July 15,1925 Effect of Low Temperatures on Bruchus obtectus 
171 
Comparing these figures (Table II) with the averages for individual 
freezings quoted above, it becomes apparent that all these stages are 
capable of resisting lower temperatures when they are not injured. 
This may be due to the cuticula preventing the rebound or, in other 
words, preventing “inoculation.” It may be that the presence of 
free water in the vicinity of the couple point may hurry inoculation 
in the case of the injured specimens. That an uninjured organism 
can better withstand extreme conditions than an injured one is a 
point that has been overlooked in much previous work. 
When an insect is impaled on a thermocouple point and cooled, the 
temperature recorded is that of the lymph and broken-down tissue 
in the immediate vicinity of the point. It may be of some physio¬ 
logical significance that the freezing point of the impaled insect's 
lymph is not the same as that of the insect as a living uninjured 
organism. 
The experiments described in the second part of this paper, when 
bean weevils in all stages were able to withstand temperatures as low 
as those shown to be the limit for injured specimens and for a greater 
length of time, seems to establish this point still further. 
RELATION BETWEEN SUPERCOOLING AND REBOUND 
Figure 2, A, B, and C, are dot charts wherein the supercooling 
temperatures have been plotted against the rebound points for each 
set of data on individual freezing. The charts suggest that there is 
a relationship between the two factors, so the Pearsonian method of 
arriving at the coefficient of correlation was used ( 12 ). 
The correlations that exist between supercooling and rebound are 
thus found to be: 
Corre¬ 
lation 
Probable 
error 
Adults ... 
+0.779 
+ .614 
+ .890 
0.053 
.084 
.028 
Pupae___ 
Larvae_ 
These high correlations indicate very definitely that the super¬ 
cooling point bears a distinct relationship to the rebound point. 
Just what this relationship is, however, is another matter. Accord¬ 
ing to chemical law (If pp. 172-180) the rebound point should be 
the same no matter how low supercooling goes. If this law applies, 
then the explanation of this correlation must be sought elsewhere. 
Knight (9) showed that when freezing P. bioculatus Fabr. a ther¬ 
mocouple placed near the insect registered a rise in temperature 
when the insect rebounded, showing that heat was given off to the 
surrounding air. This means, of course, that the thermocouple on 
which the insect is impaled does not register the total amount of heat 
of crystallization given off by the insect but that some is lost by radia¬ 
tion. These insects (. B . obtectus) are very small, not more than 3 
or 4 mm. long, so that their surface is very great in proportion to 
their mass. It follows then, that radiation to the air of the chamber 
must consume a considerable portion of the heat of crystallization. 
Because radiation varies as the difference in temperature between 
the source of the heat and the surrounding medium, it is reasonable 
