114 
A NEW METHOD OF ESTIMATING STREAM-FLOW 
(6) The upward slope to the maximum of the run-off curve, plotted with 
percentages as ordinates and days as abscissas, is probably steeper than the down- 
ward slope. 
(c) The curve probably becomes concave upward soon after the maximum. 
On the basis of these considerations, the means for Streams A and B were modified 
as shown in the last two columns of the following tabulation: 
Value of — 
(1) 
Means from 
Streams A 
and B as % 
(2) 
Factor to 
reduce to 
absolute 
value 
(3) 
Absolute 
values 
(D-(2) 
(4) 
Adopted absolute 
values for 
Solution B Bi 
(5) 
Adopted values 
expressed as 
percentages 
(4)X(2) 
Ri 
0.40 
.30 
.29 
.20 
.09 
.06 
100 
100 
100 
5.0 
2.5 
1.25 
0.0040 
.0030 
.0029 
.040 
.036 
.048 
0.0020 
.0040 
.0030 
.025 
.020 
.009 
0.20 
.40 
.30 
.125 
.050 
.01125 
R t 
R, 
R t 
R t 
R t 
From column (5) in the above tabulation, it may be verified that only 1.44 
per cent of the change in storage in the ground on any day, r h was assumed in 
Solution B Bi, to reach the lake (Superior) by the end of the sixteenth day after 
the change took place, if, during the last 16 days of that 17-day period, the storage 
is assumed to be constant. This solution contained 847 days and 759 observation 
equations. The following is a comparison between the assumed values of the 
constants used in computing — r,, — r,, — r 8 , . . . — r« and the values derived 
from the least-square solution : 
Value of — 
(1) 
Assumed 
(2) 
Derived 
(3) 
Difference 
(D-(2) 
(4) 
Difference -=- 
p.e. = (3)/p.e. 
Ei 
/+0.46 
1+0.56 
+0.69 
+ 1.17 
+0.0020 
+0.0040 
+0.0030 
+0.025 
+0.020 
+0.009 
-0.051 = 0.059 
+0.721= .155 
+0.653= .261 
-0.023= .062 
-0.383± .080 
+0.244= .068 
+0.662= .263 
-0.057= .148 
-0.049= .090 
/+o.5ii 
1+ .611 
- .031 
+ .517 
+ .025 
+ .387 
- .241 
- .637 
+ .077 
+ .058 
8.7' 
0.2 
2.0, 
0.4 
4.8 
3.5 
2.4 
0.5 
0.6 ; 
Mean 
= 3.6 
.Mean 
=2.0 
Ei 
E, 
Ri 
Ri 
R, 
Ri 
R b 
R, 
The assumed values of E h E h and E } need further explanation. It has 
already been shown in Table 35 that slight evidence was found in Solution U, that 
the evaporation curve might be concave upward for winds below 10 m.p.h., and 
that this evidence was used in setting up the equation to use in Solution BB 2 for 
testing out the evidence with a larger number of observations — 28 months instead 
of 6 months of observations. As a matter of fact Solution U> was succeeded by a 
6-months solution designated as Solution U t , before Solution BB\, immediate 
predecessor of BB 2 , was made. In Solution U t the form of the observation equa- 
tion was the same as equation (27), the one used in Solutions U, and 7\, except 
