Meiohc Nuclear Divisions of Galtonia candicans . 731 
it is soft and cloudy and finely reticulate; with acetic alcohol it is more 
coarsely reticulate; with Hermann it is stringy and becomes contracted in 
parts; with strong and medium chrom-acetic it contracts into dense 
flakes. Preparations fixed with strong Flemming and Hermann show 
indefinite areas of a dense deposit in the cytoplasm (Fig. 12). In some 
cases these areas stretch across the walls from one cell to another. Probably 
they consist of some oily storage material; they are not found in pre¬ 
parations fixed with alcohol and glacial acetic or with strong and medium 
chrom-acetic. 
To return to the nuclear framework. 
As the chromosomes break up, their fragments are seen to flatten out 
into viscous portions of linin (Figs. 6 A, 18, 19, 20, 21), in which the 
chromatin is carried in a diffuse state. In these nuclei there is no 
differentiation of linin areas and chromatin discs. The chromatic linin 
portions assume all manner of shapes and sizes, and strands are joined 
together by means of finely drawn out threads from the viscous substance. 
These threads persist throughout the prophases, and finally join the 
separate chromosomes. If it is remembered that the nuclear framework is 
of a colloidal nature, it is easy to imagine the variety of forms and shapes 
that might result. As the chromosome segments flatten out they may 
become alveolized; the central portion dissolves, leaving the sides as 
parallel threads (Fig. 21); or large alveolar windows may appear in the 
linin (Fig. 19) ; or the linin may break up into small rounded particles 
which may lie parallel to one another (Figs. 6 A and 21). Whatever form 
the linin takes it is seen to stain uniformly. From this it may be con¬ 
cluded that it is impregnated with an evenly distributed solution of 
chromatin. In some nuclei pieces of homogeneous looking chromosomes 
may still be present, whilst the other chromosomes may have already 
broken up and be indistinguishable in the nuclear framework (Fig. 21). In 
the roots of Allium Gregoire ( 9 ) (1906) has shown that each chromosome 
resolves itself into two parallel lines, composed of threads or granules. 
In the nuclei of the roots of Galtonia this parallelism is to be seen in parts 
of the linin, though in not nearly so diagrammatic nor in so regular 
a manner as in many plants. In the nuclei of the archesporium the linin 
is in a much more homogeneous and lumpy condition and shows com¬ 
paratively little trace of parallelism. The linin breaks up into still 
smaller and smaller pieces, the parallel threads divaricate, and an irregular 
kind of meshwork results. The linin framework is at first more or less 
distributed throughout the nucleus, leaving a clear space round the 
nucleolus (Fig. 21). The space becomes more emphasized as the nucleus 
approaches the so-called ‘resting’ stage (Fig. 6 b). Possibly the nucleolus 
may be exuding substances into the nuclear sap which effectually prevent 
any of the nuclear contents from encroaching on its vicinity. As the 
