening in any one year through spring, 
summer, and autumn, until the advent of the next rainy period. The 
Canon Tree-toads evidently move down and back with the fluctuations 
in the stream, as observers who have visited Tahquitz Creek in the 
later seasons of the year have found it dry in parts where hylas 
are present in numbers in the spring months. Pressure from compe- 
tition for food, or the natural tendency of a species to spread out and 
occupy all regions which it can tolerate probably accounts for this 
spreading even within the compass of a single year. In the event of a 
cloudburst in a semidesert region the streams are swollen to many 
times their normal volume. Animals like arenicolor would be swept 
down and away from their native canons onto the common grounds 
below. As the w T ater receded they would tend to follow it backward, 
but not all would necessarily get back to their own streams (this 
despite the strong homing instinct which is argued by some persons 
for at least certain of the Salientia) . They would follow the retreat- 
ing creeks and new creeks would thus be populated. Given the numer- 
ous occurrences of this sort which would happen over a brief period 
