292 ENVIRONMENTAL CONDITIONS. 



least in showing the importance of this cHmatic condition — there are 

 available just two logically planned series of atmometric measure- 

 ments in the United States. One of these series was carried out by 

 Russell for a single year, beginning in the summer of 1887. The other 

 was conducted by the present writers during the summer of 1908. 

 Approximately 20 years elapsed between these two series of observa- 

 tions, and no thorough study of this feature has been completed since 

 the last-named year, although the U. S. Weather Bureau is giving in- 

 creasing (but always secondary) attention to evaporation. It should 

 be noted that the United States Signal Service, the precursor of the 

 United States Weather Bureau, carried out the earlier of these series, 

 the second series being under the auspices of the Department of Botani- 

 cal Research of the Carnegie Institution of Washington. These two 

 series of atmometric observations, and the results derived from them, 

 will now be considered in order. 



(2) Russell's Data of Evaporation in the United States. 



Evaporation intensities for period of average frostless season. (Table 

 11, plate 53 and fig. 14.) — Russell's^ study of evaporation extended from 

 July 1887 to June 1888 inclusive, and this author prepared an 

 evaporation chart of the country, but the data thus used were cal- 

 culated. For the period from June to September 1888, Piche atmom- 

 eters w^ere exposed in louvred instrument shelters at 19 stations. 

 An experiment in a closed room, employing two Piche instruments 

 and two open pans of water, gave data from which Russell calculated 

 that the Piche instrument lost 1.33 times as much water as did his 

 free water surface.^ By use of this ratio the readings of the Piche 

 instruments in the louvred shelters were converted into losses from 

 the free water surface of the particular kind of pan used in the labora- 

 tory test, and these, as tri-daily readings, were compared with the 

 corresponding dew-points and wet-bulb temperatures within the shel- 

 ters. From this comparison Russell derived a formula by which he 

 afterwards calculated the evaporational loss from free water surfaces 

 in the shelter, for 140 stations in the United States. In his paper he 

 presents a table of the monthly calculated rates (July 1887 to June 

 1888), and also the annual totals. 



In order to make use of these data in connection with the length of 

 the average frostless season, as we have employed the latter, we have 

 proceeded as follows for each station involved. The evaporation data 

 for all whole months included in the average frostless season have been 



1 Russell, T., Depth of evaporation in the United States, Mo. Weather Rev. 16: 235-239, 

 1888. See also Kimball, H. H., Evaporation observations in the United States, Monthly 

 Weather Rev., 32: 556-559, 1905. 



* Of course this relation must vary more or less markedly with temperature and humidity 

 conditions, even where the wind influence is out of account; but Russell seems to have ignored 

 this consideration entirely, along with the other important consideration that the amount of 

 evaporation is dependent on the sort of pan used. 



