Nov., 1923] 
BAILEY-CAMBIUM AND DERIVATIVE TISSUES 
503 
of the cambium cannot be due to the elongation of transversely dividing 
cells; for, if it were, the superimposed initials must necessarily crowd by one 
another and ultimately break up the stratified arrangement. Conversely, 
if the anticlinal divisions are radio-longitudinal, the products of successive 
divisions should be grouped in horizontal rows, unless this arrangement is 
modified by differences in the elongation of adjacent elements. In non- 
stratified cambia, figures 1-3, the adjacent, overlapping fusiform initials 
vary considerably in length, but, as the writer has previously stated, the 
average length of these elements does not increase appreciably during the 
later stages of the enlargement of a stem or root. Thus, the increase in 
the number of fusiform initials in non-stratified meristems cannot be due 
solely to radio-longitudinal divisions; for, if it were, there would have to be 
a general increase in the length of the initials during all stages of the enlarge¬ 
ment of the plant. 
Such facts as these suggest that there are two fundamentally different 
types of meristematic activity—so far as the fusiform initials are concerned 
—in the vascular plants; i.e., that the increase in girth of non-stratified 
cambia is due to the elongation of transversely dividing initials and that of 
stratified cambia to radio-longitudinal divisions of these elements. Detailed 
investigations of the cambium and of the histology of its derivative tissues, 
in numerous representatives of the gymnosperms and dicotyledons, strongly 
support such an assumption. 
The writer has already referred to the fact that in all of the lateral 
meristems of conifers and less specialized dicotyledons investigated by him— 
which are, of course, of the non-stratified type—the anticlinal divisions are 
transverse or oblique (figures 8, 9). The orientation of the cell plates and 
recently formed anticlinal partitions in the stratified meristems of Robinia 
Pseudo-Acacia L. and of Diospyros virginiana L., on the contrary, is radio¬ 
longitudinal (figures 6, 7). 
Furthermore, since the fusiform initials leave a record of their activities 
in the xylem and phloem, it is possible to trace successive stages of their 
growth and division in serial, tangential sections of these tissues. 1 Although 
this indirect method of studying cambial activity must be applied with a 
considerable degree of caution, particularly in the case of dicotyledons, it 
enables one to investigate a large number of plants of which suitable material 
of the lateral meristem is not readily available. In all of the various genera 
(62) of gymnosperms and dicotyledons that the writer has examined in this 
way, the evidence indicates very clearly that the anticlinal divisions in non- 
stratified meristems are pseudo-transverse, whereas those in stratified 
cambia are radio-longitudinal. In meristems of the former type, the fusi¬ 
form initials elongate, sliding by one another, until they attain a certain 
size. They then divide by means of a more or less oblique partition into 
1 For detailed descriptions of this method, the reader is referred to Klinken’s and 
Neeff’s papers. 
34 
