362 ROYAL SOCIETY OF CANADA 
thylosal strands to the protoxylem groups he did not observe any rela- 
tion between the protoxylem, the canal and the thyloses. 
In conclusion, Thomz mentions the thyloses found in the vessels 
of the fossil fern Rachiopteris corrugata, which was described and 
figured by Williamson, and which will be discussed further on in this 
paper. Thome does not figure the canal or the thyloses, except on Plate 
VIIL, Fig. 7, where, in a cross section of the stipe of Hemitelia spec- 
tabibis, he shows a thylose next to some protoxylem elements. This 
appears to be rather diagrammatic, for no similar appearance has been 
observed in any of the preparations examined by the writer. 
From the above account it will be seen that although Thome in- 
terpreted the nature of the cells correctly, since he first identified them 
with the well-known thyloses of the Mono-, and Dicotyledons, still he 
neither grasped the significance of the structures, nor observed the 
various phases of development. 
The next reference to the subject is found in Strasburger’s Histolo- 
gische Beitrage (19). In his description of the vascular bundles, of 
Pteris aquilina the following passage is encountered (19, 443): “The 
largest, almost elliptical steles (vascular bundles) of the rhizome, show 
at one, more frequently at two places, which correspond more or less to 
the foci of the elliptical cross-section, groups of narrow lumen (englu- 
migere Gruppen), which are composed of narrow vessels and wood 
parenchyma. A portion of the narrow vessels of these groups is still 
thickened in a scalariform manner, others are already spirally thickened ; 
the greater part is disorganised and shows isolated spirals and rings. 
The elements of narrow lumen represent the beginning of the vascular 
portion,—the primary vessels. The parenchyma which accompanies 
them is for the most part thin-walled and poor in contents. In the 
vascular bundles of the stipe,-the elements of this parenchyma frequently 
swell up like bladders, and, in places, leave the bounds of their tissue, 
forming lacune. The members of the protoxylem (Erstlingsgefässtra- 
cheiden) which have not been absorbed are also capable of forming such 
bladder-like swellings. Dippel, Russow and De Bary have already re- 
ferred to similar appearances.” 
Thus Strasburger observed the formation of the canals in Pteris 
although he neither designates them as such, nor definitely mentions 
any cavities that are formed, merely describing the disorganisation of 
the protoxylem elements. He also observed the formation of thyloses 
in the stipe, but he did not recognise them as such. In the rhizome he 
missed their appearance altogether. 
As regards the statement that the protoxylem elements are capable 
of forming thyloses (bladder-like swellings) he is entirely wrong, for 

