[KIRSCH] ORIGIN AND DEVELOPMENT OF RESIN CANALS 59 
the surrounding tissues, and not on any special cells given off from 
vertical passages. 
Mayr’s view in regard to the origin of the horizontal passages has 
led him into evident error, as can be seen from the following statement 
(19, 216): “Since the horizontal passages always run uninterruptedly 
to the bast region, they naturally often encounter newly forming verti- 
cal canals in the cambial region; if they strike these tangentially there 
then ensues an open communication between the horizontal and vertical 
passages, such as Hartig (10, 137) has described and figured. If the 
axis of the horizontal canal strikes the axis of the vertical passage at 
right angles, the former will divide the latter into two and thus open 
communication between the lumina of the two canals ensues, to which 
I shall refer later on.”” Mayr thus seems to forget that the rays run 
continuously from the bast to the wood, and, on the increase in dia- 
meter of these tissues, elongate either passively by the radial extension 
of their already formed elements, or by the intercalation of 
new elements between the bast ray and the wood ray in the cambial — 
region. It would thus be impossible for a ray to “strike” a newly 
forming vertical strand at right angles, unless that vertical strand was 
forming inside the ray, and in this case it can be seen that an entirely 
different state of affairs would prevail in regard to the origin of the 
passages from each other than that assumed by Mayr. If, again the 
ray cells in the cambial region instead of proliferating equally on both 
sides tended to do so more on one side, or if additional parenchyma 
elements were laid down by the cambial tissue on one side of a ray only, 
a case would present itself where the medullary ray canal would seem- 
ingly strike a vertical canal tangentially. Both of these cases are illus- 
trated in Fig. 20, A and B. 
These drawings were taken from sections of Pinus banksiana 
obtained from a vigorous terminal shoot of a young tree, and are views 
of the same ray at slightly different levels. Fig. 20, A shows two 
canals present in the wood ray, one in the region of the protoxylem, 
whilst the other is situated in the centre of the ray near the cambial 
region. In the bast the ray is seen to take on a knob-shaped appear- 
ance, which is due to the fact that a resin cyst is formed here, the lumen 
of which is seen clearly in Fig. 20, B. In the latter, however, the 
canal near the cambial region, which was central to the ray in A, is now 
seen situated to one side, this fact being probably due to a local eccen- 
tricity of growth in the ray elements causing a proliferation on one side 
of the ray only. 
Fig. 21, taken from a section of the same nature as that from 
which Fig. 20 was obtained exhibits an interesting feature. It is here 
seen that a multiseriate ray which extended from the phloem to the 
