20 
REPORT OF THE MALARIA EXPEDITION. 
times a da\-, so that the larva; could live secure from desiccation in all but the most evanescent 
puddles. In September, however, there was a complete break in the rains, lasting three days. A 
large number of the pools, e\ en many of tliose containing water-weed, and those fed by springs 
during rain, dried up completely. Tlie question whether the larv;e liad the power of living in 
the mud at the bottom of the pools could now be testeil by direct observation. The break in the 
rains was followed b\ hea\ \ showers, which immediateh' refilled all the puddles. Had the larv;e 
conf iuied to exist in the mud, the\' would now have emerged again. As regards the puddles in 
which the mud had completeh dried, this was not the case. No larv;t at all were found in them 
for at least two da\'s after the rain had refilled them. After that interval larv;e again appeared ; 
but the\' were very small ones, evidentU' just hatclied from the egg. On the other hand, it was 
frequently observed that if the mud did not become completely dry, the larvae would emerge into 
active existence after another shower. These observations were supported by some experiments 
1)1 vitro ; and we therefore conclude that the larv;e can withstand partial, though not complete, 
desiccation.* 
(6.) Till- samr puddlrs Lontantly occiip'wd. — We have suggested S^paragraph ,7.j'] that the 
position of the breeding pools ma\ change according to the seasons ; but while we were in 
Freetown there was no change of season, and we generally found Anopheles larv.t in the same 
puddles — namely, in tliose which were suitable for them. Thus, of two puddles h ing close 
together, one would never contain larva-, and the other would always contain them. 'Die 
explanation of tiiis probabl\' is that the larva- perisii in the unsuitable pools, or tiiat the adults 
generally return to the same pools in order to la)- their eggs. It seems likel)' that the adults 
generally lay their eggs in the pools in which they themselves were bred ; and that the insects 
thus learn by experience the places most suitable for them. 
(7.) Deti'ctiou. It is eas\' to overlook Atiophi'li's larva- unless the)' are searched for in a 
bright light. 
(8.) Pupa-. The pup;e of Anophr/rs seem to be smaller than those of the commoner 
species of Ciilrx. The\' require about 48 hours to reach maturit\- /;/ vit)-o ; perhaps less in natural 
conditions. 
16. Bionomics of Adult Anopheles, (i.) Hutching. -The adults generalh- hatch out 
in the evening ; but their exit seems often to depend on the meteorological conditions of the 
moment, and appears to be dela\ ed b\' rain\ and wind\' weather. 
(2.) Food. The\' can easily be kept alive in glass cages, test-tubes, bottles, etc. We kept 
some in this manner for a fortnight, and could doubtless have kept them longer if we had wished 
to do so. We were able to confirm Bancroft's statement [18] that gnats feed on bananas; but 
they seem to prefer the fresli fruit. During the daA' tlie insects remained at rest on the walls of 
the cage ; but in the evening began to fly about and to walk over the fruit, plunging their 
proboscis into it in man\' places, so that the banana was sometimes covered with the gnats, both 
■ One of lis re;ir<;<l adults from full-grown l.n vae kupt on damp blotting-paper (in India) ; but found that the young larva; 
died \xhen kept under these conditions. 
