36 



TRANSPIRATION IN A DESERT PERENNIAL. 



gram per s(iuaro centimeter represents the average maximum for leafless 

 forms, 0.0079 gram for adult trees in leaf, and 0.0164 gram for grccn-house- 

 raised plants. Experiment VI was omitted from the series because of the 

 error in area which has already been mentioned; so also was VIII, because 

 it represents readings in the shade and does not shoAv the droj). 



The maxima for experiment IX have been reduced to white-atmometer maxima by multiplying them by 

 1.162, a correction factor which was found by averaging the results obtained by dividing the brown-atmometer 

 leadings by the corresponding white-atmometer reatlings on July 27 and on August 5. (See experiment X.) 



Considering only those experiments on plants growing under natural con- 

 ditions which show a definite drop and rise in both the actual and relative 

 transpiration, it appears that the maximum of actual transpiration reached 

 after the drop is in five cases lower than the one before the drop, and in two 

 cases the later maximum is about the same as the early one, while in no case 

 is the second one higher than the first. It therefore appears that, for a 

 certain form under similar conditions, there is a maximum water-loss beyond 

 which the plant does not go and that this is frequently reached without a 

 previous decrease in the relative rate. In the case of green-house-raised 

 plants, where, as has been pointed out, the drop is replaced by a lessening of 

 the increase of relative rate and is followed by an abrupt increase in relative 

 rate, it appears that while there is a tendency toward a constant value for 

 the actual transpiration immediately preceding diminution in the increase of 

 the relative rate, yet it is not so marked as in the case of plants in situ. In 



