Operation op the Porous Cup Atmometer. 115 



evaporating pans are operated in a row, about 50 cm. apart, 

 and the average loss from the four is considered the standard. 

 These stand on a table in the greenhouse or in a closed room 

 (to avoid air currents) and the cups to be tested are arranged 

 around them about 25 cm. apart and an equal distance from the 

 pans. The cups are mounted as heretofore, but in ordinary 

 6 or 8 ounce bottles, which bear a file mark on the neck. At 

 each observation the bottle is filled to the file mark, from a 

 burette, and the amount of water required is the reading for that 

 period. 



It was found that the standard cups of 1907 and 1908 e- 

 vaporate only 0.75 as much as does one of these pans, so that, 

 in order to render the work of various years directly comparable, 

 the actual standard used in calculation of the coefficients is 

 not the average loss from the pans, as above stated, but this 

 quantity multiplied by 0.75. A standard cup may be defined as 

 one showing a loss of 0.75 as much as the average of the four pans 

 described above, the comparison being made without air cur- 

 rents. Sunshine affects cups and water surface alike, so does 

 not alter the relation. Thus, the standard cup loses water at 

 about the same rate as 45 sq. cm. of water surface exposed as 

 above. 



The calculation of the coefficients of standardization has 

 been perform^ed as heretofore, three-fourths of the average loss 

 from the four pans being treated as the reading of the standard 

 cup. Thus, if the four pans lose in 24 hours Pi, P2, Ps, and P4 

 grams of water, respectively, and the cup to be tested lost L 

 cubic centimeters, the coefficient for that cup would be 



L 

 C — . This coefficient is expressed in two 



|xi(P.+ P.+ P3+P0 



places of decimals; it is the number by which the reading of the 

 cup must be multiplied to give the reading of the standard cup. 

 Since the evaporating power of the air for any particular 

 station can be determined for comparative purposes only by 

 means of the two coefficients of correction, it has been found well 

 to have a formal record for each cup. This is a slip of paper bear- 

 ing the number of the cup, the original coefficient and the date 

 of testing, ^ jiote of the use made of the cup, and the final co- 



