12 
As a result of this Avork Mattingly also showed that 
a climatic bridge exists between Avest Africa and Abyssinia 
Avhich may help to explain the strong Avest African element 
in the Abyssinian fauna. 
He concludes that rainfall mainly influences distribu- 
tion indirectly through its effect on vegetation. This is 
supported by the occurrence of one species in a locality 
far outside the rainfall limits knOAvn for it elscAAdiere. In 
this locality there Avas a dense humid evergreen forest 
depending not on rainfall but on innumerable springs. 
But much careful botanical and entomological Avork needs 
to be done before the relationship of the various species 
of mostpiitoes to their characteristic plant communities is 
understood. Records of seasonal occurrence, of Avhich there 
are few, would also throw light on distribution. 
You can imagine all the A’aried research needed for a 
study of this sort — apart from identification of specimens 
there Avould be searching through atlases, meteorological 
records, government reports and rainfall maps — all 
done by a museum taxonomist. The result is not merely 
knowledge of present day distribution but also evidence 
regarding past climates in Africa. This sort of Avork also 
indicates clearly Avhere there are gaps in our knoAvledge 
and provides a stimulus to other Avorkers. 
In the I^acific there is a Avidely distributed group of 
mosfiuitoes belonging to the same subgenus, ^iegomijia, as 
Mattingly is Avorking on. This is the Aeden sentcUaris 
group and members of it are found from the East Indies 
to Taliiti on almost all the islands that lie Avithin tlie 
Tropics. There are certain differences, mostly rather 
small. betAveen the forms that occur on different groups 
of islands. The taxonomist is faced Avitli the (piestioii, are 
these distinct species, or are they subspecies of one Avidely 
distributed species, tiiat could interbreed if they met on 
the same island, or are the differences just due to differ- 
ences in environment ? I tried to find an ansAATr to the last 
(piestion Avhen I AA*as in England. I had a colony of a 
Fijian form, Aedes pseudoscufellarls and reared the larvae 
at different temperatures and in 1/3 seaAvater and then 
compared the development in the adults of the characters 
that are used to separate the difTerent forms. Rome of the 
most important characters Avere not affected at all. and 
others varied a little but not to the extent they do betAveen 
different forms; a couple proved to be quite variable and 
made one doubt their reliability as taxonomic characters. 
The net result supported the treatment of the different 
forms as di'dinct species. An unexpected result came out 
()f this Avork, because the lack of variation in certain char- 
acters shoAved that specimens from Fiji and other islands 
that had been identified as pseiidoscnfellans but did differ 
