367 | B. E. Livingston * 169 
ration exceeds precipitation by a certain depth, 15 cm. in this 
example. But this means nothing at all; if the pan used had 
been larger or smaller, of different shape or material, or if a 
wet soil surface had been employed, etc., the result would have 
been quite different, and the climatic conditions would surely 
not have been altered by merely changing the atmometer. On 
the other hand, if any other form or size of raingage had been 
employed the results would be sensibly the same. The amount 
of evaporation depends largely upon the atmometer but the 
amount of rainfall recorded is practically independent of the 
raingage, so long as the latter is a raingage at all. It is legiti- 
mate to state the index of rainfall in depth units, for this is 
not seriously influenced by the internal characteristics of the 
gage, a statement that cannot be made of the index of evapo- 
ration, nor even of the index of atmospheric evaporating pow- 
er. The only logical way by which atmometric and precipita- 
tion measurements may be compared is by means of their 
ratio, in which case one set of measurements may be in depth 
units and the other in volume or weight units. They are not 
commensurable in any case, so it is best not to have them even 
appear as though they were commensurable. Other considera- 
tions, into which I cannot go in this place, lead unequivocally 
to the same conclusion. 
Fortunately, there is no serious difficulty encountered in 
the statement of the time feature of atmometric measure- 
ments. For short periods the hour is most convenient, for 
longer periods the day, week and year are all suitable. Since 
months vary in length, monthly atmometric indices are un- 
satisfactory. After the three features of the unit to be used 
have been decided upon, it is necessary to remember that at- 
mometric measurements, like other power measurements, al- 
ways apply to a certain set of circumstances and to a certain 
time period. The set of circumstances here emphasized is the 
surroundings of the atmometer, they comprise the various 
features of its exposure. The readings refer to the evaporat- 
ing power of the air only for the particular location in which 
the instrument was operated. The evaporating power of the 
