ZOOLOGY: E. B. WILSON 
323 
inal ring) which persist throughout the interkinesis and repeat the process 
in the second division, taking up a position near the spindle and dividing 
transversely at their middle points (hg. c). Each spermatid receives 
two of the final products, which give rise to a nebenkern consisting of 
two equal parts (fig. d), the axial filament growing out between them 
from a blepharoplast situated at the base of the nucleus. It thus comes 
to pass that each spermatozoon receives exactly two-eighths of the orig- 
inal ring (hence of the original chondrioma of the spermatocyte), the 
history of the chondriosome-ring suggesting in a general way that of a 
heterotype chromosome-ring, though its mode of division is very dif- 
ferent in detail. In the later stages, which are of great interest, the two 
halves of the nebenkern assume the form of flattened rods which rapidly 
elongate backwards, at the same time twisting spirally around the axial 
filam.ent (fig. e) until they finally extend throughout nearly the whole 
length of the flagellum and are converted into two very fine, closely 
twisted parallel fibrillae that form the spiral envelope of the tail. Each 
of these fibrillae, as is proved by its genesis, represents one-eighth of 
the original ring and hence of the spermatocyte-chondrioma. 
Nothing could be more surprising than the contrast to the foregoing 
mode of chondriosome-distribution seen in Opisthacanthus. It should 
first be emphasized that the two species agree closely in respect 
to the origin, staining reactions and ultimate fate of the chon- 
driosomes: they are derived from numerous minute chondrioconts ; 
they stain characteristically by Benda's method; they give rise to the 
nebenkern and ultimately to the tail-envelopes. A most striking dif- 
ference appears, however, in respect to the remaining phenomena which 
are in a general way in agreement with those described by Sokolow^ 
in the European species Euscorpius carpathicus, though some important 
differences are seen. In place of the ring appear approximately 24 
fairly large, separate, hollow spheroidal bodies scattered without dis- 
cernible order through the protoplasm. These bodies, which may be 
called chondriosome-spheres, do not at any time during the maturation- 
process show evidence of division nor do they enter into definite rela- 
tion with the spindle. Scattered irregularly about the latter (figs. /, g,) 
they are at each division segregated into two approximately equal 
groups, half of them passing into each daughter-cell; and during this 
process they seem to be quite inactive, being passively carried along 
in the protoplasm as the cell divides. Their total number is thus 
reduced approximately from 24 to 12 in the first division, and from 12 to 
6 in the second. Each spermatid thus receives six chondriosome-spheres 
as a rule, but sometimes five or seven; and these ultimately arrange 
