304 ENVIRONMENTAL CONDITIONS. 



Column 6 of table 15 presents the sums of Russell's three monthly 

 losses, for each station considered, for June 1888 and July and August 

 1887, these numbers representing inches of depth. To obtain the 

 average monthly rate, each number is to be divided by 3; to obtain 

 the weekly rate each is to be divided by 13, etc.; but since any such 

 alteration in the time increment considered would result in applying 

 the same coefficient to the entire series of indices, and since relative 

 indices for the different stations are all that are here requisite, we 

 employ the totals simply. These are charted in plate 55, where isoat- 

 mic lines are represented for increments of 5 inches, those for the 

 values 10, 15, and 30 being distinct. This chart exhibits about the 

 same features as does that based upon the length of the mean frostless 

 season as the duration factor. 



(3) Evaporation Studies of 1908. 



Presentation of data. — As has been stated, a series of evaporation 

 observations were carried out from the Desert Laboratory in the sum- 

 mer of 1908, the cylindrical porous-cup atmometer being employed. 

 This is the first and only fairly representative series of direct measure- 

 ments of this climatic feature that has been carried out for the United 

 States, and we shall here enter into considerable detail in the discussion 

 of our results. 



As has been mentioned, certain aspects of this evaporation study 

 formed the subject of a paper by Livingston. ^ The following pre- 

 sentation will proceed without attempting to distinguish between 

 what is now published for the first time and what is here repeated from 

 Livingston's paper. Since that paper was pubhshed the data have 

 been thoroughly revised, which accounts for some numerical dis- 

 crepancies between our tables and those of Livingston. 



Two standardized porous cups were sent to each of 38 stations in 

 the United States and Canada, in the spring of 1908, and were there 

 operated side by side in the open till midsummer, when one of them 

 was returned to the Desert Laboratory and restandardized. The 

 restandarized cups were then sent back to their respective stations and 

 again installed, and the two then continued to operate till the work 

 was discontinued. In the fall both cups were returned and restan- 

 dardized. Thus coefficients of correction were determined for the 

 beginning, middle, and end of the season's operation, and the amount 

 of variation in these coefficients was determined without interrupting 

 the series of observations. All standardizations were made with refer- 

 ence to standard cups, the same standard being employed as is still 

 in use for cylindrical cups supplied by the Plant World. Thus the 

 corrected readings here given are directly comparable to corrected 



' LivinKston, B. E., A study of the relation between summer evaporation intentsity and centers 

 of plant distribution in the United States, Plant World 14: 205-222, 1911. 



