Nuclear Divisions in the Rusts. 
345 
broad, plate-like centrosomes. A somewhat later stage is shown in Fig. 41. 
The nucleoles now lie some distance from their respective spindles. The 
left nucleus in this figure is particularly well adapted to show the relation 
which the dense central spindle sustains to the chromatic mass aggregated 
on its curved side. Mantle fibres are here not well differentiated ; but 
it seems quite likely that they are represented by the few dense filamentous 
structures which curve from the poles out into the adjacent mass of chromatin. 
Fig. 15 shows a somewhat similar stage of conjugate division. The 
two nucleoles are quite remote from their nuclei, and they show as yet 
no particular signs of disintegration. The double base of this figure sug- 
gests that this is a ‘ fusion cell ’, but there are no remnants of a partition 
wall left to show how such a fusion took place. 
While unsatisfactory as to the chromatin, Fig. 16 shows with remark- 
able clearness the sharply-delimited spindles. There can be no doubt that 
here the two spindles are separate and distinct structures, and that they do 
not fuse into one during division, as was claimed for conjugate divisions by 
Poirault and Raciborski. The chromatin in this instance has apparently 
been drawn almost entirely into the central spindle, and it is now in early 
anaphase condition. The two nucleoles lie near the equator of their 
respective spindles and are in an advanced stage of dissolution, as is 
indicated by the large vacuole in each. 
While it is sometimes difficult to distinguish the central spindle in the 
late anaphases, the mantle fibres, on the other hand, stand out sharply. 
A good part of the fibrous structures which appear between the receding 
centres in Fig. 1 7 probably should be referred to the mantle fibres, since 
they are chromatic in their staining reactions. In Fig. 18, also, the con- 
spicuous fibres are probably mantle fibres, and the few dim structures, 
which still persist between the daughter-nuclei, doubtless represent what is 
left of the central spindle. In many other instances, however, these late 
stages are characterized by the persistence of a quite conspicuous central 
spindle. The attenuated structures in Figs. 20 and 21, for example, which 
connect the receding daughter-nuclei are plainly central spindles, and the 
mantle fibre portion in both cases has contracted almost to the poles. 
The frequent presence of asters at the poles during these later stages 
(Figs. 17-19) undoubtedly indicates a totally different sort of polar activity 
than that which characterizes the earlier stages, since in the latter they are 
not noticeable see Figs. (13-16). 
The reconstruction of the daughter-nuclei in these forms has already 
been described for the gametophytic hyphae, and the process in the 
binucleated cells is essentially similar. Fig. 19 illustrates quite clearly 
the phenomenon. The chromatin of each nucleus lies in a clear space, and 
the strands are arranged with some regularity. One end of each strand is 
still attached to the centre, while the other apparently ends free in the 
