516 DISCUSSION : FORESTS, RESERVOIRS, AND STREAM FLOW 
Mr. Chittet as a result of forest cutting, the writer maintains that these effects are 
generally confounded with those of cultivation, and evils are charged 
to the removal of trees for which it is in no wise responsible. He will 
endeavor to show. later that the case of China is no exception. 
Dr. Smith makes what seems to the writer an essentially sound 
statement of the function of the forest mulch when he says that it 
“acts as a temporary reservoir of limited capacity, but it contributes 
its stored waters downward as well as outward.” This is the writer’s 
position exactly, although he may differ from Dr. Smith as to the re- 
lative quantities that go outward and downward under varying con- 
ditions. The writer, however, dissents in the general case from the 
further view that cultivated land does not store and transmit pre- 
cipitation to the ground as effectively as does a forest soil otherwise 
similarly conditioned. 
Dr. Smith’s reference to his observations on Table Mountain should 
be read in the light of his frequent encounter of drifts. It is pre- 
cisely that formation, whether in or out of the forest, that prolongs run- 
off far into the summer. To make his experience applicable to the 
writer’s argument, he should have mentioned what he found in those 
areas of the forest, by far the greater portion, where drifts did not 
take place. Mr. Collingwood’s remark: “In the woods * * * near 
Elmira, N. Y., the drifts remained unmelted long after every bit of 
snow had disappeared from the cultivated portions of the valley,” is 
subject to the same criticism. 
Dr. Smith and Mr. Willis both take exception to the writer’s 
citation of White River, Washington, as an example of a torrential 
stream coming from a perfectly protected water-shed. Both attribute 
its character in this respect to the glacier at its head, Mr. Willis de- 
ploring the writer’s “unfortunate” lack of acquaintance “with the real 
character of the river.” Concerning this last matter, it may be said 
that after the great flood of 1906 the writer was engaged by King and 
Pierce Counties, Washington, to investigate the flood problem of the 
Cedar, Green, White, and Puyallup Rivers, and made a special study of 
the situation on White River. He states most positively that the 
glaciers are an entirely negligible quantity in producing the floods of 
that stream. 
In the first place, at the season of the year considered, precipita- 
tion on the glacier is almost exclusively snow, a part of the aceumula- 
tion that goes on every winter. It is not believed that there is any 
large amount of surface melting at that season, and what there is tends 
more to compact the body of snow beneath it than to run off into the 
streams. It is in midsummer that the flow from the glaciers is greatly 
accentuated, Jt is then that it bears the milky appearance from 
