30 
larva nearly reaches the lower limit of the lake larva; but even 
while comparing two sacs from the two larval forms respectively 
of the same length and curvature, I have always found a diffe- 
rence in the general shape, the sac of the lake form being more 
slender than the other. The greatest transversal diameter in pro- 
portion to the greatest length (measured from one end of the sac 
to the other along the curved longitudinal axis, as it ought to be 
measured, if the sacs were straight), I have found to be, in full- 
grown larvæ of the pond form, varying from 1: 2,3 to 1: 2,5, in 
the lake larvæ from 1 : 4,4 to 1: 6,6. 
As I have not been able to examine young larvæ of the lake 
form, I cannot say for certain if their air-sacs are of another shape 
than that of the fullgrown larvæ; but as the curved sausage form 
is so marked and, so far as my experience goes, relatively unva- 
ried in the adult larvæ, it seems not likely that the form of the 
sacs of the young specimens should differ in a very marked de- 
gree from it. — The bladders of the lake larva seem to be nearly 
of the same shape as those in the larva of C. /Sayomyiaj fusca 
Staeg. described by de Meijere; here the bladders of young larvæ 
are somewhat more reniform, but not as much as the typical form 
of C. plumicornis. 
The pigment cells are generally larger and more regular in 
outline, penta- or hexagonal, in the lake larva than in the pond 
larva, and as a rule more regularly placed like the squares in a 
mosaic, exceptions are however not rarely met with. 
As it appears from this description of the air bladders of the 
two kinds of larvæ, there is a marked difference in the outer form 
and general aspect of them. A still more characteristic difference 
is to be found in the anatomical structure, especially in the thick- 
ness of the walls in adaptation to the different pressure to be en- 
dured. : 
If we examine a sagittal section of an airbladder, we will find 
that the wall is of very different thickness in the various parts. 
The thickest part is that of the concave ventral wall, next to it 
comes the wall of the dorsal convex surfåce, the thinnest parts 
being the rounded ends (Fig. 3). The bladders present, when viewed 
from outside, a similar transversal striping as the ordinary tracheæ 
of other insects; but in section the walls show a somewhat diffe- 
