A NEW METHOD OF ESTIMATING STREAM-FLOW L57 
A —A 
current day. From the nature of the factor - — l it will be seen that it decreases 
.1 
rapidly below unity at first, then more and more slowly, which corresponds with 
the rapid falling off at first in the yield of water from melting, and the gradual 
decrease in that yield as the snow disappears. 
FIXING BEGINNING AND END OF THE PERIOD OF DECREASE IN RATE OF MELTING 
It is highly important and relatively easy to fix the date on which the net 
melting begins to decrease in rate. It is relatively difficult, but fortunately unim- 
portant, to fix the last date on which any addition to storage results from melting. 
On the Wagon Wheel Gap areas the date of the beginning of the period of decreased 
rate of melting could be estimated in four ways, as follows : 
(a) by a study of the diurnal fluctuations in the stream-flow in conjunction 
with the air temperature; 
(6) by a study of the record of the disappearance of the snow at the snow gages; 
(c) by a preliminary substitution in observation equations; 
(d) from the derived flood-flow formula. 
The mean of one or more of the dates fixed by these methods was taken as the 
most probable date of beginning of disappearance of snow on Stream A. On 
Stream B the last-named method only was used. These methods will be briefly 
discussed in turn. 
The observations of stream discharge were made hourly. Each discharge 
used in the observation equations was the mean of 24 hourly discharges. The 
stream-flow in the freezing-melting period during the daylight hours is greater than 
during the night hours, corresponding to fluctuations in the temperature as 
recorded on the thermographs. As the freezing-melting period approached its 
end, the diurnal fluctuations in stream-flow increase in amplitude, corresponding to 
a more or less parallel increase in amplitude in the temperature curve fluctuations. 
At the time when the melting begins to decrease in rate, the rough parallelism 
between the two curves disappears; that is, the temperature curve continues its 
diurnal fluctuations, but the fluctuations in the stream-flow curve begin to 
decrease in amplitude. At some point, the stream-flow continues to decrease 
both in the amplitude of its diurnal fluctuations and in absolute amount. This 
was normally taken as the beginning of the period of decreased rate of melting. 
This method was not always applicable. For instance, in 1913 on Stream A, it 
was impossible to estimate the beginning of the period of decreased rate of melting 
by examination of the diurnal fluctuations in stream-flow and temperature, as 
outlined above. 
On the Wagon Wheel Gap areas, snow gagings were taken normally every 
three days. The record of mean water equivalent of the snow on the ground, and 
the number of gages at which snow appeared was available at the intervals stated. 
It was fairly easy from a study of such a record to decide when the melting began to 
diminish in rate. 
The method of estimating the date of beginning of decrease in rate of melting 
by a preliminary substitution in observation equations is as follows : With approxi- 
mate values of S c , R' h R' 2) . . . R' n , — and C, F, M and T" from previous solu- 
E w 
tions, the normal stream-flow is computed from equation (33), involving equations 
