424 
I. W. BAILEY 
of 5,000-10,000 cubic microns. The chromosomes in figure 12 appear to 
be longer than those in figures 4 and 15 because they are flat, hook-shaped, 
and visible in one focal plane, whereas those in the other figures are U- 
shaped, twisted, and in different focal planes. Of course, as admitted by 
Conklin, it is difficult to determine the volume of individual chromosomes 
with any considerable degree of accuracy, but comparisons between figures 
4, 7, and 15; 5, 13, and 16; 17, 28, and 30; and 18, 29, and 31 indicate 
very clearly- that the large, fusiform initials of the lateral meristem are not 
hyperchromatic. The striking variations in the size of the nuclei (figs. 10 
and 14) are due to differences in the volume of their achromatic portions. 
Thus, the volume of nucleolar matter is much greater in the large nuclei. 
Furthermore, the staining reactions of the "resting" nuclei suggest that 
the chromatin is more concentrated in the smaller than in the larger nuclei. 
In chromatin stains, the small nuclei of ray initials are very heavily over- 
stained long before the large nuclei of adjoining fusiform initials become 
clearly differentiated. 
It is evident, accordingly, that, although in certain cases variations in 
the volumes of cells are closely associated with fluctuations in the number 
of chromosomes (Boveri, Gates, Winkler) and in others with fluctuations 
in the size of chromosomes (Erdmann, Gregory, Keeble, Conklin, Hegner), 
the undifferentiated, actively dividing cells of the lateral meristem may vary 
greatly in size without corresponding variations in chromosomal size or 
number. 
Cytokinesis 
Certain of the fusiform initials in Coniferae are several hundred times 
as long as they are wide (radially), yet they divide longitudinally. What 
then is the nature of cytokinesis in cells of such extraordinary dimensions? 
During the telophase (fig. 17) the central spindle expands laterally by the 
addition of peripheral fibers and gradually assumes the form of a more or 
less warped disk (fig. 18). The connecting fibers, and later the accessory 
fibers, thicken to produce a cell plate in the usual manner, and then the 
fibers disappear except for a circular rim of kinoplasm. In tangential, 
longitudinal sections of the cambium, this ring-shaped aggregation of kino- 
plasmic fibrillae forms a halo about the daughter nuclei (fig. 56) The ring 
increases in circumference by the addition of new peripheral fibers and 
extends the cell plate as it does so. When it intersects the radial walls of 
the cell, it becomes more or less four-sided (fig. 57). As soon as the cell 
^ Beer and Arber (1915) reached the conclusion that binucleate cells are of common 
occurrence in the growing tissues of the higher plants. They state: "The nuclei of the 
multinucleate cells generally arise by mitosis, but there are certain exceptional features 
connected with this mitosis and with the behavior of the associated protoplasm. The 
most striking of these is that two daughter-nuclei in the telophase, between which no wall- 
formation is in progress, are often found enclosed in a hollow sphere of dense and deeply 
staining protoplasm, the appearance at first glance suggesting a cell within a cell." I 
strongly suspect that the phenomenon referred to by them is a phase of cytokinesis not 
unlike that illustrated in figure 56. 
