A Study of Kvaporation and Plant Distribution. 209 



when taken with the data for the following year which are to 

 be given below. 



The data derived in 1907 were so satisfactory, and at the 

 same time so incomplete, that it was decided to continue the 

 operations through another season. Steps were accordingly 

 taken, with very helpful suggestions from Dr. H. N. Transeau, 

 now of the Pvastern Illinois Normal School, and with Dr. h'orrest 

 Shreve, now of the Desert Laboratory, to increase the number of 

 stations and otherwise improve the character of the results. 

 The writer's absence from the United States during the year 

 1908 would have made it impossible to secure a series of obser- 

 vations for that summer, had it not been for the hearty and 

 efficient co-operation of Dr. Shreve, to whose enthusiasm and 

 perseverance the 1908 results are entirely due. Dr. Shi eve 

 took complete charge of the operations throughout the season. 



In this series, the results of which it is the purpose of the 

 present paper to report, two standardized porous cups were 

 sent to each observer at the beginning of the season. These 

 were installed side by side, each placed as in 1907, and weekly 

 readings were made on both until mid-summer, when one of 

 the cups was returned to the Desert Laboratory and restand- 

 ardized by Dr. Shreve. This restandardized cup was then re- 

 turned to the observer, by whom it was again installed in its 

 old position, and operated with its mate till the end of the season. 

 By this method determination was made possible of the amount 

 of variation occurring in the correction coefficients of the various 

 cups, without at the same time interrupting the series of obser- 

 vations. At the end of the season both cups were returned to 

 the Desert Laboratory and all were standardized by the writer 

 during the spring of 1909. All standardizations in this series 

 were made with reference to standard cups, the standard being 

 directly or indirectly derived from the original standard of the 

 fall of 1907. It is this same standard to which all coefficients 

 derived since that time have been made to refer. It will be 

 remembered that the reading of any cup is to be multiplied by 

 its coefficient of correction, to give the reading which the stand- 

 ard cup should have given for the same time and place. 



Where the coefficient of correction of any cup altered, as 

 shown by the mid-summer and final calibrations, interpolations 



