On the Dimegalous Sperm and Chromosomal Variation of Euschistus, etc. 133 
normal anapliases indicated by tlic letter s, or the „s-chromosomes” as 
they may be termed for brevity, show no sign of the constriction 
or longitudinal split that the other autosomes exhibit, which is readily 
explainable on the view that each of them had divided along the line of 
this split in the first maturation mitosis. 
How do these 2 s-chromosomes behave in the second maturation 
spindle? From what we have just learned it is certain they do not divide 
equationally for they did that in the first mitosis, and we do not know any 
case in animal spermatogenesis where a chromosome divides ecpiationally 
twice in the maturation mitoses. It is then probable they divide reduc- 
tionally in the second mitosis, i. e. that one passes to one pole and the 
other to the other pole of the spindle, just as the diplosomes do. It is 
further probable that the s-chromosomes like the diplosomes conjugate 
in the equator of the second maturation spindle before dividing there reduc- 
tionaUy, and for the follovring reason. In all the 179 polar \dews (except 
2) of second maturation spindles the condition was found represented in 
fig. 22 ; 6 autosomes and 1 bivalent diplosome (d, e, and this in niost cases 
in the centre). Now in 11,8% cases of fü'st spermatocytes there was 1 se- 
parate element more than the normal, therefore in the same percentage 
of cases there should be expected 1 more than the normal in the equator 
of the second spermatocytes; since this anticipation is not realized it is 
probable that in those second spermatocytes with s-chromosomes these 
conjugate to make a bivalent chromosome. This is the only satisfactory 
explanation of this numerical relation. The two exceptions mentioned 
among the 179 second spermatocytes were these: one with 8 separated ele- 
ments (fig. 4) due probably to the s-chromosomes failing to conjugate: and 
one case mth only 6, for which I have only the explanation that one 
chromosome may have been hidden beneath another. 
In testis no. 282 accordingly, the number of spermatogonial chromo- 
somes appears to be constant. But v.diile there are usually 8 separate 
elements in the first spermatocytes, in 11,8% of the cases 9 were found: 
this Variation is due to the precocious Separation, or failure to conjugate, 
of the 2 univalent components (s-chromosomes) of one bivalent autosome ; 
these s-chromosomes divide separately and equationally in the first matu- 
ration mitosis, and in the second mitosis in most cases conjugate in the 
plane of the equator and separate reductionaUy. Notwithstanding this 
Variation of behavior the number of univalent chromosomes in the first 
spermatocytes is 14, just as in the spermatogonia. 
In 24 spermatids the number of chromosomes was found to be in 
all cases 7. 
