Notes . 
395 
that these granules, with their curious relations to the spindle-fibres, 
were regular and normal constituents of the cell during these stages of 
division, I learnt with pleasure from my friend, Mr. J. E. S. Moore, that 
in the course of his investigations into the behaviour of dividing cells 
in animals, he had come across a similar case of multipolar spindles 
in Branchippus . His results are in course of publication, but he kindly 
invited me to see his preparations, and they agreed with those of Lilium 
in their essential characters so far as the relations of the spindle are 
concerned. Of course multipolar spindles in plants are not new, but 
those hitherto described 1 are of a different nature from those which 
occur in the Lily anthers. Moreover they are found in endosperm 
cells, whose nuclear constitution appears to vary within considerable 
limits, at least judging from the existing accounts before us. In such 
a cell-division as that which results in the production of a spore, it 
seems, however, possible that their occurrence may be of some special 
significance, but I reserve the discussion of this point for a future 
occasion. 
Hitherto I have only described the cells when in a state of active 
division, but when these stages are compared with earlier ones, several 
other features of interest become prominent. The cytoplasm up to 
the time immediately preceding the disappearance of the nuclear 
membrane, and the aggregation of the chromosomes in the equatorial 
plane, is perfectly free from stainable granules. The nucleus, when 
preparing for division, exhibits a much convoluted thread with rows of 
dots, which stain deeply, running along the edges, exactly as is 
beautifully shown in Fig. io, accompanying Guignard’s memoir 
already referred to. The nucleolus is of enormous size, and contains 
usually several endonucleoli. At a stage somewhat later, when the 
chromosomes have become individualized, and are lying irregularly 
disposed within the nucleus, the nucleolus almost always assumes 
a most curious and characteristic shape, that of an oval body with two 
polar protuberances (Fig. A ). Frequently it lies close to the cytoplasm, 
one of its pointed ends actually appears to jut out into it ; sometimes, 
however, instead of one large nucleolus, stages in its fragmentation 
may be observed, and a number of smaller bodies, which present 
precisely the same staining-capacity take its place. Their aggregate 
size, however, is equivalent to that of the large single nucleolus. 
1 Cf. Strasburger, Histologische Beitrage, Heft-i, 1 888, especially Taf. Ill, 
Fig. 34, and the text. 
