140 GENETICS: W. H. TALIAFERRO Proc. N. A. S. 
immune to another infection with T. lewisi. It is apparent that in an 
organism which shows a cycle of growth and division such as character- 
izes T. lewisi, we must make all comparisons of size and variability at 
the same stage in the cycle, and that if there is a period in which there 
is no growth and division comparisons should be made during this period. 
A study of the changes in the means and in the coefficients of variation 
of the different parts of the trypanosome throughout the course of a 
"pure line" infection demonstrates very clearly that there are different 
periods of the infection and indicates the stage at which to make com- 
parisons of size and variability between different infections. Let us 
take, for example, the coefficients of variation and the means for total 
length in rats 116 and 105 which are shown in figure 2. The infection 
in rat 116 was started from a single trypanosome and rat 105 was inocu- 
lated from rat 116. In other words, although the infection in the first 
rat started from a single specimen and in the second rat from many speci- 
mens, all of the trypanosomes in both rats are descendants of the single 
organism injected into rat 116. We can make no determinations during 
the incubation period since no one knows where the organisms are at 
this time. Let us consider first the constants for rat 105. On the first 
day of the blood infection the mean length was 24.785 =±= .423. This 
rose rapidly until the 5th day when it reached 30. 108 ±.280. This rise 
continued gradually until it reached 31.412 =±= .065 by the 19th day. From 
the 19th until the 32nd day, at which time the infection disappeared from 
the blood, there was no significant change in the mean. The coefficient 
of variation behaved in much the same manner as the mean. On the 
first day of the infection it was 26.52 =fe 1.35%. It dropped very rapidly 
for the first seven days and then much slower for the next twelve days. 
By the 19th day it reached the low value of 3.11 ± .14% and it showed no 
significant change from this value throughout the remainder of the in- 
fection. 
It is well to compare the conditions found in rat 116 with those in rat 
105. Rat 116 is given here because a longer time elapsed before the in- 
fection reached the adult stage than was the case in any of the other in- 
fections studied. This is due probably to the fact stated above, viz., that 
the infection in rat 116 started from a single trypanosome while the others 
although they were "pure lines" were sub-inoculated from 116, and in 
consequence were started with a large number of specimens. As these 
curves are probably expressions of the resistance of the host to the para- 
site, we would expect that this resistance would increase more rapidly 
when the infection is started with a large number of trypanosomes than 
when it is started with only one. We cannot compare the shape of the 
curves in the two rats very well because there are not enough points in 
the curve for rat 116. One thing is probably true, however, and that is 
that both the mean and the coefficient of variation reach a constant 
