FORESTS, RESERVOIRS, AND STREAM FLOW 263 
intluence. ‘They prove conclusively that there has been no marked 
change since the settlement of the country began, and that such 
change as there has been is on the side of higher high waters and 
tower low waters before the forests were cut off. What the record 
would be if we could go back 200 years cannot be said, but it may 
sately be conjectured that it would show both floods and low waters 
that would equal or surpass any modern record. It is the experience 
of every engineer who has the opportunity to observe the action and 
study the history of great rivers to find everywhere evidence of the 
occurrence of higher waters than any of which he has positive record. 
The up-building of bottom lands, the survival of old water marks, 
and many other indications, show that, great as are modern floods, 
those of the past were greater still. In the very nature of the case, 
it is not possible to find similar evidence of former low waters, 
because such evidence is wiped out by every succeeding high water; 
but whoever will take the trouble to study records of early expeditions 
on our rivers, when barges, keel boats and similar craft were used, 
will conclude that extreme low water is not a modern development 
by any means. Measurements of the Monongahela River, at Browns- 
ville, in 1838 and 1856, low-water years, gave discharges of 75 and 
23 cu. ft. per sec., respectively. It is quite certain that the river has 
not fallen so low in late years. At Pittsburg in 1895 (the dryest 
season in recent years), it fell to 160 ft. 
The point should be fully recognized that these records are value- 
less for establishing either side of the forestry argument unless they 
clearly indicate a new tendency in river flow. It is not enough to 
cite a few isolated cases. In a period of, say, 200 years, there must be 
a record year for high and one for low water. Is there any reason 
why it might.not occur this year as well as earlier? There must 
be clear evidence of permanent change before any conclusion can be 
legitimately drawn. In two instances such a tendency may possibly be 
claimed, the Ohio at Pittsburg and the Connecticut at Holyoke, which 
show, in the past few years, a greater frequency of high waters than 
for some years previously.* To whatever extent this may be true, it 
*In the period of 34 years from 1874, the Ohio River at Pittsburg rose above 
15 ft. on the gauge 148 times. In the first half of this period, 68 of these freshets 
occurred and 80 in the second half. The mean for the first half was 19.3 ft. and 
20.2 ft. for the second half. The mean of the lowest waters of the first half was 0.3 
ft. and 1.6 ft. for the second half. In Transactions, Am. Soc. C. E., Vol. LVIII, p. 
31, is a 20-year volumetric record of the Connecticut, which indicates somewhat 
higher high waters during the last half of the period. But in this case, as at Pitts- 
burg, higher low waters are also indicated. In fact, in both cases, the greater run- 
off in the later period was clearly due to greater precipitation. 
