PROCEEDINGS. 
521 
325th Meeting. November 24, 1888. 
The President in the chair. 
Thirty members and guests present. 
Messrs. Hazen and Abbe presented communications upon the 
Influence of Forests upon Rainfall. 
The following is an abstract of Mr. Hazen’s communication : 
[Abstract.] 
This question, discussed at the last meeting of the Society, is of vast 
economic importance and should be thoroughly investigated. The crucial 
test would be to take an extended forest region and carefully determine 
the difference in precipitation within and without the forest. Such an 
investigation as presented to the Society shows a tendency to an increase 
of about ten per cent, in the forest. It is very evident that fogs tend to 
linger much longer over a forest than a plain, and this increase of moist¬ 
ure must give more rain. The forest cannot attract rain, but it often 
prevents the evaporation of moisture and a desiccation of the air, such as 
takes place over a desert, and in consequence the rain-drops when formed 
are not dried up as they descend. 
An attempt has been made to determine this influence by a comparison 
of total annual rainfall at a large number of stations, on the supposition 
that the forest has gradually increased, and hence there should be an 
increase of rain in the last half of a series of years, even if these years 
did not embrace the same period. At Augusta, Illinois, for example, the 
records extended from 1843 to 1860, while at Davenport, Iowa, they were 
from 1861 to 1885, and so on. If forests have increased steadily from 
1843 on, and if no other influences have affected the rain, this discussion 
might be accepted as showing a purely negative effect from forests. We 
find, however, that there is a regular secular variation, which is far 
greater than the influence of the forest. Suppose a minimum epoch 
about the year 1860; then by the two series above we would have had a 
decided decrease by the first and just as decided increase from the sec¬ 
ond, neither of which, however, would be due to the forest. Taking the 
fluctuations at a station (St. Louis) having a long series of years, it was 
possible to predict, from the intervals of time at each of the twenty and 
more stations discussed by Mr. Gannett, which station would give an 
increase and which the reverse. 
It should be noted that vegetation and a deciduous forest can, in gen¬ 
eral, influence rain only during the growing season, and if the other 
seasons are taken they will serve to mask the effect sought. We have 
several stations which have a record of about forty-eight years. Taking 
the observations during May, June, July, August, and September, a table 
67-Bull. Phil. Soc. s Wash., Vol. 11. 
