Bancroft .— On the Xylem Elements of the Pteridophyta. 755 
according to his view, disintegration is complete and an empty space is left 
between the thickening bars of contiguous elements. 
Halft’s researches prove, however, that in Osmunda regalis , as well as 
in other forms, the spaces between the lignified bars are not entirely empty 
—the middle lamella stretches across from end to end. In this connexion 
Mr. Sinnott’s excellent photographs, illustrating ‘ Foliar Gaps in the 
Osmundaceaemay be mentioned. His Fig. 8 (Todea hymenophylloides) 
shows the presence of the lamella perfectly (Sinnott, 14 ). The writer 
has also observed the lamella in Osmunda Claytoniana, O. cinnamomea, and 
Todea hymenophylloides amongst the Osmundaceae, as well as in repre¬ 
sentatives of other families of the Filicales and of other groups of the 
Pteridophyta. 
In Gleichenia dicarpa good evidence of the presence of the lamella was 
afforded by a torn section, in which adjacent bars of secondary thickening 
were broken and displaced, while the lamella projected from between the 
bars, having given way at one of the angles (PI. LVI, Fig. 6). 
Halft used principally only mature material in order to prove the per¬ 
sistence of the middle substance, even in the later stages ; but he appears to 
have sectioned young xylem of Osmunda also, for he mentions that it does 
not show splits, the ‘ springing apart ’ of the lignified bars not having taken 
place at this stage. 
Not having sectioned xylem of different ages, he has failed to see that 
a certain amount of disintegration actually does occur. An examination of 
Todea hymenophylloides and Tmesipteris tannensis has shown that the splits 
between the bars of thickening of adjacent elements are formed by the 
partial disappearance of the primary wall, leaving the middle membrane 
traversing the space from end to end. Its extreme delicacy renders it 
difficult of observation, especially if the staining is light and the focus not 
extremely exact. 
Halft appears to think that the presence of a split is due to the mere 
‘ springing apart ’ of the bars of thickening, and that the primary wall 
remains between, no dissolution taking place. But the pit-membrane is 
middle lamella only, and as the primary wall consists of middle lamella 
and layers of pectic substances on either side, it would seem that some dis¬ 
integration of the primary wall must occur where the middle lamella is 
exposed and separated from the secondary bars. Where the bars actually 
join, it may be supposed that the primary layers remain, as well as the 
middle lamella. The longitudinal section of Tmesipteris tannensis (Fig. 15) 
in particular supports this view—the primary layers being distinctly thicker 
where the bars join than where they separate, thus indicating what the 
transverse sections showed to be the case—that some disintegration has 
taken place. 
It may be that a certain amount of ‘ springing apart ’ helps the widening 
