410 McNicoL — On Cavity Parenchyma and Tyloses in Ferns. 
of the stem : ‘the large diameter and short length suggest a storage rather 
than a conducting function,’ p. 248. 
At the same time the cells probably act to a certain extent as 
strengthening tissue. Thus in Phanerogams and Ferns there are instances 
of the lignification of tyloses resulting in cells which have all the appearance 
of water-storing cells. 
Cavity parenchyma is to be regarded as a special tissue formed by the 
conjunctive parenchyma cells of the vascular bundles of the petiole, which 
replaces the first-formed elements of the wood, sometimes by simply crush- 
ing the spiral vessels, but generally by means of tylose-like swellings within 
the cavity of the vessels. True tyloses, that is, soft cells which enter the 
wide tracheids or vessels and remain within them, simply closing up the 
cavities by the formation of a pseudo-parenchyma, are unknown in recent 
vascular cryptogams, except in a case mentioned by Conwentz, who found 
tyloses in the old leaf-stalks of Cyathea insignis , and in the case mentioned 
by Johnson in the rhizome of Pteris aquilina ; they occur frequently in 
monocotyledons and dicotyledons where, on account of the large amount of 
conducting tissue, the advantage of stopping up some of the wood vessels 
can be readily understood. In the fossil Fern petiole Rachiopteins insignis , 
now known to have belonged to Zygopteris corrugata , the xylem elements 
are often seen to be entirely closed up by soft parenchyma cells. It is 
a matter of some difficulty to explain the presence or function of these soft 
cells, on account of the fact that in the petiolar bundle the tracheids form a 
compact mass and are not separated from each other by any soft cells. 
Thus if the tyloses, generally so-called, are really formed by the ingrowth 
of the xylem-sheath cells, these cells must have been in a remarkably active 
condition to grow to such an extent as to fill all these wide tracheids. 
It is quite possible that such was actually the case, though one hesitates to 
ascribe to fossil Ferns an activity not present in recent Ferns, for even in 
the Tree Ferns, where cavity parenchyma attains its greatest development, 
and in Nephrolepis , where a single cell can be seen growing through two or 
three vessels, the activity of the sheath cells occurs only opposite the weak 
protoxylem vessels : the cells never enter the wide tracheids of the meta- 
xylem. The formation of cavity parenchyma is, of course, of the same 
nature ; but when one considers the usual extent of the tissue as compared 
with the extent of the whole bundle, it will at once be seen that there 
is a great difference between the activity necessary to form cavity paren- 
chyma and that which would be required to form a pseudo-parenchyma 
throughout the tracheids of the bundle. Gwynne-Vaughan, in a paper read 
at the British Association at Leicester, 1907, showed that the middle 
lamella of recent fern tracheids consists of pectose which disappears as the 
plant becomes older, the lumina of the adjacent xylem elements being thus 
placed in communication with each other by the disappearance of the pit- 
