130 BOTANICAL GAZETTE [AUGUST 
readings of the hygrometer were reduced by means of the Smith- 
sonian meteorological tables and the appropriate correction then 
applied for the bell-jar employed. Later experiments indicate that 
the correction applied in the present case must be considered approxi- 
mately rather than quantitatively exact, but this does not affect the 
value of the comparisons that follow. 
From the tables here given it is seen that for the time of observa- 
tion the rate of transpiration of the two plants respectively was: 
no. 1, 924™6 per hour; no. 2, 102™£ per hour. 
Pa Te 
j 5 A Beit Mn a 
‘Be 8 
j it ina 
oe 
BT 
GS SB a 
i 
Eas | 
as 2 aR 
j= 5 ak SSG Ga 
7 8 Ws SO eh WS + 
2 a 
We a 
Try 
ea a 
| Sa a 
| OS ai 
| 
T 
| 
J ET Por TT aT 
cal Bs RE BY 
potty 
| | | 
Fic. 4.—Curves showing rate of transpiration of two creosote bushes and amount 
transpired by no. 1 in 31 minutes and by no. 2 in 1 hour and rx minutes. 
By counting the leaves of each plant and estimating their surface 
and that of the green shoots on which they were borne, the entire 
transpiring surface was estimated as: no. 1, 1533°™; no. 2, 660%. 
For equivalent surfaces, therefore, the rate of transpiration of no. 
1, the plant on the hill, was 3.7 times that of no. 2, the plant on the 
plain below. Further experiments gave similar results. A branch of 
a creosote bush growing where the ground had been thoroughly soaked 
a few weeks before by the running over of water from the tank of the 
Desert Laboratory was exceptionally fresh and green, and its rate of 
transpiration, for equivalent surfaces, was found to be 8.9 times 4 
8reat as that of the bush on the mesa. se 
From these and other detailed experiments not here reported; It 
1s abundantly proven that after months of excessive drouth the 
