THE CAMBIUM AND ITS DERIVATIVE TISSUES 
421 
originated. Similar phenomena have been observed by Gregory (191 2), 
Winkler (19 16), and others. The work of Gregory (1909) and Keeble 
(1912) indicates, however, that in certain giant forms of Primula sinensis, 
which have larger cells than the typical form, there is an increase in the size, 
but not in the number of chromosomes. Winkler's paper is particularly 
significant, in this connection, owing to his general discussion of the relation 
between cell size and chromosomal number in the higher plants. He reaches 
the conclusion that there is a very close correlation between cell size and 
chromosomal mass, both in meristematic and non-meristematic, somatic 
tissue. He states that in embryonic types of tissue (lateral and terminal 
meristems) the cells are roughly isodiametric, of nearly uniform size, and 
always contain the diploid number of chromosomes. In non-meristematic 
tissue, multinucleate protoplasts, nuclear fusions, and changes from the 
diploid to the tetraploid or the polyploid condition are of common occur- 
rence, and many cells depart widely from the inherited, specific cell size 
of the plant. He infers that such cells tend to be hyperchromatic, much 
elongated elements containing more than one nucleus each and other types 
of large cells an abnormal number of chromosomes. 
It is evident that the cambium provides a favorable medium for testing 
Winkler's generalizations. The fusiform initials, which vary greatly in size, 
obviously are not multinucleate. Are they hyperchromatic? Before at- 
tempting to answer this question it is essential to devote some attention 
to a discussion of karyokinesis in the lateral meristem. 
Karyokinesis in Cambial Initials 
As far as I have been able to determine, the nuclei of cambial initials 
always divide mitotically. Strasburger (1891) observed fragmentation in 
young sieve tubes of larch. My own preparations indicate that amitosis 
may occur in slightly differentiated cells of the cambial layer — which later 
are to develop into tracheary elements — just before the beginning of the 
growing season (fig. 38). Vacuolated nuclei also tend to be present at this 
time (fig. 38). 1 However, I have never seen any indication of either of 
these phenomena in undoubted initials of the lateral meristem of either 
gymnosperms or dicotyledons. 
The cambium varies so greatly in its activities during different seasons, 
in different plants, and in different parts of given individuals, that the only 
rules which I have been able to formulate for obtaining division figures are, 
to collect specimens at frequent intervals, from as many individuals as 
possible, and from all parts of each plant. Initials may be dividing actively 
in one portion of a stem when those in adjacent portions are inactive. 
Similarly, numerous mitotic figures may be present in a given individual 
when a neighboring plant appears to be entirely devoid of them. 
^ I am not entirely convinced that the phenomena observed by Strasburger and myself 
are not artifacts produced by poor fixation. 
