1909] Sinnott,— Paracedroxylon, a new Araucarian Wood 167 
ber of pieces which were examined, however, appear to have been 
taken from the vicinity of one or more healed wounds. In such cases, 
the annual rings are much narrower than normally, and rows of very 
thick-walled, flattened cells oceur on the face of the summer wood. 
Here and there throughout the wounded area are distorted clusters of 
thin-walled cells, in rows at right angles to the rays. The structure of 
these cells is hard to make out, either in transverse or in longitudinal 
sections. They are probably parenchyma, though possibly very 
much modified tracheids. These groups of cells strongly suggest the 
clusters of parenchyma representing abortive resin-canals in certain 
of the Abietineac, notably Tsuga. They may possibly be open to the 
same interpretation. 
At two places observed, there was a band of wound-tissue at the 
annual ring (Fig. 7). In the transverse section, this is seen to begin 
as a narrow strip of cells which increases in width toward the apparent 
position of the wound, though no piece showing the actual wound was 
found. The tissue is composed of very thick-walled, much-pitted 
parenchyma. Where the band is wide, large mucilage-filled spaces 
appear in it, surrounded by the parenchyma. ‘hese cavities are 
sometimes nearly circular in cross-section, but are more often some- 
what flattened laterally. In the intervals between them, the rays 
cross the wound-tissue (Fig. 8). 
A longitudinal radial section exhibits these mucilage-filled spaces 
interrupted here and there by the passage of rays across them (Fig. 9). 
In this region, the horizontal and often the end walls of the ray-cells 
are much thickened and pitted, in strong contrast to their normal 
condition (Fig. 10). The mucilaginous contents of the spaces is often 
'acuolated, and contains here and there masses of material much 
darker than most of it. The wound-tissue is usually bounded, next 
the normal wood, by rows of septate tracheids (Fig. 11). 
A longitudinal tangential section through the wide part of the band 
of traumatic tissue, where these cavities appear, shows that they form 
a net-work of mucilage-spaces, separated only by the rays or groups 
of rays which cross them (Fig. 12). In the narrower part of the strip 
of wound-tissue, the cavities grow much smaller, cease to be connected, 
and finally disappear. 
The resemblance between these cavities and traumatic resin-canals, 
especially when seen in cross-section, is close, and as wound structures 
are often abnormal and monstrous, it seems safe to infer that the 
