Bancroft.—On the Xylem Elements of the Pteridophyta. 757 
be much less than where the connexion of the bars is wider. It is tentatively 
suggested that there may be a mechanical explanation of the variation in 
type. In the case of Osmunda , for example, where there is variation within 
the organs of the same plant, the mechanical needs are clearly greater in 
the petiole than in the rhizome, and the efficiency of the xylem as a 
strengthening tissue is here raised by the wider connexions of the bars. 
Again, where the connexion of the thickening bars is slight, there will 
be a larger area through which transfusion of water and solutions may take 
place, in a lateral direction as well as vertically. 
The different parts of the conducting system in this case will be in 
more facile communication than where the exposed areas of lamella are 
smaller and transfusion is consequently somewhat retarded. 
With regard to the detection of true vessels, 1 such as were found by 
Halft in Pieris Aquilina , longitudinal sections are necessary, for it seems 
that the perforation of the end walls of the xylem elements depends, in 
a great measure at least, upon the amount of their inclination, and that 
where end and side walls grade into one another and are not properly 
differentiated, as is the case in the pointed elements of Osmunda regalis , 
pit-membranes remain throughout; but where end walls are well defined 
and very little inclined to the lateral walls, the need for true perforations 
may arise. In all the cases where longitudinal sections were used as 
controls to the transverse (as in the majority of the types mentioned), true 
pointed tracheides were found, confirming the opinion of Halft that these 
are the rule in Vascular Cryptogams. 
Summary. 
1. The xylem elements of Pteridophyta are typically pointed tracheides, 
the pits on both end and side walls being closed by a membrane formed by 
the persistent middle lamella. 
2. The middle lamella is believed to be exposed in the pit areas by 
the disintegration of the rest of the primary wall. 
3. The varying amounts of disintegration give rise to the prevalence 
of the ‘split’ or ‘non-split’ appearance between the secondary lignified 
layers, as seen in transverse section. 
4. The width of the split depends largely upon the degree of separa¬ 
tion of the secondary layers in the areas surrounding the pit openings. 
The grateful acknowledgements of the writer are due to Dr. H. C. I. 
Fraser for drawing attention to this problem ; to Professor Carr for material 
and opportunities for research ; and to Professor Gwynne-Vaughan and 
Mr. H. S. Holden for valuable criticism and advice. 
1 Halft employs the term * trachea * in its limited sense meaning * vessel whilst Gwynne- 
Vaughan apparently follows De Bary’s usage, giving it a comprehensive application. In the present 
instance also, ‘ trachea ’ is understood to include both 1 vessel ’ and ‘ tracheide 1 . 
