94 
AMERICAN JOURNAL OF BOTANY 
[Vol. 9. 
The size and shape of the capillary water films within the plant tissues 
undoubtedly have a marked influence on the amount of undercooling 
necessary to start crystallization. Also, the amount of water held by 
imbibition depends on the colloidal properties of the bud tissues. The 
degree of hardiness is so modified by these conditions that undercooling 
and freezing-point measurements of expressed saps do not in all probability 
indicate a true index of hardiness. 
Numerous theories^ have been advanced concerning the mechanical 
cause of injury due to low temperatures. Whether the harmful results come 
from failure of the protoplasm to regain water lost when ice crystals form 
or from precipitation of proteins or from other metabolic changes accom- 
panying low temperatures, there is evidence that the injurious effects come 
at the time that ice crystals form or shortly thereafter. It must not be 
understood, how^ever, that injur\ always results from water crystallization 
alone. Experiments were carried out during the early part of 1921 (January 
to March) with the idea of comparing the undercooling of sap within fruit 
buds of two peach varieties. The method used was somewhat similar to 
that described by Harvey.^ A copper-constantan thermo-junction made of 
number 40 wires and encased in a small glass tube (0.4 mm. diam. by 3.0 
mm. long) was cemented to the end of a piece of hard rubber tubing. When 
temperature measurements were made this junction was inserted into the 
upper end of a detached bud and lowered into a double-walled chamber 
surrounded by a freezing mixture of ice and salt. The bud was so suspended 
as to be free from contact with the walls of the cooling chamber. The 
cooling chamber was of such a size that it fitted very conveniently through 
the cork stopper of a pint thermos bottle containing the freezing mixture. 
The constant-temperature junction was placed in a similar thermos bottle 
filled with melting ice. This constant-temperature junction could very 
easily be kept at 0° C. for two days. A galvanometer similar to that 
described by Shreve^ was calibrated and used to measure the temperature 
differences of the junctions. The copper leads were connected directly to 
the binding posts of the galvanometer so that a continuous reading could 
be made. Between each two successive readings the zero point on the 
galvanometer scale was obtained by a key connected directly w4th the 
binding posts. A temperature difference of 0.1° C. could very easily be 
detected. 
In the experiments here reported fruit buds from two trees (Elberta 
and Greensboro) were studied. These trees, located near each other in 
a nine-acre peach orchard, were growing under very similar conditions. 
* A brief review of a number of these theories and a bibliography of 50 titles dealing 
with the hardening process in plants is given by Dr. R. B. Harvey in "Hardening process 
in plants and developments from frost injury." Jour. Agr. Res. 15: 83-1 11. 1918. 
^ Harvey, R. B. Importance of epidermal coverings, Bot. Gaz. 67: 441-444. 1919- 
^ Shreve, Edith B. A thermo-electrical method for the determination of leaf tempera, 
ture. Plant World 22: 100-104. ^9^9- 
