450 BOTANICAL GAZETTE [ DECEMBER 
through the leaves. Twenty-five healthy branches, taken from 
the same tree which furnished the diseased ones, were prepared 
ina similar manner. All the bottles were then brought into the 
laboratory and weighed twice daily for ten days. At the endof 
this time the branches and leaves were weighed and the amount 
of water evaporated per gram of dry weight was calculated. It 
was found that the diseased leaves evaporated approximately 
one-fifth more than the healthy, these results holding good 
through all the experiments, as will be seen by a study of the 
preceding diagrams. 
The fact that cutting off the diseased parts of the leaves 
prevented them from falling has already been pointed out. It 
was assumed that the removal of the injured portions stopped 
the excessive evaporation and enabled the leaf to heal the wound 
made by the cut, which it could not do in case of the fungus. 
To prove this a number of experiments were made, of which the 
following is an example : 
Twelve pairs of fresh leaves were selected, six pairs being 
diseased and six healthy. These were divided into four sets of 
three pairs each. Set no. 1 (healthy) had one-half inch cut from 
each leaf and the tips charred with red hot glass to prevent an 
excessive loss of water and turpentine. After charring, the tips 
were coated with a varnish, which prevented the entrance of 
fungi and slightly reduced the evaporation of water from the cut 
ends. Set no. 2 (diseased) was treated in the same way, the 
diseased ends being removed. Set no. 3 were healthy uncut 
leaves and set no. 4 diseased uncut leaves. All these sets were 
weighed at given periods and the loss in weight determined. 
The results are shown in the accompanying diagram. 
During the first twenty-four hours the healthy cut leaves 
lost 9™, the diseased cut and the healthy uncut exactly 7m each, 
and the diseased uncut 11™8. The same relative loss held good 
through the next forty-eight hours. oa 
The ends of the healthy cut leaves lost water more rapidly 
than the ends of the diseased cut leaves, making the loss appt” 
imate more closely to that of the diseased uncut leaves. *"° 
q 
a 
; 
| 
: 
4 
. 
