268 
Journal of Agricultural Research 
Vol. VII, No. 6 
varieties. This differentiation seems to hold with remarkable constancy 
notwithstanding the wide geographic origin—West Indian, Bahaman, 
Central American, Mexican, and Hawaiian—of the seeds or budwood 
from which the tissues dealt with originated. 
The type which is characterized by the slightest freezing-point lowering 
of its extracted sap—that is, the type in which the expressed sap freezes 
at the highest temperature—is the one which has been shown by horti¬ 
cultural experience to be the least capable of enduring cold. That 
capacity to withstand low temperatures is not solely due to differences 
in the freezing point of the sap is evident from the slightness of the dif¬ 
ferences in the cryoscopic constants of the West Indian as compared 
with the Mexican and Guatemalan types. Furthermore, horticulturists 
believe that the plants of the Guatemalan type are intermediate in hardi¬ 
ness between those of the Mexican and West Indian types. There is, so 
far as our data go, no discernible difference in the freezing point of the 
sap of these types. 
The problem is evidently one of considerable complexity. To what 
extent other characteristics contributing to the capacity of the organism 
to withstand low temperatures are correlated with sap properties remains 
to be investigated. 
It seems highly probable from the evidences presented in the paper 
that in the case of the tropical perennials, a knowledge of the freezing- 
point lowering of the sap would be of some service in predicting ability 
to withstand cold. At least the subject is one deserving of more extensive 
investigation. We would have been glad to carry out the present study 
on a far more extensive scale, but the determinations given practically 
exhaust the trees of flowering age in the collection of Guatemalan and 
Mexican types at the Miami Plant Introduction Field Station, and it 
will probably be several years before a better series is available. 
