373 | B, H. Livingston 175 
day to day, for the same place, which makes it necessary in 
climatic discussions to resort to averages and means. While 
the index of relative humidity for any instant may be readily 
interpreted by use of the corresponding air temperature, there 
is no possible way by which an average of several such indices 
may be so interpreted; the average temperature for the period 
is of no use for this purpose, since the march of temperature 
for the period is not necessarily at all related to that of the 
moisture condition. The only way to give definite meaning 
to a relative humidity mean is to obtain the original humidity 
values from which the mean was derived (together with the 
corresponding air temperatures), to substitute for each indi- 
vidual value the corresponding vapor pressure deficit, and to 
derive the mean of the deficits, thus discarding relative hu- 
midity altogether. 
For biological experimentation, for hygienic studies of the 
air moisture condition in dwellings, and for general climato- 
logical purposes, it is very obvious that the whole concept of 
relative humidity is hopelessly misleading; the sooner this 
concept can be forgotten the more rapidly will knowledge ad- 
vance. When it is not desirable or expedient to employ the 
index of atmospheric evaporating power itself (as determined 
directly by some form of atmometer), the moisture condition 
of the air should be stated in terms of the vapor pressure de- 
ficit, which demands no correction for air temperature and 
may represent evaporating power in all comparisons where the 
index of effective air circulation may be considered as constant. 
