396 ROYAL SOCIETY OF CANADA 
side of the bundle and contain mucilage or tannin. A few of these cells 
are, however, found on the outer side of the bundle as well. 
Alternating with these on the inner side of the bundle are groups 
of protoxylem which are fairly thin-walled, but empty of contents 
(Photo 16). Most of the protoxylem groups at this stage were intact, 
but some of them were starting to break up, and thus formed small 
spaces which represent the beginnings of the canals. The secondary 
xylem in these sections was not yet formed, and its cells showed an 
abundance of protoplasmic material (Photo 16, xy). 
Sections obtained from a mature rachis of Todea barbara showed 
the thyloses beautifully. The protoxylem groups seen in the young 
rachis were not present at any point, and their position is occupied by 
large, thick-walled thyloses (Photo 17). The thyloses frequently con- 
tain masses of protoplasm in which nuclei stand out very prominently. 
The thylose cells are of about the same size as the mucilage or tannin 
cells which alternate with them, but are less regular in outline, and for 
the most part possess thicker walls. When stained in hematoxylin they 
are easily differentiated by the lighter color they assume. Remains of 
the disorganised protoxylem are frequently observed lying among them. 
In the specimens examined twenty-five to thirty thylosal groups were 
present in each bundle. 
In sections obtained from buds of fronds of Osmunda regalis almost 
the same appearances as in the young stipe of Todea barbara were ob- 
served. The tissues as a whole are very immature, the tannin or mucil- 
age bearing cells being the most evident. Between the groups of tannin 
cells are the protoxylem groups (Photo 18), which are practically intact, 
and are the only elements of the xylem defined at this stage. There are 
about twelve such groups in each bundle. 
In sections taken from.older stipes the tissues were already well 
defined, and thyloses had made their appearance (Photo 19). The thy- 
loses in this form, however, do not stand out very clearly, but resemble 
the surrounding parenchyma a great deal both in color and in contents. 
Moreover, the parenchyma cells are very large here and this renders the 
resemblance even more striking. However, on close examination, it can 
be seen that the thylose cells are very irregular in outline, and that they 
contain very clear lines and circles which turn out to be the highly re- 
fractive remains of the broken protoxylem elements. 
This difference in appearance from the thyloses in Todea is pro- 
bably due to the different age of the two. In Todea it is seen that the 
thyloses are even thicker walled than the xylem elements, and this is no 
doubt due to very active growth in the thylose cells. The thyloses would 

