365] B. E. Livingston Ley 
and one black, when operated together, make up the radio- 
atmometer, for use in studies of radiation as an atmometric 
condition. 
Whatever type of evaporating surface is employed, this sur- 
face must be clearly defined, so that the data obtained will not, 
by any chance, be regarded as comparable with other data 
derived from another type of surface. This means that the 
essentials of the instrument must be described, but this can 
be accomplished by merely naming the instrument and refer- 
ring to some previous description. Thus, it may be stated 
that a given set of data were obtained by means of the U: 8. 
Weather Bureau pan, the Briggs and Shantz shallow pan, 
the Livingston standard sphere, ete. If a new type of in- 
strument has to be used it requires a complete description. 
In stating the amount of water lost from the given instru- 
ments during a unit of time, it is of course unimportant what 
water unit is employed, so long as it is definite enough for the 
work in hand. Since the whole aim of atmometry is to meas- 
ure a power to do work, and since the amount of liquid water 
vaporized per unit of time is considered as a measure of this 
power, weight units rather than volume units should be used. 
Nevertheless, if the temperature does not vary too much, from 
reading to reading, and generally if there is no need for ex- 
treme precision, volume units may be used, and we may con- 
sider that a cubic centimeter of water weighs a gram. 
Obviously, the volume or weight of water lost from a certain 
type of instrument for a unit of time may always be multiplied 
by any value that the worker may like, so long as this value 
is stated, and so long as it is always applied to all readings 
from this same type of instrument. This treatment does not 
alter the relative values of a series of comparable readings and 
the results remain comparable. This principle makes it logi- 
cal to use depth units instead of volume units, for free water 
surfaces, for the depth of water lost from a given cylindrical 
pan is the volume lost, multiplied by the reciprocal of the sur- 
face area, this coefficient being a constant for the instrument. 
