[kirscx] CERTAIN STRUCTURES IN THE PTERIDOPHYTES 363 
they do not possess this power. Owing to the disorganisation of the 
spirals, however, they frequetly simulate bladder-like swellings, this 
appearance being due to the disposition of the stretched spirals. This 
is a purely mechanical appearance consequent on the rupture of the 
spiral tracheids and has no physiological significance. 
It is a curious fact that Strasburger does not mention Thome’s 
observations on this phenomenon, although he certainly knew of his 
article, since he refers to it in another connection a little further on 
(19, 445). 
This brings to a close the account of the structures as observed and 
interpreted in modern Pteridophytes, and a short account will now be 
given of the occurrence of these structures in fossil forms. 
Thyloses in fossil ferns were first described by Williamson (26) 
in his account of the vascular bundle of Rachiopteris corrugata. The 
structure and mode of occurrence of the thyloses are figured on Plate 
VI., Figs. 15 and 16. “In fig. 15 the dark, strongly marked walls of 
the vessels enclose densely packed masses of cells (i). The longitudinal 
Bection® (Mig. N16) 000300 shows it to be similarly crowded with thy- 
loses through its entire length.” (26, 214). Williamson takes the ap- 
pearance of thyloses in plants of so ancient a period to afford a striking 
example of the persistence of types of elementary tissue, additional to 
those we already possess. In a later paper (27, 506) Williamson de- 
scribes Rachiopteris insignis, and figures thyloses in the vessels of the 
vascular cylinder (Plate XVI., Figs. 19, 20, 24). The thyloses here fill 
up the large vessels of the xylem in the same manner as in Rachiopteris 
corrugata. 
But an appearance which is much more significant, in view of what 
has been observed in existing ferns, is a group of cells on each side of 
the large vessels, which Williamson describes as “liber” structures 
(Plate XVI., Fig. 20, g). These groups are exactly in the position of 
the protoxylem of the bundles, and are of the same size and structure as 
the thyloses which fill the vessels, and so they undoubtedly correspond 
to the thylosal cells which occur in the position of the protoxylem in 
recent ferns. That this is the case is confirmed by the following state- 
ment by Williamson: “ Fig. 21 represents the central part of a trans- 
verse section of a similar stem to Fig. 19, but made so obliquely as to 
be almost a longitudinal one. From it we learn that all the cells c aud 
d enclosed within the middle cortical layer b, excepting g, are long, 
narrow, and have square ends. We also see that the large vessels, e, and 
the smaller ones, f, are alike barred, the former being filled with thylose 
as in the transverse section.” This is identical with the appearance in 
