nennen 
ON THE STEM OF DAMMARA ROBUSTA. 443 
rendered so brittle and friable by decay, by the thinning and 
partial removal of the woody prosenchyma, that fracture ensues 
by its own weight, or from the first slight violence that befalls 
it. Long before this occurs, the cortical parenchyma of the dead 
part is completely demarcated from that of the living; this is 
effected by transverse division of the parenchymatous cells and 
the formation of an intervening plane of cork. In this process, 
also, the cambium readily participates, its cells undergoing trans- 
verse subdivision to produce phellogen and cork,—it plays the 
part of a cork cambium under the altered relations of the parts, 
as readily as it before served for the production of wood and 
bast. After the complete detachment of the dead part, a ring of 
cork-covered callus advances from the cambium over the fractured 
surface of the dead wood, and ultimately covers it completely in, 
though the new tissue, which becomes differentiated into wood 
and an external cortical system, contracts no adhesion with the 
dead surface. 
This short account may suffice to contrast these cases with 
those in which branches are detached by disarticulation. No 
structural provision exists in the one case whereby a dead part 
may be detached; the provision which allows of disarticulation 
in the other case differs slightly in different examples. 
In Dammara robusta a longitudinal section (Pl. XVII. figs. 1, 
2) displays at once the means by which this ready separation or 
disarticulation is allowed. The base of the branch presents a 
marked enlargement, due almost solely to an increase of the cor- 
tical parenchyma ; this excess serves to supplement the wood, in 
this situation, in supporting the branch ; the cortical parenchyma 
generally and the medulla as well contain a considerable proportion 
of branching sclerenchymatous idioblasts. As regards the wood, 
the section displays an extreme thinning of this in the base of 
the branch; that is to say, where the wood of the branch is 
continued into that of the stem. 
This disproportionate tenuity of the wood is demonstrab | 
In transverse sections, by means of which it appears that, associated 
With the retardation of woody growth in this situation, the 
primitive condition, in which the fibro-vascular bundles are dis- 
sociated, is scarcely passed. This primitive disposition of the 
wood at the base of the branch and the associated undue propor- 
tion of parenchyma is especially well seen at the branch-junctions 
in Populus tremula, where on stripping off the cortex, after 
le also 
