INVERTEBRATA, CRYPTOGAMIA, MICROSCOPY, ETC. 977 



granular. When tlic diflferentiation of the nucleus begins, sharply 

 defined round granules appear in it, the protoplasm becoming less 

 transparent as they increase in number. These granules have much 

 the appearance of starch-grains. They frequently collect at one side of 

 the cell or form a ring round the nucleus ; this ring or sphere after- 

 wards becomes broader, the granules retreating towards the cell-wall, 

 while the nucleus remains surrounded by a transparent and nearly 

 homogeneous protoj^lasm. 



In the final stage of the differentiation of the nucleus, the threads 

 form a dense convohition, the substance of which is obviously finely 

 granular. The line where they meet the periphery of the nucleus is 

 very fine, and apparently not smooth, but granular ; while in the centre 

 of the convolution the outlines of the threads are not clearly dis- 

 cernible. 



The next processes comprise the second stage of development, the 

 phase of the division of the cell-nucleus. The outlines of the threads 

 become sharjier and their substance apparently denser and more homo- 

 geneous ; and at the same time they begin to change their position, so 

 as to lie more or less nearly parallel to one another, the convolution 

 increasing at the same time in size, and assuming the form of a plate 

 composed of serpentine threads. Some among the cells are dis- 

 tinguished from the rest by their size and transparency, although their 

 nuclei have not more than the ordinary size. The threads now pre- 

 sent the appearance as if they were elastic, and constantly endeavouring 

 to free and straighten themselves ; at all events the convolution breaks 

 up eventually into small fragments. The nucleus is now practically 

 composed of a great number of separate filiform fragments, without 

 any special distinguishable intermediate matrix, the threads appearing 

 to be surrounded by the same transparent cell-contents which fill up 

 almost the entire cell-cavity. 



The division of the convolution, and the development of the two 

 new nuclei, usually proceed in the following way : — The plate, com- 

 posed of transverse coils of threads, becomes thicker, often passing 

 almost across the cell-cavity, and having sometimes somewhat of a 

 stellate form. Tlie tlireads at the same time break up into shorter 

 rod-like fragments, their outline becoming also less sharp. The split- 

 ting of the disk now commences at right angles to tlie direction of the 

 elements of the nucleus, beginning usually at the margin and advanc- 

 ing towards the centre. The two halves begin at once to separate 

 from one another, but remain for a time connected in the middle. 

 The intermediate space becomes filled with a very dense, opaque, and 

 usually granular protoplasm, in wliich a delicate striation is sometimes 

 to be seen. ]>ut tho elements of the two lialvcs of the nucleus possess 

 und(jubtedly from the first a certain polarity, that is, a tendency to 

 separate in t)pposite directions, as may be seen from the position taken 

 up by single free fragments. The two halves soon sej)arate completely 

 from one another, and whvu they have approaclied ch)so to tlie cell- 

 wall tliey are still composed of distiiiguislmblo rods or gi'anules. The 

 elements of the nucleus gradually fuse together, but its structure is 

 still apparently not homogeneous. 



