11 
useful additional character for taxonomic study. 
From biochemistry let us turn to physiology as a tax- 
onomic character. In Africa the important malaria carry- 
ing moscpiito, Anopheles gamhiae breeds in fresh water. In 
East Africa R. C. Thomson has shown that there is 
also a distinct form breeding in salt water. No constant 
distinguishing morphological characters have been found, 
but captured females can be identified by physiological 
tests on the larvae hatched from the eggs they lay. Larvae 
are hatched in fresh water and transferred to 75 /( sea 
water. All larvae of the salt water form survive at least 
six liours, while all larvae of the fresh water form are 
dead within two hours. This is rather a tedious way of 
identifying a mosquito, but by using it Thomson has been 
able to show that the two forms differ in behaviour and in 
their role in malaria transmission. Further tests on crossing 
the two forms are needed before it can be decided whether 
they are distinct species. 
That sort of taxonomy has to be done on the spot. 
P. F. Mattingly, at the British Museum in London, has 
been worki^ig on another group of African mosquitoes, 
species of Aedes (sid)genus Stegom ijia) . After checking 
many indentifications he took records from over 500 lo- 
calities. for which he listed latitude, longitude, altitude 
and annual rainfall. Then he plotted the distribution rec- 
ords of one species on a map and considered it in relation 
to altitude and rainfall and what was known of the habits 
of that mosquito. 
For example, by drawing boundaries marking the 
limits of 40 inches with three dry months per year, and 
100 inches with four dry months he obtained a hypothetical 
distribution which agreed very well with the known dis- 
tribution of Aedes fraseri. At the same time these boundar- 
ies suggested the possibility of its being found in certain 
areas for Avliich no records exist. It is known that another 
species, Aedes luteocephaluSj will not enter closed canopy 
forest and so in mapping its potential distribution, the 
big areas of equatorial forest had to be excluded. 
The putative distribution of Aedes dendrophilus Avas 
mapped Avithiu the limits of not more than tAvo dry months 
in the year. It is a species found in fringing forests in 
Avest and central Africa and there is an isolated population 
in southeast Africa. Phis suggests that it is a more ancient 
species than some of the others and that in the past the 
southern forests Avere much more nearly continuous Avith 
those of Central Africa than they are to-day (a hypothesis 
already put fonA^ard for the same reasons by workers on 
birds and butterflies). 
