102 
A Study of Trypanosome Strains 
(a) I ask what difference is made when a strain is passed through various 
animals (goat, monkey, dog, rat) or through a single animal alone. Taking the 
wild-game strain discussed by Sir David Bruce and others*, we have: 
Microns. 
10 
11 
12 
13 
u 
15 
16 
17 
18 
Totals 
Wild-Game Strain ) 
(from various animals) ( 
2 
34 
85 
172 
119 
63 
22 
3 
500 
Wild-Game Strain ) 
(from a single I'at 510) j 
1 
9 
53 
137 
163 
117 
14 
6 
500 
Here we find 65'37 and P < 000,000,1. In other words the distribution of 
lengths of the trypanosomes of the wild-game strain obtained fi-om various animals 
differs so enormously from that obtained from a single rat that the two cannot be 
looked upon as samples of the same population. The moment this result is realised 
we appreciate that (i) it is impossible to compai'e two strains developed in a variety 
of animals unless we have previously tested on the same strain the equal valency 
of these animals, (ii) a series of animals of even the same species may quite 
possibly give widely divergent results from those obtained for a single animal. 
Thus passing from a variety of animals in wild-game strain to a varietj' in wild 
G. morsitans strain makes less difference (P = '000,008) — although great enough — 
than passing from a variety of hosts to a single rat in the wild-game strain. 
This rule is not universal, but it illustrates the absolutely essential need for 
testing the effect of change of host before questioning the identity or non-identity 
of two strains. 
{h) I now turn to the Mvera cattle strain, and ask what differentiation is 
produced by the dog and goat as hosts. The data are very sparse and unless we 
get a high degree of resemblance may be worth little. They runf: 
• R. S. Proc. Vol. 87, B, pp. 6 and 8. 
t R. S. Proc. Vol. 87, B, p. 3. I tested the relative interchangeability of goat and sheep in the case 
of r. ca^n-ae. The data are as follows : (R. S. Proc. Vol. 86, B, p. 280) 
Microns. 
T. caprae 
18 
19 
SO 
31 
aa 
23 
24 
35 
36 
37 
38 
39 
30 
31 
33 
Totals 
Goat 
3 
7 
11 
35 
43 
50 
33 
28 
27 
17 
5 
1 
260 
Sheep 
1 
10 
12 
29 
39 
31 
28 
20 
5 
3 
1 
1 
180 
leading to x"=18'088 and P = -1133 or the resemblance is considerable although not so great as we find 
between goat and dog for the Mvera cattle strain. 
