195 
certain Varietal and Hybrid Ferns . 
times a few threads may diverge towards one of them here and there. 
In this respect they are to be contrasted with what occurs in many other 
plants, e. g. Lilium , E guise turn , &c. 
As the later stages of the division in this species were not traceable in 
sufficient detail to enable us to add materially to our knowledge of the pro¬ 
cess, we shall pass on to the consideration of the other plants. 
Polypodium vulgare (Type form). 
It was desirable to examine this plant, inasmuch as its variety elegan- 
tissimum is one of the parents (or putative parents) of the hybrid. We have 
already remarked on the large number (about 90) of the haploid chromo¬ 
somes in this species. In so far as the development of these bodies from 
the resting nucleus is concerned the process is similar to what obtains 
in P. aureum , but as the larger size of the nucleus (cf. PI. XVI, Figs. 4 and 
10) is more than compensated by the number of the chromosomes them¬ 
selves, the details of chromosome-formation were not satisfactorily followed. 
But the character of the spindle is very different in the two plants, and 
it also exhibits unexpected abnormalities which recur in the varietal form, 
as well as in the hybrid. 
The granules in the cytoplasm, already described for P. aureum , are 
also formed here, and some at any rate are derived from the chromatic 
constituents of the nucleus. They are seen to be budded off from the 
nuclear contents into the cytoplasm just as has been described by one of us 
for Galtonia d 
As the nucleus of the spore mother-cell enters on the stage of synapsis, 
the kinoplasmic nuclear sheath is easily demonstrable, but instead of the 
bipolar type of spindle a quadripolar (or multipolar) structure (PI. XVI, 
Figs. 7, 8) is differentiated. 2 The arms (nearly always four in number) are 
very prominent objects in the cytoplasm, and they persist as well-marked 
cones which extend from the nucleus to the periphery of the cell. It is not 
till diakinesis is passing over and the nuclear wall breaks down that the 
quadripolar arrangement merges into the bipolar one by the fusion and 
concentration of the poles in pairs (PI. XVI, Fig, 9). 
There are certain features connected with the nuclear contents that are 
important in relation to the first differentiation of the quadripolar spindle. 
As the synaptic tangle passes into the more open thread stage which always 
immediately succeeds it, the chromatic linin is seen to be first less evenly 
distributed within the nucleus than is more often the case in plants. On 
the contrary, it is, though rather obscurely, massed at four spots about equi¬ 
distant from each other just within the nuclear wall. As soon as this becomes 
1 See V. Derschau, Max, Beitrage zur pflanzlichen Mitose : Centren, Blepharoplasten. Prings- 
heim’s Jahrb. f. wiss. Bot., xlvi. Digby, L., Observations on ‘ Chromatin Bodies’, &c. Annals of 
Botany, xxiii. 
2 Cf. Lawson, A. A., Studies in spindle formation. Bot. Gaz., xxxvi. 
