e Heterotypical Reduction ' in Somatic Cells. 787 
the ‘protoplast’ (Fig. 1). Fig. 17 shows, however, that later, the two 
nuclei of a tetraploid cell generally lie widely apart in the latter, as though 
repelling each other, in accordance with Gerassimow’s account of the 
behaviour of two healthy nuclei when occurring in one cell. In connexion 
with Gerassimow’s hypothesis that such repulsion is due to an ‘ electrical 
force \ and also with the general problem of the nature of nuclear activity, 
it is of particular interest to find that at certain stages of division two 
nuclei will fuse together, instead of repelling each other. This is seen in 
the telophase shown in Fig. 21. Here the contiguous telogroups of two 
diasters occurring in one cell have amalgamated to form a large central 
daughter nucleus ; there is no crushing together of these two groups and 
a clear connexion can be traced between their respective networks of linin. 
Figs. 17-19 illustrate another point of interest in this connexion, 
namely the close similarity between the states of activity of the two nuclei 
in a tetraploid cell. In all cases where these nuclei present a normal 
appearance, they are at precisely the same stage, their spiremes, diasters, or 
telophases showing practically identical features. Where any difference is 
discernible between them, one of the nuclei without exception exhibits 
signs of degeneration. This similarity of activity seems to suggest that the 
various phases of rest and division, through which a nucleus passes, are 
initiated by some stimulus resulting from its interaction with the cyto¬ 
plasm. 
Figs. 1, 12, and 13 show, in agreement with Hertwig’s theory of a 
* Kernplasmacoefficient ’, the existence of a certain correlation between the 
size of a cell and the amount of its contained chromatin. As was noted 
above, after an interval of some hours the tetraploid cells, at first no larger 
than those containing one nucleus, show a considerable increase in size. The 
recovery by the tetraploid nuclei of their normal appearance, and presumably 
of their healthy activity, seems indeed to be invariably accompanied by 
marked growth of the cell, which may attain fully twice its usual dimensions. 
The behaviour of the nucleoli after treatment of the root with chloral 
hydrate requires comment. At the earlier stages of the experiment 
these appear to fragment into innumerable minute particles, which are 
scattered through nucleus and cytoplasm, a condition which possibly 
indicates the occurrence of rapid excretion. Later, they very generally 
divide into two or more bodies (it should be noted that this condition 
occurs, though to a less extent, in normal roots), and the shape of some of 
the nuclei containing them, which tends to be of a dumb-bell character, 
suggests that amitotic division may perhaps take place at this stage, 
although it probably does not do so earlier, or as a direct response to 
the action of the poison. Although, however, the occurrence, at this period, 
of numbers of cells containing two small nuclei of rounded outline might 
support such a suggestion, the data are insufficient for a decision on this 
