162 BTLLETIX 1059, U. S. DEPARTMENT OF AGRICULTURE. 
these perforations. The disk is firmly attached to, and flush with, 
the upper end of the stem. 
Over the wick is a cover only slightly larger than the disk, whose 
flanged edge extends down over the edge of the disk. This is 
held down by two .-crew-, which engage the flanged edge. The 
cover is flat, but seamless, and completely excludes rain or snow. Its 
surface is finished with nickel and an instrument having this polished 
surface absorbs practically no radiant energy and is called a " shade " 
instrument. To obtain the effects of radiation, an instrument whose 
r has been coated dead-black is used. The "shade 53 instrument 
is now entirely abandoned, since it has been -een that the difference 
between the two is not a mea>ure of sunlight intensity, but a measure 
of the addditional effect of sunlight in producing evaporation. This 
r can not under any circumstances be ignored in ecological 
studies. 
The tests which have been made show that the lo-se- from the 
blackened instrument of this type follow more closely those from 
potted trees, under a great variety of atmospheric and solar condi- 
.-. than do the lo-ses from any other type of instrument at 
present available. The instrument has also shown itself remarkably 
free from annoying characters, and responsive to all degrees of 
evaporation stress. It may be considered, however, something of a 
disadvantage that the amount evaporated is relatively small. The 
losses for short period- may. therefore, only be determined by very 
precise weighing. 
OliSEEVATIO^S. 
With the Forest Service evaporimeter the daily observations con- 
sist in weighing the instrument, usually on an inexpensive balance 
such as the Harvard trip scale. Eefilling is undertaken as often as 
necessary to maintain between 100 and 200 cubic centimeters of water 
in the tank. The " closing "" weight and " refilled " weight are en- 
tered on the field observation form, together with the number of 
the instrument. Computations of los-es are usually made when 
entering the data on Form 8. and the correction in accordance with 
the calibration of the instrument is usually applied only to the total 
I for 10-day period-. 
With porous cups the daily or periodic observations will usually 
consist in an entry of the amount of water required to All the reser- 
voir to datum level. A graduate i> taken into the held to measure 
this amount. 
In the case of the Piche evaporimeter. losses are calculated from 
the reading- of the graduated reservoir. Note should always be 
made of observed overflowing or drying-up of the filter paper or 
evaporating surface, and the probable correction due to these fail- 
ures jf the instrument should in all cases be estimated and entered 
