EVAPORATION AS A CLIMATIC FACTOR. 47 



tube about fifteen inches in length, the latter passing through a 

 cork stopper into a Mason jar, the container for the supply of 

 water. Cup and tube are filled with distilled water and the tube 

 is carefully inserted into the jar, which has been previously filled, 

 in such manner that no air is allowed to enter the tube. Evapo- 

 ration takes place from the surface of the cup and water to 

 make up for the loss thus occasioned is drawn from the jar be- 

 low. The jar bears a file mark on its side to indicate the stand- 

 ard level, and is filled to this mark at each reading, the amount 

 of distilled water required for this being measured and furnish- 

 ing the reading of the instrument. The apparatus was placed 

 with the jar beneath the soil surface, the cup projecting about 

 six inches above. Readings were taken once a week, which was 

 found to be often enough for the work in hand. Considerable 

 trouble was experienced this season with a consignment of im- 

 pure water, which so clogged the pores of the cups that the latter 

 became useless. Fortunately for the work, an instrument had 

 been installed at the University of Arizona, by Dr. W. B. Mac- 

 Callum, and his records have been used in place of those from 

 the instruments which were injured. The University campus 

 is so situated that the evaporation conditions there are very 

 similar to those of the plantation where this work was carried 

 out. 



Among the plants tested were : the garden nasturtium 

 (Tropaeolum) , morning-glory, marigold, sunflower (Helianthus 

 anniius), mustard, castor bean, muskmelon, teasel (Dipsacus 

 sylvestris), and jimson-weed (Datura Stramonium}. The seeds 

 were sown in May, in the midst of the spring dry season, and the 

 behavior of the plants was watched till early September, when 

 the experiment was brought to a close. 



From the time of planting, the drought conditions continu- 

 ally increased in severity until July 6th, when the summer rainy 

 season opened. From May I3th to July ist the weekly aver- 

 age of the daily temperature maxima rose from 87 to 107 F., 

 while the corresponding average of temperature minima rose 

 from 45 to 75 F. For the same period the weekly average of 

 maxima and minima of relative humidity decreased from 45% 

 (maximum) and 32% (minimum) to 31% (maximum) and 



