STUDIES IN THE EXPERIMENTAL ANALYSIS OF SEX. 165 
larly over the spindle and distributed apparently at random 
to either end of the spindle in the division. Synapses 
apparently do occur between certain chromosomes or even 
groups of chi'omosornes, but the whole process is plainly 
abnormal. I have not observed any cases of multipolar 
mitoses such as Guyer describes, but otherwise in the 
inequality of size and distribution of the chromosomes our 
observations agree. 
But now occurs tlie remarkable process which accounts for 
the hybrid spermatozoa being of double normal size. After 
a careful search through the material of all three hybrids I 
can find no trace of the second maturation division ever 
occurring at all. Instead of this the secondary spermatocytes 
are at once converted into spermatids of twice the normal 
size, which proceed forthwith to transform themselves into 
spermatozoa. Many of these spermatids are of normal appear- 
ance (figs. 9 h and c and life and c), except that their nuclei 
and cell bodies are just twice the normal size, and they become 
elongated in the ordinary way to form the normal double- 
sized spermatozoa. Others, however (figs. 10 and lid), 
appear abnormal from the Brst, quite apart from their large 
size, the nucleus being uniformly stained a deep black with 
hacmatoxylin, and it is evident that these give rise to the 
abnormally twisted, branched or beaded forms of spermatozoa, 
some of which are figured on PI. 8, figs. 3-5. 
The process of transformation of these abnormal spermatids 
into deformed spermatozoa is shown in some cases in figs. 10 
and 1 1 d. The fact that the second maturation division is sup- 
pressed accounts completely for the double size of the majority 
of the hybrid spermatozoa ; the occurrence, however, of a com- 
paratively small percentage of small or normal-sized spermatozoa 
in the hybrids leads one to suppose that the second divisioji 
does occasionally take place, though so rarely that it would 
easily be missed. 
'I'he point at wliich the spermatogenesis in the hybrid begins 
to become abnormal is at the first maturation or reduction 
division. A comparison of the spermatogonia of the normal 
