1909. ] LIFE-HISTORY OF THE AGRIONID DRAGONFLY. 273 
February, as it moulted once during that period, but so far as my 
records go it was not ! 
Table 4 (p. 272) gives the dates of appearance of the imagines 
and also the number of stages through which the nymphs 
passed. 
With regard to the question whether a high temperature in 
winter tends to reduce the number of stages in the life of the 
nymph, the material is too scanty to form any sound opinion. 
For such a question only one species can be considered at a time, 
as probably all species do not moult the same number of times 
under similar conditions. Also the individuals of any one species 
will not, under apparently similar conditions, all moult the same 
number of times, so that the average of a fair number of 
individuals would be necessar y to get any definite result. 
With regard to Agrion pulchellum, the only thing that can be 
said is that, with the single exception already referred to, the 
only nymphs which completed in 10 stages were three w hich had 
been incubated during the winter, while the only ones which 
passed through the ie stages were two which had been kept cold 
during the winter. On the other hand, three of the winter- 
warmed nymphs passed through 12 stages, while two of the others 
got through in 11; andlam inclined to think, after studying the 
rate of growth, that the incubator did nothing towards reducing 
the number of stages. I shall discuss this in relation to the 
question of body-length. 
Bopy-LENGTH. 
The average length of a nymph of Agrion pulchellum in its 
final stage is 18°0 mm., the smailest nymphs being about 14-0 mm. 
and the largest about 22°0 mm. These figures were obtained by 
measuring a number of nymphs with 7-segmented antenne 
collected without any regard as to the number of stages they had 
passed through, and they show a wide range of variation in size. 
In the newly-hatched nymphs there was but little difference in 
body-length, but at each succeeding moult the differences became 
more marked, so that, for instance, after the eighth moult, two 
-nymphs measured respectively 6°7 mm.and 14-0 mm. In noting 
this fact I also noted that the smaller nymph passed through 
13 stages, while the larger one completed in 10; and as the latter 
had been incubated during the winter, while the former had not, 
I at first thought that there would be evidence of reduction in 
the number of moults in the winter-incubated individuals. To 
test this I took the body-lengths of all the nymphs, so far as I 
had the figures, at each stage, and compared the length of each 
nymph in relation to the number of moults it ultimately went 
through, and the results are rather interesting and are given in 
Table 5(p. 274). It will be noticed that the tendency all through 
is for the larger nymphs in any stage to be those which ultimately 
complete in the smallest number of stages; and since this is 
noticeable in the beginning of the second stage and becomes 
more marked in succeeding stages, it seems that the earlier stages 
