[KIRSCH] ORIGIN AND DEVELOPMENT OF RESIN CANALS 53 
division. However, as their extension in a radial direction is rather 
limited they would keep on growing and proliferating in either a tangen- 
tial or a vertical direction or both. In this manner, vertical strands of 
medullary ray cells would be formed by the vertical proliferation of all 
or some of the elements of a ray in the cambial region of the tree. More- 
over, since medullary ray. are capable of arising at any time during the 
cambial activity of a tree, it must be taken for granted that they arise 
whenever the economy of the plant demands their presence. This 
would be either when there is a greater amount of food to conduct than 
the rays already present are able to handle, or when some of the older 
rays would for any reason become nonfunctional. If, however, the 
former is the cause for the appearance of new ray elements, these would 
follow the tendency of the other ray cells to extend vertically and would 
thus add to and also join the vertical strands of cells laid down by the 
rays already present. A strand of cells would thus spring into existence, 
which would be undergoing rapid growth and division in both length and 
breadth, being of greater or lesser extent as seen in transverse section, 
and extending in the longitudinal direction as long as the conditions 
which called it forth persist. 
Where such conditions are only local, viewing the tissue in trans- 
verse section, but still occur at points in the same tangential zone, there 
would be formed groups of parenchyma cells in contact with the medul- 
lary rays from which they originated, and forming a ring in the vascular 
cylinder, as was the case in the annual shoots of Pinus banksiana de- 
scribed above. As the same conditions might recur more than once 
during the same season, several such rings may be present. 
Where the necessary conditions for the initiation of these strands 
is fairly general, however, there would arise a more or less continuous 
tangential band of parenchyma, which, in extreme cases, would form 
a continuous ring, interrupted only by the elements of the rays from 
which they arose. Where such conditions, on the contrary, are purely 
local and desultory, the masses of vertical parenchyma would be found 
scattering through the whole annual ring, although occasionally tending 
to occur in the same zone. 
That the increase in the number of medullary ray elements wher- 
ever there is an abundant supply of food is not merely hypothetical, 
but supported by various observations is shown by the following 
examples :— 
Anderson (2, 337), describing the wood of the diseased branches 
of the witch broom of Abies balsamea, says as follows: “ Although the 
annual rings in the affected branches are as wide and often wider than 
the normal, they still contain fewer tracheids per square unit of surface 
area. ‘This is due to the fact that in the diseased branches more medul- 
