361] B. E. Livingston 163 



the nature of the pans themselves must be exactly alike. It 

 is thus both theoretically and practically impossible to express 

 the surface factor in the atmometric index by a unit that rep- 

 resents merely extent of surface. 



Since the complex of internal conditions that make up the 

 capacity of any surface to produce evaporation is very difficult 

 of analysis, we may avoid the necessity of this analysis by 

 simply using atmometer surfaces that act alike. Then the 

 surface factor of our unit of measurement becomes the surface 

 of our instrument (with whatever characteristics it may have), 

 and we do not need to enquire what may be its area, etc. In all 

 the studies so far carried out with porous clay and paper sur- 

 faces for measuring the evaporating power of the air, I have 

 never been led to determine the area of the surface employed ; 

 it would have been useless to do so, although such a surface 

 is easily measured. We are thus led to the proposition that 

 the atmometric index is to be expressed in terms of (1) a 

 weight unit of water, (2) a time unit, and (3) a given stan- 

 dard instrument. All these desiderata are supplied in such 

 a statement as this : that the evaporating power of the air in 

 a given locality and for a given period is such as to produce 

 the evaporation of so and so many grams of water per hour 

 from a standard spherical porous-cup atmometer. . No unit of 

 area is considered, although all the internal characteristics of 

 the instrument are implied by its name. 



It is clear that it makes no difference what sort of surface 

 we may use as standard, but we must use the same standard 

 throughout any series of comparative measurements, and when 

 several instruments are needed we must be sure that their in- 

 ternal characteristics are as nearly alike as possible, as far as 

 these characteristics may influence the rate of evaporation. 

 The only feasible way to compare a number of instruments 

 in this last regard is to place them all in the same environ- 

 ment (as far as environmental characteristics may influence 

 the rate of evaporation) and then compare their evaporation 

 rates. If these rates differ this must be because of internal 



