180 
A Study of Trijpanosome Strains 
based on 1000 individuals. Here there is a mode about 24"0, with possibly a sub- 
mode at 19 microns, but the evidence for dimorphism has largely disappeared. 
It is very desirable that we should know the details of this curve, i.e. the nature 
of the hosts and so forth, for it apparently replaces the earlier data and remains 
the standard T. briicei distribution. It certainly shows nothing of the definite 
heterogeneity (or dimoiphism) of the previous Uganda material. 
Its constants are as follows : 
Mean 23-5290, 
yu, = 19-30583, = 996-877(54, 
/ti, = 10-54837, yu,, = 2146-37930. 
24(/ - 101-86182^ - 4-0057f/ + 140-6937y^ + 62-0835?^ 
- 29-3940g3 + ll-237lf/ + l-4413g + -0331 = 0. 
No suitable root of this equation exists and accordingly it would appear that 
this distribution is not rigidly reducible to Gaussian components. This result is 
so remarkable in view of the obviously bi-modal character of the earlier T. brucei 
distribution, and the resolution into two components of all the other seven 
distributions, said to be allied to T. brucei, that I determined to consider the 
matter further by fitting Gaussians to the 'tails' of the T. brucei distribution*. 
I chose as the right-hand ' tail ' the frequency from 28 to 88 inclusive, and as the 
left-hand 'tail' the frequency from 13 to IS microns inclusive. The two resulting 
components were : 
T. minus. T. majus. 
m, = 20-0817 (19-83), Wo = 26-4359 (25-95), 
0-1 = 2-8685 (1-85), a. = 3-6399 (3-03), 
n, 628-16, Jh = 467-52. 
The totals populations for each component are clearly not very good and their 
combination exceeds by 9-6 "/^ the total observed population ; but the means are 
not widely divergent from the avei'age values resulting from our six resolutions, as 
the numbers given in brackets testify. Accordingly I determined to select the 
means of the components at values near the mean values of six reductions, and 
after one or two slight betterments, determine the sizes of the populations and 
their standard deviations so as to give the mean, and second and third moments of 
the observed population. These provided : 
T. minus. T. majus. 
77.1= 19-8244, h7o= 26-1122, 
(Tj = 2-6439, a., = 3-4134, 
Ml =410-83, n, = 589-17. 
Bioiuetrika, Vol. ii. p. 1 and Vol. vi. p. 65. 
