886 ROYAL SOCIETY OF CANADA 
to or within the otherwise persistent xylem, sometimes so that, though 
the latter is originally formed, the tracheæ become destroyed and de- 
graded as the parts expand.” 
“In numerous Monocotyledons, the Equiseta, and some Dicotyle- 
donous water-plants, at the inner edge of the bundle, where the primitive 
tracheæ are placed, a passage is formed by the peripheral extension of 
the surrounding cells, t.e. schizogenetically, while the external part of 
the xylem attains perfect development and is persistent............ sd 
“The width attained by the passage is various, sometimes equal to that 
of a moderate vessel, sometimes to the cross-section of the whole per- 
sistent part.” 
The phenomenon described above is very common in the Pterido= 
phyta, as the results obtained will demonstrate, but was not observed 
by De Bary or by the authorities he refers to, except in the Equiseta. 
The next mention of these structures is by Terletzki, who discusses 
them rather fully (21), but he gives neither their origin nor their func- 
tion, and fails to interpret their true nature. Like the others, he only 
describes the thyloses, and makes no mention whatsoever of the canal. 
The first reference to these structures in his article is in his de- 
scription of the anatomical characteristics of the rhizome (Auslaufer) 
of Struthiopteris germanica (21, 463), where he proceeds as follows :— 
“A considerable accumulation of parenchyma is frequently found 1+ 
one, more rarely at two places in proximity to the ends of the woody 
tissues (Holztheils), and indeed always very near the spiral and annular 
tracheides. Where the woody part (Holztheils), as is frequently the case, 
curves in at one side in a hooked manner, such a parenchyma group 
always lies on the concave side of the curve. In cross section a group 
of cells remarkable for their wide lumen, stands out in the centre of the 
parenchyma accumulation (Fig. 1n and Fig. 9n).” 
“Tn longitudinal section it is evident that we have to deal with a 
strand of peculiarly built cells which run parallel to the primary xylem 
(Erstlingsholz). This strand consists of parenchyma cells which are 
not elongated like the other conducting cells (Geleitzellen), but are as 
broad as they are long. These groups of cells are further differentiated 
by possessing large intercellular spaces, which are formed by the irre- 
gular polyhedral or contiguous globular constituent cells. Owing to the 
shape of their cells, I shall indicate the structures under discussion s 
‘Stumpfzell-Stringe’ throughout the course of my work, or, where no 
confusion with other structures is possible, as ‘ Zellstrange.’ ” 
“The cell walls of these are very little thickened and show the same 
pores as the conducting cells previously described. Similar Stumpfzell- 
