26 



The plants employed were seedlings of the creosote bush 

 (^Coz'illed) and palo \'erde {Parki)isonia Torreyajia and P. aatlcata) 

 growing in cans and supplied with measured quantities of water 

 at stated intervals. The rate of transpiration was determined by- 

 placing the plants under a bell-jar, with suitable precautions to 

 prevent the absorption or escape of water vapor, the amount of 

 water transpired being derived from readings of a hygrometer. 

 As details will be given elsewhere, a brief resume of experiments 

 and results will be sufficient for the present purpose. 



Beginning with the palo verde, two sets of plants, one serving 

 as a check on the other, were used. August ii, the plants hav- 

 ing been well watered the day before, the rate of transpiration 

 was determined. The following day, August 12, the plants 

 meantime having received no water, but having been treated 

 precisely as before, as regards light and other controllable condi- 

 tions, the rate of transpiration was found to be only 52.6 and 

 38.5 per cent, as high as it was on the preceding day, a result 

 apparently attributable to nothing else than the diminished 

 quantity of water in the soil in which the plants were growing. 



The same plants were again placed under observation August 

 18, having been given no water since August 15. External con- 

 ditions were favorable to transpiration, full sunlight, a fresh 

 breeze, and rather high temperature. At 1 1:40 A. M., after the 

 rate of transpiration had been noted, number i was given one 

 ounce, and number 2 three ounces of water. At 1:15 P. M., the 

 rate of transpiration of number i was found to be the same as at 

 the time of the preceding observation, while that of number 2 

 was twice as great. At 4 P. M., observations were again made, 

 and at this second afternoon reading it was found that number i 

 was transpiring twice and number 2 four times as rapidly as at 

 the time of the forenoon observation. 



The following forenoon the rate of transpiration of number 2 

 was found to be nearly four times as great as that of number i, 

 a striking difference when it is considered that only twent\'-four 

 hours earlier their rate had been the same, explainable, it would 

 seem, only by recalling the fact that when the observations began 

 on the morning of August 18, both sets of plants were in dry 



