[kirsca] _ CERTAIN STRUCTURES IN THE PTERIDOPHYTES 373 
formation of a new tissue. The nuclei are very evident in these dividing 
cells, which fact probably means that they take an active part in the 
division as is the case in ordinary tissues. 
From this region to the base very little change was noted; the thy- 
loses in no case filled the canal completely, and remained thin-walled 
throughout. The secondary xylem exhibited walls of considerably 
greater thickness, but protoplasmic remains were still visible in the 
lumens of the constituent tracheids. Thus the whole length of the 
various stipes was still in a formative state. 
In stipes collected at the same time, but which exhibited much 
more vigorous growth, being 8 cm. broad, or about twice the diameter 
of the others, many interesting appearances were observable. 
The canals in the bundles are very large and extend deeply into 
the wood parenchyma groups (Photo 5). That there is a good deal of 
lysigenetic degeneration here in the formation of the canals is frequently 
evident from the manner in which the wood parenchyma cells are broken 
away and ruptured, and from the presence of free protoplasmic remains 
in the cavities. In this way the canal is no doubt enabled to attain its 
large dimensions. 
In many cases, where canals originate near each other in the 
bundle, in two or three places, the rupture of the wood parenchyma 
lining them leaves very thin partitions of cells between the cavities; 
or the parenchyma between them is entirely destroyed, thus causing the 
confluence of the two or more cavities. In Photo 5, the partition con- 
sists of about five layers of cells, in others there are only one or two 
layers, and in others again these layers are completely ruptured, the tora 
ends gradually disintegrating and disappearing, thus leaving a single 
large canal. Remains of protoxylem elements in the form of rings and 
more or less uncoiled spirals line the walls of the canal, and are scattered 
in the cavity. 
In these stipes various stages of thylose formation are represented 
in the canals of the same transverse section—from places where there 
is only the slightest trace of thylose formation as exhibited in the slight 
swelling of the cells bordering the canal (Photo 4), to places where the 
thyloses are very large and fill up the whole cavity of the canal (Photo 
6). The complete blocking up of the canal in transverse section in 
these stipes, is local, and not general as in the more mature stipes. This 
fact is proven by an examination of the same bundle in sections taken 
at different levels, when it is seen that at some levels the canal is com- 
pletely filled with thyloses, while at other levels it is only partly filled 
This difference is due to localised conditions of growth and activity. 
