114 BOTANICAL GAZETTE [FEBRUARY 
decreases more uniformly. These facts, taken together with the 
appearance of the mesophyll, seem to show beyond question that 
the leaves on a steamed stem are dying, and that death is not 
entirely due to a lack of water. 
Applying heat by means of wax heated to 110° C. does not cause 
so much discoloration of the contents of the sieve tubes, or so much 
stoppage of the lumina of the vessels, and the mesophyll of the 
leaves does not show so much immediate plasmolysis. The leaves 
above stems so treated remain turgid three times as long as those 
on steamed stems. Plants do not show such an immediate decrease 
in the transpiration as is the case when the stems are killed with 
steam. The effect upon the transpiration rate and upon the injury 
to the leaves in using hot wax is not a progressive one, as it is when 
steam is used. Altogether the method of using hot wax to kill 
a section of the stem seems more satisfactory than steaming, with 
reference to the question of the relation of the living cells to sap- 
flow. And it is still plainer in stems killed in this way that the 
death of the stem cells does not per se directly affect the sap-flow. 
My experiments with poisons, in which 5-10 cm. of the stem were 
treated for 36-72 hours, show that it is possible to kill a portion of 
the stem without completely disorganizing the killed stretch and 
without reducing its conducting capacity. Not only does a sufficient 
quantity of water pass through the poisoned portions to supply the 
transpiration needs for a comparatively long period of time (3 
months in the case of CuSO,), but also to allow the development and 
growth of new parts. As has been shown, the mesophyll cells 
remain perfectly normal; no discoloration of the chloroplasts and 
no contraction of the protoplasts follow this method of treatment, 
if care is taken that the poison does not reach the leaves. In the 
poisoned portion of the stem the cells are apparently “fixed”? when 
alcohol or picric acid is used, there being no plasmolysis., When 
CuSO, is used, the parenchyma of this region has its protoplasts 
contracted, but the vessels remain normally open and apparently 
unaltered. 
From my experiments, in which poisons were used to kill the 
whole plant, it was at once evident that the kind of poison greatly 
influenced the subsequent rate of evaporation of water from the 
