194 BOTANICAL GAZETTE [MARCH 
or double thread, but that of a band composed of a series of 
chromatin masses joined together in rings, each ring being sepa- 
rable into four primary masses of chromatin. Later, at the time 
of the formation of the equatorial plate and when the chromatin 
has reached its maximum size, the four masses have by their 
growth fused with one another into a ring, the succession of 
rings resulting in the formation of a more or less continuous 
tube. In figs. 73 and 76, longitudinal sections of periblem cells 
giving surface views of the coils in the nucleus, some of the 
loops have been cut transversely and thus give end views of the 
spireme. Views of both the ends and sides of the threads are 
of course necessary for the formation of an opinion as to the 
real condition of the chromatic substance at this stage. From 
a comparison of such views it is evident that the thread is now a 
tubular structure, the walls being composed of the four more.or 
less completely fused elements of the quadripartite stage. The 
more completely fused the chromatin masses, the straighter and 
more sharply defined is the edge of the spireme thread. Its 
contour varies with the age of the spireme and with the nature 
of the cell. In many cases (jig. zo) the spireme appears as 4 
double homogeneous thread with clean-cut edge. In other cases 
(fig. 9) it has a moniliform appearance, the ring structure being 
more evident as the component chromatin masses are further 
apart. In the cross-section of the cell shown in fig. 7 all the 
sections of the coil are transverse to the direction of the spireme 
thread. Groups of the enlarging tetrads can be seen joined to 
one another by the connecting linin substance. When the chro- 
matin is not overstained and a surface view of the spireme is 
obtained, its appearance is that of a longitudinally split thread, 
the split, however, never being sharply defined. /igs. 17 and 16 
show this apparent longitudinal splitting, while in figs. 9 and 14 
the appearance, instead of resembling a split, is like a series of 
separate vacuoles. The apparent longitudinal splitting results 
from looking down upon the succession of rings formed by the 
incomplete fusion of the tetrads. The rings are less dense at the 
junction of the component masses, so that a surface view is that 
of two dense chromatin masses joined by the less dense linin ; 
