60 BULLETIN 1059, U. S. DEPARTMENT OE AGRICULTURE. 
PRECIPITATION*. 
Precipitation should be measured at as many special stations as 
possible, but only at those which are fairly permanent. In general, 
the regular Weather Bureau data collected at a large number of sta- 
tions will suffice for the purposes of forest investigators. Because 
of the difficulty of obtaining an average exposure under canopies, 
precipitation should always be measured in more or less open situa- 
tions, or above the crowns, except of course when it is desired to de- 
termine the amount intercepted by crowns (89). 
Since precipitation has no important action on plants until it is 
added to the moisture of the soil, there can be no object, in a biologi- 
cal study, and especially in a study of forests, in analyzing precipi- 
tation data very closely. For this reason there is no need of hourly 
precipitation records except possibly in a few localities to study the 
general character of the storms, which, of course, will vary only 
slightly with the forest types. For this purpose the tipping-bucket 
rain gauge (93) should be used. Standard eight-inch rain gauges 
(93). if properly exposed, will serve in most cases, though more 
valuable results will be secured where it is possible to install shielded 
gauges. On the whole, however, the gain in catch through the use 
of the Marvin shielded gauge 9 is hardly of enough significance to 
justify the additional expense of the installation, at least for any 
practical benefit to ecology. The methods of measuring precipitation 
are too well known to need description. 
Under certain circumstances, as in situations which can not be 
conveniently visited every day. it is possible to increase considerably 
the value of the record by keeping some kerosene in the rain gauge, 
which will cover the water and in large measure prevent its loss by 
evaporation. In this event it will be desirable either to pour off the 
kerosene before attempting to measure the water or to pour both 
into a glass graduate in which the amount of water can be seen in 
a few moments, after which as much of the kerosene as possible may 
be replaced in the gauge. This method needs little modification for 
the winter period, if the snow is melted before measuring, as ordi- 
narily it would be. It is, however, very desirable to have the snow, 
as it melts naturally, drop into a seamless basin containing some 
kerosene. This may be accomplished by placing a loose funnel near 
the bottom of the gauge. 
Exposure of Gauges. 
While the measurement of precipitation in gauges is very simple, 
the securing of a true ;c catch " is much more difficult, and for this 
reason the greatest care should be used to install gauges in such 
9 Designed by the present Chief of the United States Weather Bureau. 
