This is mostly supposed to consist of unicellular organisms, fresh water algae, diatoms, but the larvae of Anophelince may also exhibit cannibalism. The exact nature of the food of larvae has been little studied. Most of the data on the subject refer to the vegetable organisms that are found in the water from which the larvae were collected. It is generally stated that they feed on fresh water algae,-Spirogyra, Mougestia, Protococcus, etc.,- but in the stomach may also be found Protozoa, and it is possible that these form a considerable portion of their food; on the other hand, vegetable matter is certainly consumed, and may be found in the intestinal contents, and the green color of larva? is in part but not entirely due to this cause. A peculiar fact which one often observes in larva? derived from foul puddles is that the body is almost entirely covered with a mass of Vorticellce or allied organisms. Again, to determine the vegetable organisms present in every water frequented by larva? would be a large task, and it is doubtful how far one would derive any especial advantage therefrom. We may, however, name the following plants, found in waters infested with A. maculipennis, and which were collected by Ed. and Et. Sergent in Algeria. They were Myriophyllum, Elodea canadensis, Elatine alsin astrum, Ranunculus (Batrachium), Nymphcea alba, Nuphar luteum, Trapa natans, Limnanthemum nymphceoides, Hydrocharis morsus rance, Alisma natans, Potamogeton spp., Lemna spp. With regard to Lemna, we have repeatedly observed that " tanks" in India and pools covered with it do not harbor larvae, though in favorable situations; whether or not the exploration is, as has been suggested, that it offers a mechanical obstruction to the movements and existence of the larva? remains to be shown; on the other hand, a small amount of Lemna, according to Grassi, favors the growth of larva?. In the intestinal contents we find also bacilli, and among these the " molds," identical with those which, as we have previously seen, Schaudinn finds in the esophageal diverticula of the adult.
In the young larva the thorax is not so broad as the head, but in the adult nearly twice as large. The thorax is furnished with numerous hairs or bristles, which are not of importance for purposes of recognition of species, but in some species we find on the thorax hairs identical with those on the abdomen known as palmate hairs. They are situated posteriorly and laterally, and often require careful examination for their detection. They are well developed in M. culicifacies, M. listoni, and P. jeyporensis; they are not present in the majority of Anophelince, and are absent, for instance, in A. maculipennis. Those larva? in which palmate hairs occur on the thorax are found in streams of water where there is considerable movement. The abdomen of the larva consists of nine segments, the first seven of which are similar in shape, decreasing in size as they reach the posterior end. The eighth segment is peculiarly modified, bearing the respiratory stigmata, and the ninth segment, on which the anus opens, is cylindric and bears various appendages. The first two segments in A. maculipennis bear at their posterolateral margins a pair of long branched hairs. The third segment carries a single similar pair. The other segments bear smaller unbranched hairs, and, further, there are shorter tufts of hairs on either side of the longer hairs, but these hairs do not appear to vary much in different species. On some of the segments, however, we find hairs of quite a different character. From their resemblance to a fan or the leaf of a Palmyra palm they are termed "palmate hairs." They lie nearer the middle line than the lateral hairs, and about half way between the front and back margin of each segment, though their position shifts toward the posterior margin as the hin dermost segments are reached. They really consist of a number of leaflets or rays set upon a central little knob or stalk; when opened out, the leaflets are extended like the rays of a fan, and practically form a complete semicircle, or even a circle. They are seen in their extended position when the larva is suspended under the surface film of water, and, in fact, their function is to support the larva in this position. At other times, when the larva is beneath the. surface, the fan may be seen closed and the leaflets huddled together. In the freshly hatched larva the fan is ill developed, but about the third day it has developed half a dozen lanceolate leaflets, which later take on the peculiarities of the particular genus and species to which they belong. The fully developed hair consists of about 15 to 20 leaflets, and from the position and characters of these hairs we can derive great assistance in the identification of species. We stated that the hairs are found only on certain of the first seven segments, and occasionally on the thorax. Thus we have: (1) Well developed hairs on all segments, one to seven, including the thorax in P. jeyporensis, M. listoni, M. culicifacies. (2) Well developed hairs on the second (or third) to seventh segments. (3) Rudimentary hairs on the first segment and even on the second, and on the thorax in N. stephensi, N. maculatus, N. theobaldi. (4) On the third, fourth, fifth, sixth, and seventh segments only M. sinensis, M. barbirostris, A. maculipennis. (5) On the fourth, fifth, and sixth segments only M. turkhudi. The commonest arrangement is the second one, viz., with palmate hairs on the second to seventh segments. But beyond the distribution of the hairs, which may not furnish much help, we have in the characters of the individual leaflets characters of specific importance. We can divide the leaflets into two portions-the basal portion, like an elongated wedge, and the terminal portion or filament, which joins the base of the wedge by a series of notches; or, without these, the basal portion gradually tapers into the filament. In some leaflets there is no sharp distinction between basal portion and filament; in others the filament is long and slender, and yet in others short and stumpy. We may find, then, the following types of leaflets: (1) The leaflets are unbrokenly lanceolate in shape, notched along the terminal half, and without a terminal filament-e. g., Mr. barbirostris, Mr. sinensis; (2) filament distinct and long-M. rossii, M. culicifacies, M. listoni, N. fuliginosus. Further, in this type we can distinguish one species from another; thus in M. rossii the filament is as long as the leaflet and the notch at the point of junction is wanting, while in M. listoni it exists. (3) The filament is very short-a spike like process-e. g., N. maculatus, N. theobaldi, N. maculipalpis. In M. turkhudi the filament is short and blunt. With regard to the great majority of larva? no observations on these points have yet been made.