116 
A Study of Trypatiosome Strains 
There are still further results of importance to be ascertained, however, from 
our table of human strains. Let us compare Strains IV and V, which we found 
resembled each other closely even for compounded hosts. We now reach 
14 035 and P=-5229. 
Or, the probability that these two strains are identical has been reduced by 
selecting out the rat data only. But the result is still so high that no one would 
hesitate to assert that Chipochola and Chibibi were suffering from a disease due 
to the same strain of trypanosome. The correspondence is so close that we have 
combined Strains III and V for all other comparisons. In the case of Strain III, 
we have added together the results for Rats 952 and 953. Such addition is less 
reasonable for Rats 726 and 728, but without doing this, it is impossible to decide 
which rat is to represent the E strain. I have then made the following com- 
parisons : 
Strains IV and V with III : = 525 67, P < -000,000,01. 
There is accordingly no similarity at all between the Chituluka strain and that 
common to Chipochola and Chibibi. 
Strains IV and V with II : = 64-70, P < -000,001. 
Thus the strain from the European E from Portuguese East Africa diverges 
from the Nyasaland strain widely, but not as widely as that of Chituluka does 
from those of Chipochola and Chibibi. 
Strain I with Strain III : = 126-13, P < -000,000,1, 
Strain I with IV and V : = 217-82, P < '000,000,01. 
Thus the trypanosomes from Mkanyanga are widely divergent from those of 
the three other Nyasaland cases. Nor are they any closer to the European E : 
Strain I with Strain II : x^ = 331-37, P < -000,000,01. 
Thus with the exception of the Chipochola and Chibibi strains, the trypanosome 
distributions from human sources differ widely. Nor is this to be wondered at, if 
the human beings owe their trypanosomes to Glossina morsitans, for in that case 
we should expect the human strains to be as diverse as we have found those from 
the tsetse fly itself It would remain to explain the close similarity of the 
Chipochola and Chibibi cases. It would be interesting to know the history of 
these cases with regard to locality and to the possibility of a unique source 
of infection. 
(c) In the case last dealt with, namely that of Chipochola and Chibibi, we 
have the remarkable feature that the strains although significantly identical, 
whether treated in the rat alone or in compounded distributions from various hosts, 
resemble each other somewhat less closely in the single host series. This is not 
generally the rule. Some of the big divergencies we have already noticed become 
far less appreciable, nay, even become resemblances when we confine our attention 
to one species of host. The chief misfortune which then too often arises is the 
