368 J.D. Dana—Flood of the Connecticut River Valley. 
at the two rates of flow in the river, about 359 and 478° cubic 
miles.* 
The amount from the precipitation to be deducted does not 
diminish greatly these numbers, even if all of it went to in- 
crease the floods. With a mean annual rain-fall of 120 inches, 
or 10 feet, this amount for the part of the entire drainage-area 
north of Massachusetts, would be only 16:1 cubic miles; anc 
for the whole drainage-area to the Sound, 21°8 cubic miles. 
This amount, 21°8 cubic miles, is too large by the amount 
which would have been lost through evaporation, soil-absorp- 
tion, chemical changes in and beneath the glacier making 
oxides and hydrates, and in other ways. An estimate of these 
losses may be derived from the modern Connecticut. 
The annual discharge of the Connecticut, as determined by 
measurements made under the direction of General G. K. War- 
ren, of the U.S. Engineer Corps, by General T. G. Ellis,t at 
Hartford, Ct., where low water in the stream is almost exactly 
mean-tide level, it being only ‘045 feet below it, was as follows 
for the years mentioned : 
Year. Discharge i bie feet Year. Discharge in cubic feet. 
1872 638,070,000,000 1875 570,649,000,000 
1873 727.160,000,000 1876 706,291,000,000 
1874 733,103,000,000 1877 526,261,000,000 
The largest of these sums, that for 1874, corresponds nearly 
to 5 cubic miles (more exactly 4°95). 
The amount of annual precipitation over the drainage-area 
north of Hartford to the sources of the Connecticut (allowing 42 
inches as the mean for the part north of Massachusetts and 46 
inches for the rest north of Hartford), averages nearly Cli 
cubic miles. Consequently the amount discharged at Hart- 
ford in 1874, was 69°5 percent of the precipitation ; so that 
the loss in all other ways was little over 30 per cent. In 1877, 
a year of minimum discharge, the amount was 50 per cent. 
The proportion of 70 per cent more correctly represents the 
condition in the era of the glacier-melting than that of 50; for 
it is the proportion during a year of great Connecticut floods ; 
and these floods are dependent, in nearly all cases, (1) on the large 
amount of snow and ice of the winter or spring season, and to 
a considerable extent (2) on the frozen state of the ground over 
the hill slopes and much of the country, favoring an easy 
slipping of the waters over the surface without absorption by 
the soil,—conditions eminently characteristic of the era of the 
* The specific gravity is here taken too high because of crevasses and want of 
ier-ice; but this error. for the calculation of which no specific 
£ impo} . Helland found the specific gravity of the 
ice of a large iceberg only 0°886, this low rate being due to linear air bubbles 
which thoroughly permeated it. : 
+ Annual Report of the Chief of Engineers for 1878. Appendix B 14. 
4 
