108 
A Study of Trypanosome Strains 
may simply be the result of the rat strain having been developed in the dog as 
host. The divergence between the dog and goat strain may again be solely due to 
the greater variety of sources in the goat. Tlie data from the wild G. tnorsitans 
experiments seem to indicate that the observed divergences between the strain 
from rat and the strain from goat or dog may not be due to difference of host, 
but to difference of source from which the material was drawn, and to difference of 
treatment of the individual stock of trypanosomes, e.g. the number of hosts, etc., 
through which it has passed. 
It seems absolutely certain that at the present time most light would be 
thrown on the conditions for asserting sameness or diversity of strains, by well 
devised experiments on strains from single sources passed through different species 
of hosts in different manners, in order to determine the exact measure of divergence 
produced by host and by treatment, and ultimately to devise a standard treatment 
for all strains which we desire to compare. 
The exact nature not only of host, but of standard treatment is most vital. We 
can demonstrate the influence of treatment at once by considering the " percentages 
of posterior nuclear forms among short and stumpy forms " recorded by Sir David 
Bruce and his colleagues for the wild-game strain*. AH the trypanosomes were 
from rats, and although the date of infection of the rat is, I think, not stated, the 
dates of first extraction will be after much the same interval, and we can therefore 
classify by date from first extraction. We find the following table : 
Wild-Game Strains. 
Percentage of Posterior-Nuclear Forms among 
Short and Stumpy Forms. 
From first Extraction 
and under 
2S °/„ and over 
Totals 
6 days and under 
7 days and over 
18 
6 
G 
18 
24 
24 
Totals 
24 
24 
48 
Using Sheppard's formula for the four-fold table, we have for tetrachoric r 
r = -707 ; 
or, the correlation between this character of the trypanosome and the time 
after infection of extraction is very considerable. It will be obvious that in a 
standardised treatment this time of extraction will play a most important part. 
But it again is not independent of the species of trypanosome, for if we take the 
wild Glossiy^a morsitans strainsf, we find : 
* R. S. Proc. Vol. 86, pp. 396—404, Tables III, VI, IX, XII and XV. 
t R. 8. Proc. Vol. 86, B, pp. 410—418, Tables III, VI, IX, XII aud XV. I have added one percentage 
by random selection from the complete table by lot in order to give 60 cases, and save labour in 
fractionising. 
