28 THE TUNDRA AND ITS FAUNA 
curious that they should be abundant in this cold 
region, but the cause is to be looked for in the un- 
developed condition of the drainage system, already 
mentioned. The absence of any run-off in winter gives 
the tundra, as we have noted, the characters of a semi- 
arid region, with the result that the water formed in 
summer by the melting of the snow tends to accumu- 
late in innumerable lakes and pools. In these the 
mosquito larvae swarm, and the natural checks, e.g. in 
the shape of small fish who feed on the larvae, must be 
insignificant. From the pupal stage the adults emerge 
towards the middle of June, and they are then extra- 
ordinarily abundant till the middle of August, when 
they disappear, after having laid their eggs in the pools. 
These pools freeze or dry up with the onset of winter, 
but the eggs or larvae remain uninjured, to hatch or 
recommence active life with the coming of water in the 
spring. It is possible that the life-history is lengthened 
by the cold, but the rapidity of reproduction is certainly 
not checked, except in the coldest regions. In addition to 
mosquitoes a considerable number of other flies occur, 
especially about the houses of the Eskimo, &c., breeding 
in the refuse thrown out. Indeed, flies predominate 
among northern insects, the percentage increasing as 
we pass northwards. The reason lies apparently in 
the fact that the larval stage is usually short, and 
can be run through in the brief period of warmth. 
With the exception of the curious parasites known as 
Strepsiptera, all the other orders of insects are repre- 
sented, the bees and butterflies fulfilling here their 
usual role as fertilizers of flowers. The very short 
active period of vegetation has a curious effect upon 
the development of plant-eating insects. As is well 
known, in more temperate countries butterflies usually 
