definitive host by mosquito bite. Whether it may obtain an entrance by any other channel or medium it would be hard to prove and rash to deny. That the young filariæ can live in water for a time (about seven hours) is certain; it is conceivable that some of them, such as those which at the completion of their stage of development in the mosquito find their way into the abdomen of the insect, may escape into water when the mosquito lays her eggs or dies.*^[1]

Technique.— Mr. Max Colquhoun made for me beautiful sections illustrating the metamorphosis of the filaria in mosquito by the following technique. The insects, preserved in glycerin, were soaked in dilute acetic acid 5 per cent, for one day only; then in formalin 50 per cent., water 50 per cent.,for one day; next in absolute alcohol for one day; absolute alcohol and ether, equal parts, one day; finally, in celloidin for one day. They were then sectioned, placed in strong hæmatoxylin for two hours, decolorized in acid, hydrochloric 1 per cent, and alcohol, washed in water, passed through aniline oil, xylol, and mounted in xylol balsam. Paraflin sections are not a success.

In working with fresh mosquitoes, all that is necessary is to tease up the thorax of the insect with needles in normal salt solution, and, after removing the coarse fragments of the integuments of the mosquito, apply a cover-glass and place under an inch objective. The filariæ are readily detected. If the slide is dried, fixed, and stained with a watery solution of some aniline dye or with logwood, and mounted in xylol balsam, beautiful permanent preparations will be obtained.

1. ↑
   • In my original observations on this subject in 1879 and 1883 (Trans. Linnean Soc., 1883) I supposed that filarial metamorphosis, so far as the mosquito was concerned, is completed in from six to seven days. This I now believe was too short an estimate.
   Bancroft (Journ. of Trop. Med., 1899) has since shown that it is necessary to feed the experimental mosquito (at all events, C. fatigans), so as to keep it alive for at least sixteen days, to obtain the final stages of the metamorphosis. (Fig. 106. ) I have little doubt now that the insects on which my observations were made had fed repeatedly, without my being aware of it, after their meal on the filariated patient. Bancroft has shown that by proper feeding (he used banana for the purpose) mosquitoes may be kept alive for several months, any filariæ they happen to harbour remaining alive, but not advancing in development beyond the final stage alluded to in the above description. In the observations referred to I conjectured that the fully metamorphosed filaria escaped from the insect either at her death or when she deposited her eggs, and that thus, in drinking-water, it obtained a chance of gaining access to the stomach of a human host. Before Low's observation Bancroft had very nearly guessed the truth, for he had expressed the idea that the filaria may be injected into man by the mosquito when it re-feeds, or that it may be swallowed by man while it is still enclosed in its intermediary insect host.