422 
I. W. BAILEY 
The karyokinetic figures in ray initials (figs. 15, 16, 30, 31, 32) generally 
resemble those that occur in parenchyma and terminal meristems. The 
figures vary considerably in fusiform initials, depending upon the shape 
and size of the nuclei and cells, and upon the plane in which the cells are 
dividing. Of course, the principal plane of division in elongated initials 
is periclinal or parallel to tangents to the circumference of the stem or root. 
In other words, the fusiform initials divide in a tangential, longitudinal 
plane which is a division plane of maximal area. Although the tangential 
diameter of the cambial initials increases to a certain extent during the 
earlier stages of enlargement of stems, roots, and branches (table i), it 
falls far short of being sufficient to compensate for the rapid increase in the 
periphery of the cambium. Nageli (1864) inferred from this fact that the 
fusiform initials must divide periodically in a radial, longitudinal plane. 
These hypothetical radial, longitudinal divisions are described and figured 
in many botanical textbooks, but I have been unable to find them in any of 
the gymnosperms and less highly specialized dicotyledons that I have 
studied. As shown by Robert Hartig (1895) and Klinken (191 4) — for 
Pinus sylvestris L. and Taxus haccata L. — the fusiform initials elongate, 
sliding by one another, until they have attained a certain length. They 
then divide, by means of a more or less oblique, transverse partition into 
two short cells, which in turn elongate and divide. Thus, the increase in 
the periphery of the cambium is due primarily, not to radial, longitudinal 
divisions of the fusiform initials, accompanied by lateral enlargements of 
the products of such divisions, but to the sliding growth of periodically 
elongating and dividing cells. During the process of elongation, between 
successive pseudo-transverse divisions, the initials continue to divide in a 
tangential, longitudinal plane. 
In dicotyledons {e.g., Robinia) having short initials and small nuclei, 
the polar axis of the karyokinetic figure in a longitudinally dividing cell is 
placed at right angles to the main axis of the cell (fig. 25). In Pinus 
Strohus and other gymnosperms, on the contrary, it tends to assume a 
diagonal position during the late prophase, metaphase, anaphase, and early 
telophase (figs. 4, 5, 17, 18, 19). That this is not an artifact, due to the 
displacement of an ordinary spindle during fixation or sectioning, is indicated 
by the fact that the whole mitotic figure is asymmetrically developed in 
conformity with its oblique position (figs. 17, 18, 19). Furthermore, it 
should be noted that the radial diameter of the initials is so short (fig. 50) 
that there is not sufficient room to permit the elongated nucleus to shift 
into, or the karyokinetic figure to develop in, a transverse position. Of 
course, the position of the mitotic figure in pseudo-transversely dividing 
fusiform initials (figs. 7, 8, 9, 11, 12, 13, 28, 29) cannot be due to such 
factors as these, but is closely correlated with the orientation of the division 
membrane. When the partition is exactly transverse, the polar axis of 
the karyokinetic figure is parallel to the long axis of the cell (figs. 7, 9, 28, 
