404 DISCUSSION: FORESTS, RESERVOIRS, AND STREAM FLOW 
Mr. Leighton, not appear like a cessation of timber-cutting. On the contrary, 
Turners Falls is well down on the river, and, if the observations of 
lumbermen be true, and if the statistics of the Forest Service be 
justified, this amount of timber does not represent the extent of actual 
deforestation. This quantity of timber, from 50000000 to 80000000 
ft. b. m., represents an enormous and constant drain on the primeval 
forest on the Upper Connecticut. So much for the facts. The line 
of inquiry suggested in Colonel Chittenden’s foot-note will now be 
developed. 
The diagrams, Figs. 8 to 15, inclusive, have been prepared to 
show: 
First, the record of rainfall in each place for the period of observa- 
tion; 
Second, the number of days in each year during which the gauge 
stood at or above a certain designated poirt which may be considered 
as the flood point. (The point chosen in these cases may or may not 
be the true flood point; but the fact is immaterial. The presentation 
of data in this way is, in its essential result, merely the cutting off 
of the top of the hydrograph and the summing up of the number of 
days. As long as the same point is used throughout the record on each 
river, the exact location thereof is of little consequence, because it is 
the relative flood duration, rather than an actual numerical record of 
total height, that is desired) ; 
Third, the expression of the number of days in terms of precipita- 
tion. Each diagram is the expression of precipitation as a straight 
line, and the further expression of the number of days of flood in 
terms thereof. It will be observed that such a record obviates the 
necessity for a two-century record of observation, because it gives the 
actual results of a unit’s precipitation. 
The fourth diagram on each sheet is the expression of the total 
rainfall in each 10- or 5-year period, such periods being progressive. 
Thus, the first period may be 1880-89, inclusive, the second 1881-90, 
inclusive, and so on. Such an expression gives the concrete view of 
progressive flood tendencies. The wide annual variations in precipita- 
tion and the resultant run-off effects thereof, by this method of ex- 
pression, are smoothed out and compensated, and afford a less involved 
idea of the actual trend of events. The fifth and sixth expressions are 
those of total days of flood and the relation of one to the other, all in 
progressive periods, as in the fourth expression. It will be appreciated 
that the last three are composed of the same original data as the first 
three, the only change being the method of assembling the facts. A 
mere glance will show that flood tendency on all these basins is 
increasing, and an examination of their history will prove that while, 
in some cases, a part of the change may be due to other factors, still 
the relative importance of those factors is insignificant in comparison 
