372 ROYAL SOCIETY OF CANADA 
found in any living ferns according to Molisch (11, 88), who made a 
most careful examination of about 700 plants with a view to ascertain- 
ing the distribution of these curious cells. Nor is it surprising that they 
should be absent from Ferns and Gymnosperms, when we consider that 
their main function is probably to close continuous vessels. For these 
latter are generally absent from the two groups of plants referred to, 
which, possessing tracheids, have less need for this form of closing a 
wound. Their mode of origin, too, upon which Molisch insists, as in- 
growths from surrounding parenchymatous cells, seems to preclude their 
formation in Gymnosperms where wood parenchyma is generally absent.” 
As will be shown below, the canal formed in the protoxylem region 
is a definite continuous vessel with a fixed function, only it has no 
bounding wall of its own; its wall consisting of the wood parenchyma 
cells and persistent protoxylem elements which line it. The outgrowths 
from these parenchymatous cells fill the requirements of thyloses in 
every respect, and must be recognised as such, for, as will be proven 
later, they arise in the same manner as the thyloses in the Mono-, and 
Dicotyledons and owing to the same causes. 
That this phenomenon has not been noted before in the ferns (except 
by Thome, who only observed some of the stages, and whose work seems 
to have remained unnoticed), is no doubt due both to its peculiar mode 
of occurrence, and to the fact that all the stages of growth have to be 
examined in order to interpret it correctly. As has been shown in the 
historical part of this paper, the mature structure was observed by 
several investigators, who, however, failed to interpret it correctly with 
the data at their disposal. 
The cause of the formation of these peculiar cells in the Pterido- 
phyta will be dealt with below, and an attempt will be made to show the 
purpose and origin of thyloses in general, both in the Cryptogams and 
in the Phanerogams. 
Returning to the sections it is seen that the whole transverse sec- 
tion still shows very little maturity of parts, and the canal with the 
accompanying thyloses is the structure most defined after the appear- 
ance of the protoxylem. There may be several canals in a bundle, the 
number varying with the number of protoxylem groups. Torn paren- 
chyma cells are now and then visible on the walls of the canal, and this 
fact shows that considerable lysigenous disintegration occurs in the 
formation of the canal. This point will be dealt with more fully later. 
In sections taken about 10 cm. below the tip, the tissues exhibited 
much greater maturity of parts, the secondary xylem showing striated 
walls of considerable thickness. The thyloses in this region filled up 
the greater part of the cavity of the individual canals, the ingrowing 
thylosal cells frequently showing dividing walls and thus initiating the 
eS ee NS PP 
