4 
PROFESSOR W. C. WILLIAMSON ON THE ORGANIZATION 
to detect any true walls enclosing them. They appear to be intercellular spaces, and 
to have served as gum-canals. In one longitudinal section only (fig. 6), in which a very 
small example of one of these canals accompanies a vascular bundle, the whole being 
magnified 65 diameters, the canal (c) appears as if it had originated in a row of enor- 
mously enlarged and attenuated fusiform cells which overlap each other at their oblique 
extremities. Since this is the only example of the kind which I have met with, I will 
not venture to affirm that all the canals have originated in the same way. In the young 
twigs these canals have a diameter varying from '016 to *004, being usually much larger 
than in the more matured petiole (fig. 1), in which they rarely exceed '01 in diameter. 
In another very fine example of a matured rachis, for which I am indebted to Captain 
John Aitken, of Bacup, they are very much smaller (fig. 7, c), their diameter ranging 
between '004 and '0025. In another somewhat matured rachis, a portion of a transverse 
section of which is represented in fig. 9, the canals (fig. 9, c) are about '007 to '004, 
approximating to the diameter of those of the young twigs (figs. 3 & 4). It thus 
appears that these canals vary in their dimensions in different specimens ; but I cannot 
discover any indication that such variations have any specific value. 
The vascular bundles may next be examined. These vary both in different specimens 
and to some extent in different parts of the same specimen. In fig. 1 we can readily 
discern that these bundles are so arranged as to form a certain pattern or design. There 
is obviously a peripheral ring of them indicated by the letters d, whilst others occupying 
the more central parts of the section appear less regularly arranged ; nevertheless a 
second irregular circular series may be traced, reminding us of what exists in certain 
recent ferns to which I shall call attention. This disposition to form a pattern is a 
feature that does not occur in endogenous plants. Each bundle consists of a cluster of 
vessels of various sizes, as seen in figs. 11 & 14, which represent two of the bundles of 
fig. 1. In fig. 11, for instance, the vessels ( d ') are compressed and somewhat deranged in 
position, and are in close association with an enlarged gum-canal (d), the entire cluster, 
including the canal, being invested by an imperfect sheath ( e ) formed of small cells. 
In fig. 11 we have one (d) of the smaller peripheral clusters, containing but three or 
four vessels (d). 
The Vessels. — Fig. 7 represents a fine cluster of these as seen in a transverse section 
made from Captain Aitken’s specimen. In this example the larger vessels (d) compress 
each other so slightly that they retain much of their cylindrical form, the intervals 
between them being occupied by small cells. The largest of these vessels has a diameter 
of about '01, and the smallest of about '0012. The larger vessels are always aggregated 
on one side of the bundle, and the smaller ones (d!) are clustered together much more 
compactly on its opposite side. Fig. 9, d represents part of a section with two bundles 
from another specimen in which the vessels are much fewer in number, and the very 
small ones seen at fig. 7, a l 1 are almost wholly wanting. In some instances I don’t find 
more than two or three large vessels and one or two small ones. In the young rachides 
(figs. 3, 4, & 4*) it is almost impossible to discover these bundles in the transverse 
