138 
Charles E. Allen 
includes fibers belonging to three Systems: (a) connecting fibers, running 
from pole to pole ; (b) mantle fibers, running from either pole to the chromo- 
somes, a bündle of fibers from each pole being attached to each chromo- 
some; and ( c ) a few short fibers which run from the poles, mostly in the 
general direction of the equatorial region, and seem to end freely in the 
eytoplasm, or rarely (Fig. 35) on the plasma membrane. In spindles 
of the later androgonial divisions (Figs. 38 — 40), as a result of the smaller 
amount of fibrous material, the grouping of fibers of dass (J) is not so 
noticeable, and those of dass (c) are fewer or entirely laeking. The spindle 
is always broad-poled, as is to be expected from the method of its forma- 
tion, although the fibers are commonlyslightlyconvergent toward the poles. 
I have not observed in any case, previous to the Separation of the 
daughter ehromosomes, a dustering of the connecting fibers (of dass a), 
such as was found by Timberlake at this stage in the root-tip cells of the 
onion and the pollen mother cells of the lareli. If such a grouping occurs, 
it is masked by the more eonspicuous dustering of the mantle fibers (dass 
b); and my observations here agree with those which I have previously 
made (Allen, 1903) on the pollen mother cells of Larix. On the other 
hand, as I shall describe later, there beeomes apparent during the anaphases 
(PI. VII, Figs. 50 — 53) a grouping of the connecting fibers exactly like that 
observed by Timberlake. 
Karyokinesis. 
The androgones of all generations are substantially alike as to the 
structure of their nuclei and the behavior of their nuclear substances 
during mitosis. A resting nucleus (PI. VI, Figs. 1 — 18) contains a single 
large, deeply-staining mass, sometimes rounded (Figs. 7, 19), but usually 
more or less irregulär in shape (Figs. 1 — 6, 8 — 18). Between this and the 
nuclear membrane is a sparse, irregulär retieulum, whose staining reactions 
indicate that it consists of chromatin and linin. It appears that in general 
the large central mass is not composed whollv of nueleolar material, but 
that its commonly irregulär outline results from the gathering of more or 
less chromatin around and incontaetwithanucleole. The proportion of the 
chromatin which is thus massed about the nucleole varies; but, as noted 
above, some chromatin is always scattered through the nuclear retieulum. 
In haematoxylin preparations the compound nature of the central 
mass is not commonly obvious, although sometimes, as in the cells shown 
in Figures 2, 13, 17 and 18, its extremely irregulär shape suggests that 
it is not a single body; and in other cases (e. g., the lower cell of Fig. 1), 
it is evident that small granules are elustered about a larger body, although 
