N. H. SWELLENGREBEL AND J. M. H. SWELLENGREBEL-DE GrRAAF 1 83 
It is often extremely difficult to procure mosquitoes in numbers sufficient for 
the determination of the N. I.; and, moreover, if they are found to be all 
uninfected, it may still be asked whether there was no flaw, and whether the 
species under examination is not after all a good vector. In experimental 
work no such doubt exists, provided the gamete carrier is a good one, and 
that another species, known to be a good vector, is used as a control. It is 
also easier to obtain large numbers of mosquitoes artificially bred than caught 
in nature. On the other hand, there are many factors which determine the 
chance of the mosquito becoming infected, and which are not taken into 
account in experimental work, and so do not influence the E. I., whereas they 
may materially affect the N. I. In this way it may happen that the deter¬ 
minations of E. I. and N. I. yield different results. Thus we found this to 
hold for M. sinensis, which was easily infected experimentally, but uniformly 
showed a low N. I. In this connection we may quote the observations recorded 
by Stephens and Christophers, and by James, which revealed a similar dis¬ 
cordance between the N. I. and the E. I. of M. rossii. As to which of the two, 
the N. I. or the E. I., should be trusted in practice, the above-named observers 
unhesitatingly express a preference for the N. I. and we agree with them. 
Experiments show that under favourable conditions certain species may 
transmit malaria, but thev leave us in doubt as to whether these conditions 
obtain in nature. Our observations may leave us uncertain whether under 
changed conditions a given species will or will not become a good vector, but 
at least they tell us that under the conditions existing in a given country, 
a particular species actually does or does not harbour malaria parasites. If 
such observations carried out in various localities, yield uniform results, the 
latter, in our opinion, are of more practical value than those obtained by 
experiment. We have therefore adopted the determination of the N. I. as our 
standard method to discover the principal vectors of malarial parasites, and, 
according to the conditions formulated above we have extended our researches 
to various localities. 
The figures representing the N. I. of one and the same species in different 
places, all of them highly malarious during the period of observation, often 
show a marked divergence (Table III). One of the principal causes of this is 
the fact that gamete-carriers are more numerous and potent in regions of 
epidemic malaria, than in countries where endemic malaria prevails. Where 
there is endemic malaria, the spleen-rate is much the same in both children 
and adults, and it is invariably high (80 per cent, and more). The parasite- 
rate, in young children especially, is considerably higher than in adults. 
Comparing the rate of infection in children and adults, with the three species 
of malarial parasites separately, it appears that in adults, the three species 
do not disappear at the same rate, the decrease of simple tertian and quartan 
being more marked than subtertian. In the case of subtertian parasites, 
a striking feature may be noted, namely, that the decrease of crescents is 
greater than that of rings and the number of gametes present in the crescent- 
