STUDIES IN THE EXPERIMENTAL ANALYSIS OF SEX, 161 
the normal male pigeon and dove (figs. 1 and 2) vary within 
narrow limits, and that the heads are straight and of a fairly 
uniform thickness throughout. In the case of the three 
hybrids (figs. 3-5), the spermatozoa, drawn to the same scale, 
are seen in the majority of cases to be about twice the length 
of the normal spermatozoa, while some of them have 
thickened beads of chromatin along their length and are 
obviously deformed. 
A few of the hybrid spermatozoa drawn in the figures are 
of about normal size, but they are only few compared to the 
large or deformed varieties. In considering the size of the 
spermatozoa, only the heads are taken into account, as the 
tails are so delicate and stain with such difficulty that it was 
found impossible to deal with them. 
As it wa.s desirable to establish what proportion of 
the hybrid spermatozoa were twice the normal size, camera 
tracings were made of a large number of spermatozoa taken 
at random from the normal and hybrid males, and the 
camera tracings were measured with dividers. 
Spermatozoa showing beads upon thetn or complex coils 
were not used, only the comparatively straight and uniformly 
thick heads being measured. The samples of the hybrid 
spermatozoa were therefore not perfectly random samples, 
but since it is the large spermatozoa which usually exhibit 
the beads and coils, the error introduced would tend to 
diminish the proportion of large to small spermatozoa. 
As to the method of measuring the camera tracings with 
dividers, it is obvious that a rather large experimental error 
is introduced owing to the spermatozoa heads being curved, 
but by carefully stepping with the divider’s over the sharper 
curves and not attempting to take account of variations less 
than a millimetre, the main results may be relied upon. 
Tlie I’esult of the measurements of the spermatozoa heads 
of the three hybrids and of three normal .male pigeons and of 
a normal male dove are given in the subjoined table, the fre- 
(luency with which the various sizes of spei’inatozoa heads 
occur in each case being read off in the horizontal columns. 
VOL. 58, I’ART 1. KKW SERIES. 11 
