909 
May is, 1925 Possibility of Sex Control by Artificial Insemination 
The measurements of swine sper¬ 
matozoa from seven different slides are 
shown in Figures 8 to 14, together with 
the theoretical normal curves 6 for 
populations of those sizes with those 
means and standard deviations. Slides 
Nos. 519, 531, and 536 were made of 
semen out of the testicle of one boar, 
slides Nos. 522 and 526 were made 
from the fluid recovered from a natural 
service by the younger boar, and slide 
537A was made of semen from the 
testicle of still a third boar. Two 
separate sets of measurements were 
made on slide 531 (figs. 10 and 11). 
Fig. 8.—Actual (solid line) and theoretical (broken 
line) frequency distributions of the head lengths 
of untreated swine spermatozoa taken from the 
testicle of an old boar castrated three or four hours 
before this slide was prepared. The probability 
that the actual distribution is only a chance 
deviation from the theoretical is 0.017. (Slide 
No. 519) 
Fig. 9.—Actual (sona nnei and theoretical (broken 
line) frequency distributions of the head lengths 
of centrifuged swine spermatozoa. From the 
extreme outside end of the centrifuge tube after 
five minutes of centrifuging at 1,600 r. p. m. 
From the same samDle of semen used in preparing 
Figures 8, 10, and 11. The probability that the 
actual distribution is only a chance deviation 
from the theoretical is 0.31. (Slide No. 536) 
Fig. 10. —Actual (solid line) and theoretical (broken 
line) frequency distributions of the head lengths 
of centrifuged swine spermatozoa. From the 
extreme inside end of the centrifuge tube after 
being centrifuged five minutes at 1,600 r. p. m. 
From the same sample of semen used in preparing 
Figures 8 and 9 and from the same slide from 
which Figure 11 was prepared. The probability 
that the actual distribution is only a chance 
deviation from the theoretical distribution is 
0.041. (Slide No. 531) 
Fig. 11.—Actual (solid line) ana theoretical (broken 
line) frequency distributions of the head lengths 
of centrifuged swine spermatozoa. From the 
extreme inside end of the centrifuge tube after 
five minutes of centrifuging at 1,600 r. p. m. 
From the same sample of semen used in preparing 
Figures 8 and 9 and from the same slide from 
which Figure 10 was prepared. Figures 10 and 
11 were made from measurements made on 
spermatozoa from different parts of the same slide 
and thus serve as a control upon the accuracy of 
the measurements. The probability that the 
actual distribution is only a chance deviation 
from the theoretical distribution is 0.30. (Slide 
No. 531) 
6The theoretical normal curves were constructed according to the method explained in West s Intro¬ 
duction to Mathematical Statistics” (1,2). The curve is complete y determined by^three namely, 
its mean, standard deviation, and the number of individuals involved- These> three 
retical curve are taken from its corresponding actual curve. The the Goodness o^ 
a chance deviation from the theoretical normal one is obtained by the X* method of testing the goodness of 
fit of one set of observed data to a corresponding set of expected data (15 ). 
