520 Gwynne- Vaughan . — On the Real Nature of 
sections indicate that the tracheae undergo a considerable amount of sliding 
growth which must take place while they are still in a thin-walled condition. 
The first signs of lignification in the secondary layers appear at the corners. 
Shortly afterwards they become lignified all over, and so do the angular 
thickenings of the primary walls (Figs. 6 and 7, unshaded), and also the 
regions separating the vertical series of pits. All the rest of the primary 
wall remains pectose, and as soon as the lignification of the above-mentioned 
regions is completed it begins to show signs of disintegration. Its substance 
breaks up and becomes granular and then gradually disappears without, 
however, altering its composition, for to the last it is coloured by the pectic 
stains (Fig. 8). Its complete re-absorption leaves an empty space between 
the two bars of secondary thickening, or, in the region of a pit, sets the 
cavities of the two contiguous tracheae into continuity. The pectose 
portions of the primary wall seem to be reabsorbed more rapidly in some 
specimens than in others, and in all cases some remains are still to be 
observed coating the extremities of the persistent lignified parts (Figs. 3 
and 4). Where a trachea borders upon a cell of parenchyma the primary 
wall of the trachea remains intact, and so also does the middle lamella 
between it and the wall of the parenchymatous cell. It follows from this 
description that the holes in the tracheal walls are due to the re-absorption 
of the whole thickness of the primary wall of the young tracheae at the points 
in question, and not merely to the disappearance of the middle lamella. 
Longitudinal sections are not particularly helpful in demonstrating 
these holes. To obtain an ideal section the razor should pass twice through 
the same series of pits and then the several bars, being loose and uncon- 
nected, all fall away from one another. In thicker sections the pectic 
remains that coat the ends of the persistent portions of the primary wall are 
visible through the unstained lignified parts and simulate a continuous 
middle substance. If, however, the cut edge of a vertical wall be brought 
into focus it is seen that the pits are extremely narrow, and that the bars of 
thickening often vary greatly in width. They usually project much further 
into the cavity of the element in the neighbourhood of the corners than 
elsewhere. This accounts for the overlapping outlines so often seen in 
transverse section (Fig. 8). 
The same type of element was found in stem, leaf, and root of all the 
species of Osmunda and Todea that were examined, and it may also be 
confidently inferred in the fossil representatives of the order. A reference 
to Fig. 11, which is a photograph of the xylem of Osmundites skidegatensis 
will illustrate the peculiar and at first inexplicable appearance of the 
tracheal walls that initiated these observations. It now appears obvious 
that the black bars represent the bars of thickening, while the white spaces 
represent holes that occur in the substance of the walls. In the other 
fossil members of the order the xylem-walls vary very much in appearance 
