635 
Growth and Structure of the Wood of Larix. 
sectioned at every few inches, the position of the principal branches being 
carefully noted at the same time. Such a method was precluded from the 
present investigation because sections had only been obtained at every 
fourth foot. On a future occasion I hope to throw some light on this 
interesting matter. 
(b) Anomalous Formation of Autumn Wood. 
In certain rings the defoliations have produced a striking anomaly in 
the autumn wood formation, traceable in every section from the top of the 
tree to the bottom. Such rings are denoted in the tables by the letter ‘a ’, 
and show as it were a tentative formation of autumn wood. There is first 
a zone of the rounded thickened cells that ordinarily occur in Coniferous 
timber as a transition between the thin-walled spring tracheides and the 
tangentially flattened, thick-walled ‘ Grenztracheiden ’ 1 that limit a normal 
annual ring, but then, instead of the typical f Grenztracheiden ’, these 
abnormal rings pass on to a renewed production of thin-walled cells to 
form their outer limit (Fig. 4, innermost ring). It is as though the tree 
had begun to make its autumn wood but the food supply ran short, and 
consequently the thickening of the cell-walls could not be continued right 
up to the boundary of the ring. The actual appearance of this thick- 
walled zone recalls somewhat the isolated bands of f mechanical ’ tissue 
commonly found in the spring wood of the young shoots of many Conifers. 2 
These bands have been referred to the effect of supposed alterations of 
pressures acting upon the cambium, and Schwarz, 3 as has been said, goes 
even so far as to regard the longitudinal stresses occasioned by bending as 
the main stimulus to the thickening of the cell-walls of normal autumn 
wood. If this view is correct, it is possible to account for the want of 
thickening in the walls of the outermost cells of these abnormal rings on the 
theory that after defoliation the wind could not sway the tree so much as 
before, and consequently the bending forces would be less. 
Besides these two suggestions, viz. that the increased proportion of 
autumn wood is due to lack of food, or to reduced pressure, or perhaps to 
a combination of both, there is a third possible factor to be considered. 
It has been mentioned that Strasburger, with Haberlandt and others, held 
that spring wood was formed just so long as the shooting of the young 
leaves required freish provision of water-conducting tissue. According to 
this view a second flush of leaves would mean a renewed formation of 
spring wood, and the reason why ‘ Lammas ’ shoots do not always cause 
a ‘ double ring * is that the second crop of leaves may burst out before 
autumn wood formation has begun. Strasburger 4 mentions the case of 
a twenty-year-old larch which once at least had formed a double ring after 
1 Cp. Hartig ( 1 ), Fig. 3, p. 12. 2 Cp. Sanio, p. 101. 
3 Schwarz, pp. 2 and 160. 4 Strasburger, p. 949. 
