1911] OVERTON—TRANSPIRATION AND SAP-FLOW 105 
experiment, therefore, it seems possible to kill a certain region of 
the stem without injuring the leaves, and to show that the dead 
portion is capable of water conduction so long as actual breaking up 
can be prevented. 
The objection may be raised that the killed region in the above 
experiment was too short to show conclusively that living cells 
are not necessary for sap-flow. I have succeeded in killing 10 cm. 
of the stem by another poison without injury to all of the parts 
above, and thus obtaining results quite as striking. A stem 23 cm. 
long, with a healthy crown of 11 involucral rays with 2 branches 
each 8 cm. long and 2 others each ro cm. long, was surrounded for 
4 cm. with a saturated aqueous solution of CuSO, for 36 hours on 
May 7. The CuSO, was absorbed and the stem discolored for 
20 cm., or up to the base of the crown. The liquid was removed 
and a plug of cotton put loosely about the stem in the tube inclos- 
ing the upper portion of the 10 cm. which was discolored. On 
May 20, or after 13 days, the plant was in perfect condition except 
one branch, some of whose leaves were drying, and the leaf tips 
of the crown, which showed signs of yellowing. This branch 
and one of the others finally dried and died, and 3 leaves of the 
crown died, but the remaining 8 leaves of the crown and the branch 
have remained perfectly fresh, green, and turgid up to the present 
writing (July 18); the other branch has grown several centimeters. 
The stem is partially collapsed, as indicated by its smaller size, and 
has colonies of mold (Penicillium) on it, owing to the fact that. the 
incasing tube was not closely sealed. The steni is dark brown, 
almost black, and is certainly dead. This experiment shows, there- 
fore, that it is possible for 10 cm. of a dead stem to conduct water 
for 8 leaves and 2 branches for an indefinite period of time. A 
photograph of this plant is shown in fig. 3. The results of the 
above-described experiments appear below in tabular form. 
From these experiments with picric acid, alcohol, and CuSO,, 
we see that it is possible to kill a portion of the stem without 
completely disorganizing the killed stretch and without interfering 
with its conducting capacity. 
I have further attempted to determine quantitatively the 
amounts of water that may be evaporated from perfectly dead 
