205 
(at the foot of the Himalayas) it is as high as seventy- 
two (Fig. 67). We found, however, that there was one 
important matter in which the Duars differed from 
Calcutta, and that was in its Anopheline fauna. 
Whereas in Calcutta M. rossi was the predominant 
species, in the Duars M. listoni was the commonest 
Anoftheline. 
Again, in the Jeypore district (Madras), we had 
a district of uniformly high endemic index (50-100), 
and here we found an Anofheline , P. jeyporensis , which 
we had not encountered elsewhere, so that the view 
seemed tenable that the high endemicity of these 
districts was dependent on their special Anopheline 
fauna. To test to what extent species was concerned 
in determining endemicity, we then made use of 
another more exact method, viz., determining by 
dissection whether any difference occurred amongst 
the different species in the percentage of infected 
specimens : we were able to carry this out in the case 
of M. rossi and M. culicifacies. We caught these 
species in the same huts in the same villages at the 
same time, and determined by actual dissection the 
percentage of glands infected with sporozoits. The 
results were most striking, and fully confirmed our 
previous idea, based on more general considerations 
of the importance of species. They were as follows 
I. Mian Mir (Punjab) 
Number 
dissected 
Number 
with sporozoits 
Percentage 
M. culicifacies 
259 
12 
4-6 
M. rossi 
496 
O 
O 
