[KIRSCH] ORIGIN AND DEVELOPMENT OF RESIN CANALS 51 
Essner (6), in his investigation of the number and height of the 
medullaryrays among the Coniferæ, came ‘to the following conclusions :— 
1. The number of the medullary rays differs with the age of the 
plant. They are most numerous in the first year of growth, then 
become less numerous for a number of years, and then for a certain 
period begin to increase again. In one and the same growth ring the 
number decreases with increasing height. 
2. Like the number, the height of the medullary rays also stands 
in certain relations to the age of the plant. In the first growth ring, 
where they are most numerous, the shorter ones predominate; with 
increasing age the rays of greater height gain the ascendancy. How- 
ever, even different individuals of the same species show great inequali- 
ties in regard to the number of the medullary rays, and so Essner con- 
cluded that the number of rays, affords no point for the determination 
of Coniferous woods. From a consideration of our results, however, 
his failure to secure any data of diagnostic value in regard to the number 
and height of the rays is not at all surprising, as these characters 
depend almost entirely on the vigor of growth, and may approach 
closer in individuals of different species but of equal vitality, than in 
individuals of the same species which present considerable differences 
in their vigor of growth. 
In many cases, however, another phenomenon is bound to make 
its appearance under certain conditions. Where the ray cells are very 
active and there is an abundance of food available, they would be 
capable of keeping pace with the elongating elements in two ways, as 
the following will demonstrate. Where the rows of ray cells are being 
torn apart at their horizontal walls, the central cells, by a rapid division 
could afford a relief of tension and, in this manner, give rise to a structure 
as illustrated by Figs. 6 and 7. Both of these drawings were obtained 
from a radial section through a portion of a second year shoot of Pinus 
banksiana on which the shoots described in lot one grew. Fig. 6 shows 
how the central cells of the ray in the phloem region have divided into 
a considerable number of rather high narrow cells. The central cells, 
however, do not divide in a longitudinal direction only, but also undergo 
division in a tangential direction, by means of a radial walls, and hence 
the ray becomes multiseriate in the central region, as seen in tangential 
section. The marginal cells, on the other hand, acting as the medium 
by means of which the central cells are being torn apart in the rapid 
extension of the vertical elements, usually elongate and may even 
divide by means of cross walls, but do not, as a rule, divide tangentially. 
In this manner the ray receives its characteristic fusiform appearance. 
From the above it is evident that the rapid division of the central ray 
cells is induced not only by the large, and in many cases, unusual amount 
