a 
SEPTEMBER 4, 1913] 
NATURE 21 
become noticeable. In the first place, inside the head 
a spasmodic pulsation is visible, at first at long in- 
tervals, but later more or less continuously. I have 
observed this pulsation in eggs of other water-beetles, 
and also in those of the dragon-fly, and although I am 
not sure that the interpretation is the same in dragon- 
fly and water-beetle, 1 am satisfied in the latter case 
the pulsation is really a swallowing process. 
The larve of all the water-beetles I have examined 
possess a special sucking apparatus known as a 
“pharyngeal pump,”’ the use of which I shall describe 
directly, and in the embryo this pump apparently 
comes into use to absorb the fluid which surrounds 
the embryo in the shell; the embryo merely drinks this 
up. 
After this sucking-pump begins to work, various 
other movements of the internal organs can be ob- 
served, including peristalsis, and also at infrequent 
intervals the whole body moves slightly in the shell, 
the tendency being to push the head into the end. 
One other movement is to be noted, and that is an 
up-and-down motion of the head, at first very slight, 
but later becoming very marked. 
On either side of the head is a small papilla, at 
the apex of which is a minute, slightly curved spine. 
When the embryo is at rest, this papilla lies in a 
slight depression, but when the sucking-pump is at 
work the papilla bulges outward, so that the spine 
touches the shell. Thus when the head moves up and 
down and the sucking-pump works at the same time, 
the two spines scrape along the inside of the shell and 
ultimately burst it open. They are, therefore, ‘‘ hatch- 
ing spines,” and similar instruments differently 
situated have been observed in a few insect embryos 
of other orders. 
You see, therefore, that the shell bursts open at 
the head end; immediately it bursts the compressed 
larva bulges out, and by slight writhing movements 
works its way clear of the shell, the whole operation 
taking less than two minutes. As soon as the larva is 
clear of the shell the tail straightens out, and the 
legs and mouth parts assume their natural position. 
In the embryo there is a peculiar fold in the upper 
part of each jaw, but within two or three minutes of 
the larva’s escape this fold has completely disappeared. 
From the moment the larva escapes it begins to 
grow in length and breadth. The long air-tubes in 
the body are flat, but have a bright silvery appearance, 
suggesting that some gas has been secreted in them; 
but the larva is heavier than the water, and therefore 
sinks to the bottom. For a time, half an hour or 
more, it rests quietly and shows no desire to get to 
the surface, but sooner or later it gets restless and 
swims to the surface, using its feathered legs as oars, 
and raises its tail to the surface film and remains 
suspended for a few minutes. After this the newly 
hatched larva is buoyant, and cannot remain away 
from the surface without holding on to the submerged 
vegetation. The buoyancy is, however, only tem- 
porary, as older larve frequently require to swim to 
the surface to renew their air-supply. 
In the insect, breathing and blood-circulation are 
normally not intimately associated as in other 
animals. In a human being or a fish, or even in a 
snail, air is taken into special organs—lungs or gills 
—where the blood takes up the oxygen and carries 
it through the whole body. In the insect the blood 
has usually nothing to do with the aération of the 
different organs, the whole body being permeated by 
innumerable air-tubes. y 
In all the water-beetle larva which come to the 
surface to obtain their air, these innumerable air- 
tubes communicate with two large air-tubes which 
run the length of the body, one on each side, and 
these open on the last segment. Hence, when a larva 
NO. 2288, VOL. 92] 
requires to renew its air-supply it comes up tail first, 
bringing the openings of the two lateral trachez into 
communication with the air, and by contracting and 
expanding the body it exhales the used-up air and 
inhales fresh air. 
For a day or so after hatching the larva is sott and 
is not hungry, but once its skin and jaws have . 
hardened it begins to look about for food. I found 
that tadpoles and pieces of chopped worm were suit- 
able food, but under natural conditions small newts, 
water-shrimps, and insect larvag—including brothers 
and sisters—constitute the normal diet. It is impos- 
sible to keep two larva together in one small vessel, 
as one invariably attacks and kills the other within 
a few hours. Even when I gave a tub to four speci- 
mens only one survived after a few weeks, so that 
in a small loch, where at least some thousands of 
these larve hatch out, the death-rate must be 
enormous. 
The method of feeding of the larva is peculiar. 
The two long sharply pointed jaws are each pierced 
with a fine tube, of which one end opens on 
the inner side just below the apex, and the other end 
opens on the upper side just near the base. When 
the jaws are closed the inner ends of these tubes 
communicate with the corners of the mouth, but 
when the jaws are open the inner ends of these tubes 
do not communicate with the mouth at all. The 
mouth itself is also peculiar. In a front view of the 
head it is visible as a long narrow slit between the 
bases of the jaws, but if this slit is examined it is 
found that across the lower side of it is a raised ridge 
which fits into a groove running across the upper 
side of it. When the jaws are wide apart the ridge 
and groove are separated, and the mouth is open, but 
as soon as the jaws come together the ridge fits into 
the groove, and the mouth is closed. As soon, there- 
fore, as the larva seizes its prey its mouth is closed, 
and the only communication into it is through the 
tubes in the jaws, the basal ends of which now open 
into the corners of the mouth. 
Immediately behind the mouth is the powerful suck- 
ing-pump, the pharynx, which I mentioned in connec- 
tion with the embryo. By expansion and contraction 
of its muscles it sucks in the juices of the prey 
through the tubes in the jaws. But if this were the 
whole process of feeding there would be a consider- 
able waste, as a worm or a tadpole consists of a 
large amount of solid material; and yet, if one 
watches one of these larva feeding, one will find that 
almost nothing is left of the prey except the skin, 
This is due to the fact that at short intervals the 
sucking-pump stops working and saliva is poured into 
the prey. This saliva digests and dissolves away 
the solid parts of the food, which are then sucked in 
by the larva. The process of digestion, which in most 
animals takes place internally, is carried on in these 
larvz outside the body. 
With regard to the duration of the larval period, in 
my examples this varied from six to nine weeks. 
This period is divided into three stages, there being 
two moults prior to the final one which produces the 
pupa. Each of the first two stages only lasts about 
ten days. so that the last stage is a very long one, 
as it is in all other insects. 
This last stage is also divisible into two parts, the 
first occupying four or five weeks, during which the 
larva feeds and grows as in the previous stages, the 
second occupying two to four weeks, being spent out 
of the water making a cell in the earth, and resting 
preparatory to becoming a pupa. 
In the few cases which I had the opportunity of 
observing, the full-grown larva always left the water 
in the morning between eight and ten o’clock; but 
whether this is the rule with this species, or whether 
