There is no collection of water, however insignificant, which it is safe to disregard as a possible source of larvae. There are few places, however dry, where by careful search some unexpected source of water will not be found, and the existence of mosquitos, otherwise difficult to account for, readily explained. From large open sheets of water to the small collections in the hollows of leaves in the jungle larvae may be found. In concealed cesspits they may exist in myriads; in covered tanks, in deep wells, in cisterns on the roofs of houses, in the bilge water of ships, in running streams, in boats, in crab holes, besides all the well known sites about domestic dwellings, pots, pans, tubs, etc.-it would be rash to say that there could be no larvae there; nothing short of careful inspection should determine the point. Larvae of Culicidce are not infrequently found in brackish water, and Culex sollicitans will thrive in sea water (Taylor).
From deep wells, although mosquitos may be dislodged in crowds, it may be difficult to collect larvae; as by letting a bucket down the larvae are disturbed and frequently escape capture, so in inclosed cisterns, such as are used in the tropics for storing water, the proof of the existence of larvae may require special apparatus; but it is a safe rule to adopt that, when hard pressed, mosquitos will breed in any source of water.
In the detection of larvae mere inspection is not sufficient, especially if the water contains weeds of any sort. "Dipping" should always be resorted to. A white enameled cup or tin is suddenly plunged beneath the surface, and water and weed are brought up together. A little time is allowed to elapse for the water to clear, and in a minute or so the larvae come to the surface. Weedy ponds, which to the closest inspection show nothing, when searched in this way often yield many larvae. On the contrary, many promising pools yield no larvae at all. The explanation often lies in the fact that they contain small fish. Again, ponds covered with species of Lemna are often free from larvae. Not only fish, but a variety of water creatures prey upon larvae, and yet where both fish and these predatory enemies exist, larvae may also exist, provided there is a sufficient supply of coarse water weed to protect them from the view of their enemies. Not only have larvae many enemies, but they are also cannibalistic. In fact, not only do the different species prey on each other, but also the large members of one species will devour the small ones. It would appear, also, that some larvae are almost entirely cannibalistic-e. g., Psorophora sp., C. concolor, Mucidus scata phagoides, Megarhinus sp. Larvae are frequently absent from large open spaces of water, and not infrequently from deep waters. They also do not appear to frequent clear cool spring water. They are difficult to find in moving water, e. g., where a lagoon is affected by the flow of the tide.
In some places they are not found in the cement containers surrounding fountains, though in a cement tank in a shady garden, especially if it contains rotting leaves, larvae may abound. Again, of two tanks, one made of metal and the other of wood, it is the latter that is most likely to contain larvae, and in the wooden tubs used by natives for storing larvae in the tropics they exist often in amazing numbers, so that a native village may produce literally millions of insects (chiefly Stegomyia). Anopheline larvae put in a collecting bottle and exposed to the shaking of a long transit often appear dead, and they sink to the bottom of the water. They may, however, often be revived by carefully floating them out on the surface, when, shortly, signs of movement will be exhibited.
Taylor gives the following data: S. fasciata, minimum, six days; Culex confirmatus, four and a half days; Culex jamaicaensis, three and a half days; Culex sollicitans, four and a half days; Cellia argyrotarsis, twelve days. The larval stage of Megarhinus sp. is, however, considerably longer, lasting fifteen to twenty one days (and the pupal stage six days) at a temperature of 80° to 90° F. The shortness of these periods is remarkable, for Christophers found for Stegomyia sp. in India, where the temperature of the water was 96° to 102° F., that the duration of the larval stage was seven days, pupation taking place on the eighth day. For M. rossii Christophers gives eleven days as the duration of the larval stage, with pupation on the twelfth. These longer periods for the two anophelines noted above are very interesting, and extended observation on this point is necessary.
The nymphae are the globular headed, comma shaped bodies into which the larvae develop. They have a "tail" more or less curved under the head. They are not so easily seen as the larvae, and readily take fright, diving rapidly to the bottom, whence they arise again in a series of skips and hops. The differences be tween the nymphae are not so great or so readily perceived as in the case of the larvae. The main differences exist in the shape of the siphon tubes, a pair of which project from the bulbous thorax. These differences can be appreciated only by microscopic examination. In the Anophelince the siphons have a square, truncated end, and are shorter than those of Culex. The siphons of Culex are long and narrow and have a slit like opening. In Stegomyia sp. the siphons are broadly triangular and are very characteristic. Differences occur in many other genera, and many of the genera are still undescribed. The nymphae that may be most easily confounded with those of the Culicidce are those of Chironomus and Corethra. That of Chironomus is characterized by its conspicuous respiratory tufts. The pupa lies at the bottom of the water, with only the tufts and thorax projecting. The pupa of Dixa has not been very fully described: it has a general resemblance to that of Culicidce pupae. A close inspection would probably recall many points of difference. The "head" of the pupa of Corethra is not so bulbous as that of the Culicidce, and the tail fin is broader. The trumpets are pointed, with a slit like opening. The pupae of the sand flies (Simuliidce) lie concealed in cocoons attached to water plants, the respiratory tufts only of the pupa projecting.