262 FORESTS, RESERVOIRS, AND STREAM FLOW 
During the spring of 1908 occurred a record-breaking flood in 
Western Montana, nearly all the streams on both sides of the Con- 
tinental Divide going far over their banks. As might have been 
predicted; this occurrence was promptly cited as another example of 
the effect that a forest-barren country has upon floods. Nevertheless 
it is as certain as anything of this.kind can be, that if the country 
affected by this extraordinary downpour (in some places breaking all 
previous records) had been thickly forested, and the ground still 
covered, as it would have been, with a solid layer of saturated snow, 
the flood would have far exceeded in magnitude and destructiveness 
that which actually took place. Wherever forests existed in the 
higher altitudes they did have this effect.* 
Having now considered the influence of forests upon stream flow 
from a theoretical standpoint, let the records themselves be examined 
as far as they are available. These records in the United States, 
unfortunately, are not so useful as might be wished, because of their 
brevity. No continuous records on any of our streams run back for 
more than eighty years, and most of them less than half as far. This 
is far short of the 200 years considered by certain European engineers 
who investigated Wex’s theories as the minimum period “necessary 
in order to draw a reliable conclusion” upon this subject.” It does 
indeed seem absurd to take present-day records, as is constantly done, 
and draw conclusions one way or the other as to comparisons with 
the past, of which records are entirely wanting; but such as they are, 
a few of these records are given in Table 1. They include in most 
eases both high and low water, although the low-water records cannot, 
in the nature of the case, be of very much value. Works of channel 
improvement on most of the streams have probably affected some- 
what the low-water stages for the same discharge, while, as is well 
known, a given stage, even in a natural stream, does not mean the 
same discharge at different times.t It is really the discharge of the 
streams rather than the stage that forms the correct basis for com- 
parison; but data for discharge are almost wholly wanting. 
An examination of these records shows how utterly impossible it 
is to find anything in them to support the current theory of forest 
*In the Weather Bureau report, Montana section for June, 1908, it is stated 
that “the rainfall was phenomenally heavy over most of this district, and, combined 
with the water from the rapidly melting snow in the high mountains, caused unpre- 
cedented floods in nearly all streams.” 
+During the past 20 years the low-water stage of the Mississippi 
has been materially modified by reservoir action. eae ene 
