TORRENTS IN FRANCE. 947 
described by very competent and trustworthy observers, whose 
reports are more easily accessible than those published in other 
countries.* 
The provinces of Dauphiny and Provence comprise a territory 
of fourteen or fifteen thousand square miles, bounded north-west 
by the Isere, north-east and east by the Alps, south by the Medi- 
terranean, west by the Rhone, and extending from 42° to about 
45° of north latitude. The surface is generally hilly and even 
mountainous, and several of the peaks in Dauphiny rise above 
the limit of perpetual snow. Except upon the mountain ridges, 
the climate, as compared with that of the United States in the 
same latitude, is extremely mild. Little snow falls, except 
upon the higher mountains, the frosts are light, and the sum- 
mers long, as might, indeed, be inferred from the vegetation ; 
for in the cultivated districts, the vine and the fig everywhere 
flourish ; the olive thrives as far north as 433°, and upon the 
coast grow the orange, the lemon, and the date-palm. The forest 
trees, too, are of southern type, umbrella pines, various species 
of evergreen oaks, and many other trees and shrubs of per- 
sistent broad-leaved foliage, characterizing the landscape. 
The rapid slope of the mountains naturally exposed these 
provinces to damage by torrents, and the Romans diminished 
their injurious effects by erecting, in the beds of ravines, bar- 
riers of rocks loosely piled up, which permitted a slow escape 
of the water, but compelled it to deposit above the dikes the 
earth and gravel with which it was charged.t At a later 
* Streflleur (Ueber die Natur und die Wirkungen der Wildbdche, p. 38) 
maintains that all the observations and speculations of French authors on the 
nature of torrents had been anticipated by Austrian writers. In proof of this 
assertion he refers to the works of Franz von Zallinger, 1778, Von Arretin, 
1808, Franz Duile, 1826, all published at Innsbruck, and HAGEN’s Beschret- 
bung neuerer Wasserbauwerke, Konigsberg, 1826, none of which works are 
known to me. It is evident, however, that the conclusions of Surell and 
other French writers whom I cite, are original results of personal investiga- 
tion, and not borrowed opinions. 
+ Whether Palissy was acquainted with this ancient practice, or whether it 
was one of those original suggestions of which his works are so full, I know 
