96 
AMERICAN JOURNAL OF BOTANY 
IVol. 9. 
as compared with 19.4° F. three days previous. The undercooHng value of 
the Greensboro buds rises but 0.5° F. during the same period. On March 14 
the Elberta buds were about a day further developed than the Greensboro, 
as indicated by the difference in time required for their petals to open when 
Fig. I. Graphs showing variation in the freezing point (upper pair) and undercooling 
(lower pair) of fruit buds in the Elberta (broken lines) and the Greensboro (continuous 
lines) peach. 
placed in the greenhouse. Whether or not these undercooling measure- 
ments mean that Elberta and Greensboro fruit buds are of practically the 
same degree of hardiness from January 21 to within a few days of the time 
the petals open can not be determined from the scant data of this one year. 
The extent of undercooling is probably a more important index than the 
actual freezing point. The freezing point in all probabiHty is a more 
constant value depending largely on the concentration of the cell sap, while 
the extent of the undercooling is further conditioned by agitation of plant 
tissues due to wind movement, size of capillary films, rapidity of cooling, 
etc. The undercooling here obtained in the laboratory is not necessarily 
the undercooHng that would occur outside under natural conditions. In 
the former case the cooling was very rapid as compared with similar changes 
under the usual natural conditions. Neither must these undercooHng 
temperatures be confused with the actual killing temperatures. As a 
matter of fact, the actual minimum temperature^ between February 18 and 
^ These temperatures were recorded by Mr, Thomas H. White, who has charge of the 
local U. S. Weather Bureau Station. 
