The nuclear componeuts of the sex cells of four species of cockroaches. 511 
As metaphase gives place to anaphase, the lugs become pulled out 
(I, fj while the horizontal parts of the ring become smaller and 
smaller until the ring as such has disappeared entirely and given 
rise to two sets of daughter chromosomes joined end to end. It will 
be seen that this division is a longitudinal one, separating daughter 
chromosomes in au equational division. The second spermatocyte 
division is reductional (I, g) separating elements originally conjugated 
in synapsis. 
The description of H. S. Davis is very different in detail frorn 
that just given, but the end result is the same. Here the rings 
appear superficially identical with those of the foregoing account, 
but in reality, as Davis believes, they are constructed on a different 
plan. The ring (II, d) is virtually a loop formed from an originally 
longitudinally split bivalent chromosome (II, c), (he free ends not being 
united as in the preceding case, but crossed upon one another. The 
extending portions of the ends of the chromosomes forming the ring 
may be termed the lugs. The position of the ring in the spindle is 
vertical and not horizontal as in Suttox’s case while the lugs lie 
parallel to the plane of the ring proper. Düring anaphase (II, e), 
the lugs are pulled out so that ultimately a reductional division is 
effected (II, f) by the Separation of the chromosomes joined end to 
end in synapsis. The second division (II, g) is an equational one 
and consequently longitudinal. Hence there is a reversal of sequence 
of the two methods of Separation of the daughter chromosomes in 
Davis's description, as compared with that of Sutton, although the 
end result is the same. 
It was stated above that the descriptiou of the Schreiner’s for 
Tomopteris is essentially similar to that given for the species consi- 
dered in the present paper. For this reason a brief account will be 
given of the\r observations (text-figures I, v). 
In their material, terminal attachment of the chromosomes to 
the spindle fibres never occurs. The rings, unlike those of Periplanetci 
are formed in the prophase (IV, c) by the opening out of the daughter 
chromosomes upon one another along the line of the longitudinal 
split (IV, b), which represents the line of the spermatogonial chro- 
mosomes in synapsis (IV, b). The rings lie in the spindle vertically, 
the lugs being horizontal in the equatorial plane. The first division 
is exactly the same as in Periplaneta (IV, f) and is identified by the 
Schreixer’s as reductional, the second being equational (IV, g). 
Figure IV, e, represents a chromosome with sub-terminal attachment 
Archiv f. Zellforschung. III. 34 
