No. 408.] EFFECT OF HEAT UPON GROWTH. 955 
well as of the beginning and ending of the period of rapid 
imbibition of water. 
2. Under the conditions of the experiment, the absolute dry 
weight appears to undergo little change. There is apparently 
a slight loss from the begin- aüigrammes 
ning of the experiment to the 
period of maximum percent- 20 ES. 
age of water in the embryo. __ PEF 
In studying the data I do not / P= a 
find any constant relation be- 10 "d 
tween this decrement and the  , / 
temperature. Inasmuch as 4 
the prevailing difference in 06 5 10 15 20 25 30 35 
dry weight is so slight, it IG. 5. — Bufo. Graphic representation of Table 
scems probable that the sud- III, column 4, showing average total weight on 
den fluctuations of weight -7 rS das of experiment 
seen in some of the measure- 
ments may be due partly to ,,|. | | 
individual variation in the A | 
. 
amount of food taken and in 3? 
the presence of facal matter 1 
in the digestive tract, and \ 
partly to errors in weighing, l Y 
which may amount to as 15 
much as o.5 milligram. In .. | b. 
some of the smaller weigh- '. ipn Eom 
ings this would mean a pos- sis : 
sible error of 20%. Of course | "E 
the probable erroris much less * pays” 
than this. It appears safe to ric. 6. — Bufo. Graphic representation of the 
: . ratio of dry substance to the total weight. (See 
say that during the period em- Table III, column 7.) The broken line indi- 
braced in my experiments, Sas cates the probable posi on o the curve ; 
on hatching amb the at second measurement is doubtless inaccurate. 
tainment of the maximum percentage of water, — the dry weight 
is unaffected by temperature. It follows, therefore, that the 
acceleration and retardation experienced at ug p in cras 
growth of the larvae by reason of different temperature conditions 
is due almost entirely to the changed rate of imbibition of water. 
5 
S 
e 
M 
> 
wN 
E 
to 
© 
w 
e 
So 
S 
eo 
