65 
consequently, the periods during which greatest injury is done. Deter- 
niinatiojis of this character were made in connection with the previ- 
ously mentioned breeding work. 
Measurements made of the length of the body are of but little use, 
since the}^ may be varied so much by the movements of the larva. 
There is a rather constant and gradual increase, however, from birth 
to maturity, as shown b}^ the measurements given below, which are 
averages of twenty-live larvae. 
Table XX. — Showing amount of growth, during, and length, at end of, each instar. 
Instar. 
Amount 
of growth 
during 
each in- 
star. 
Lengtli at 
end of 
each in- 
star. 
I 
7nm. 
2.02 
2.65 
7nm. 
3. 52 
fi 17 
II 
III 
4 53 1 10. 70 1 
IV 
V 
VI 
7.68 
10.51 
13.35 
18.58 
29.39 
42. 74 
Thus it will be noticed that a larva shows 
a greater amount of growth during each 
succeeding larval instar. This is shown 
graphically in the accompanving curve 
(fig. 6). 
More accurate measurements are those 
based upon the size of the head, which does 
not change between the molts, and thus 
serves b}^ its almost constant size to identify 
the instar of any given larva. By meas- 
uring large numbers of head casts thrown 
off at the molts, it was found that they are 
fairly constant for each instar, sufficiently so 
to enable one to tell the stage of any larva 
positivel3^ The following table is compiled 
from measurements of over one hundred 
specimens: 
II III IV V VI 
Fig. 6.— Diagrammatic represen- 
tation of comparative rate of 
growth of larvse daring dif- 
ferent instars (original;. 
Table XXI. — Showing average width of head casts for respective larval instars. 
Width and range. 
Stage I. 
Stage II. 
Stage III. 
Stage IV. 
Stage V. 
Stage VI. 
Width 
mm. 
0.30 
(«) 
mm. 
0.48 
0. 42-0. 54 
mm. 
0.78 
0.67-1.00 
mm. 
1.33 
1.21-1.50 
mm. 
2,05 
1.73-2.28 
vim. ' 
3 95 
Range 
2 85-5 00 
a Practically constant. 
