From Alan Bowman and David Thomas, Vindolanda: the Latin writing
tablets London: Society for the Promotion of Roman Studies,
1983, pp. 26-29
The physical characteristics of the writing-tablets,
particularly those of the leaf type, are of some importance. The
plane dimensions of the complete diptychs, where they are preserved
or can safely be inferred, show only small variations. The largest
are of the order of ----> 20 cm by 9 cm, the smallest of the order
of 16.5 cm by 6 cm (the arrows indicate the direction of the grain
of the wood in relation to the measurement). More striking, in a
significant number of examples, is the extreme thinness, less than
0. 25 mm; the majority of the leaves are of the order of 1- 2 mm
in thickness, some are as much as 3 mm. The surface of the leaf
is generally very fine-grained and smooth and, with a few possible
exceptions, had clearly been prepared expressly for writing in ink.
The following features of wood anatomy have proved most useful in identifying the tree species from which the Vindolanda leaf tablets were cut28:
1. The arrangement of vessels within each annual ring.
2. The structure of vessel endplates.
3. The formation of medullary rays.
4. The size and distribution of pits in vessel side-walls.
Other features of wood anatomy were also recorded when present. The samples were finally identified by comparing their wood anatomy with published descriptions and by comparing them with reference slides of wood taken from different tree species.
Thin sections were taken from the leaf tablets by hand, using a sharp razor. Sectioning was confined to the edges of the tablets, taking care to avoid areas where text covered a break in any fragment. The written surface of the tablets was not sampled, to avoid damaging the text. Sections running parallel to the grain of the wood were obtained relatively easily, but transverse sections running across the grain were extremely difficult to make since the wood lacked strength and crumbled easily. However, with care, transverse sections were taken from most of the sample tablets.
In order to prepare slides, sections were taken from the dry tablets by hand, using a double-ground razor. These were then placed in a mixture of glycerin and ethanol (1:1 by volume) to prevent any possible damage to the highly desiccated wood slivers through hygroscopic movements. Sections were left in this mixture for between thirty and ninety minutes, and were finally mounted on microscope slides in Canada Balsam mountant and dried at 60 C. In initial preparations some sections were bleached for thirty minutes in a 1 % solution of sodium hypochlorite in glycerin. This removed all pigmentation from the sections but made the tissue structure less visible. Photomicrographs were taken on Pan F film using a Zeiss photomicroscope II.
Eleven leaf tablets have been examined botanically, eight by Dr. Tapper and three by Dr. J. Turner, who has also examined two stylus tablets29. Of the leaf tablets, nine have been positively identified as being cut from alder or birch, while from two it has proved impossible to obtain sufficient detail to permit firm identification.
The characteristic features of alder and birch
are shown in Table III.
|Annual ring structure
|Vessel pit dimensions
|1.4 X 2.5 pm to 3.6
to alder Vessel
|36 to 72
|uniseriate and multiseriate
rays up to four cells wide
Both woods are diffuse-porous, possessing vessels scattered throughout
each years growth. The endplates of the vessels are scalariform
in both species and have between 12 and 25 slit-like perforations.
The pits in the side-walls of vessels are also similar in the two
species. In both cases they are arranged in close reticulate pattern,
between anastomosing bands of thickening in the vessel wall. They
are circular to elliptical in shape, the average dimensions being
3.6 llm X 1.4 . However, detailed observations
suggest that, while the dimensions of the pits are the same in both
species, pits in vessels of birch tend to be polygonal rather than
with a smooth outline.
A major difference between wood anatomy in alder and birch is the presence of medullary rays that are up to four cells wide in birch, while those in alder are never more than one cell across. In some parts of alder wood, especially near places where a branch diverges from the main trunk, the medullary rays may be closely aggregated to each other, but this was not observed in samples taken from the leaf tablets which were examined.
Leaf tablets cut from alder (Alnus glutinosa)
Seven out of the eleven leaf tablets which
have been examined are made of alder. Positive identification was
made from the presence of uniseriate, homogeneous medullary rays
only, which distinguishes alder wood from that of birch, and by
the presence of scalariform endplates with at least 12 bars. Tearing
in the sections limited the number of bars that could actually be
counted (PL. XIV, 3 and 4 - launch Plates viewer window). Apart from
birch, no other tree species in which scalariform endplates are
common possesses homogeneous medullary rays.
The diameter of vessels ranged from 35 to
72 , and this
range was fully spanned by vessels from tablet No. 45 (341).
Most of the tablets showed a less extreme range, while tablet No.
was found to possess smaller vessels ranging from 35 to 50
in diameter. Pit dimensions were variable from 1.4-3.6 mmm across
and were circular to elliptical in outline. Medullary rays are shown
(i) in cross-section, (ii) tangential longitudinal section (TLS)
and, in side view, (iii) radial longitudinal section (RLS). All
rays observed were homogeneous and uniseriate. The uniseriate rays
were mostly 10-12 cells high, with occasional rays of 8 cells or
less. They are darkly coloured from the accumulation of staining
matter during the long period over which the wood has been buried.
In a few sections the rays are bent and distorted (presumably as
a result of desiccation followed by burial) compared to sections
of modern wood. Side-views of the rays are shown in PL. XIV (launch Plates viewer window), 2 (RLS), in
which the ray cells can be seen lying horizontally over vertically-arranged
vessels and tracheids. Plasmadesmata connections between ray cells
are visible in PL. XIV, 2 and have the appearance of short chains
of beads. These connections permit horizontal movement of water
and other materials along the rays.
Most of the alder tablets consist of small fragments approximately 2 mm in thickness, which possess a fairly uniform writing surface. They are almost all wedge-shaped, being rather thicker at one end than the other. Two of the tablets possess additional features of note. Tablet No. 5(TVI(I191) has a thin sliver of bark attached along one of the sides. The width of this strip is no more than 3 or 4 mm and indicates that the bark of the alder from which this tablet was cut was rather thin. The surface of the tablet also includes several small knots. Tablet No. 43 (325) bears what appear to be horizontal cutting marks on its surface.
Leaf tablets cut from birch (Betula spp.)
Only two tablets cut from birch have been identified.
In tablet No. 29 (241),
which was examined, transverse sections (TS) clearly revealed the
presence of an annual ring (PL. XIV, 7). This was the only ring
seen in any of the tablets, and appears to run obliquely across
the edge of the tablet, rather than parallel to it. Similar rings
must be present in the alder and other tablets, but have not been
observed because of the difficulty of preparing transverse sections.
A uniseriate ray and multiseriate rays can clearly be seen traversing
the layer of small-diameter tracheids formed at the end of the growing
season and they pass into wood formed in the succeeding year. The
vessels are large in this sample, with the largest in the spring
wood attaining a diameter of more than 100 .
The pits in the side-walls of vessel elements are polygonal rather than circular in outline in this tablet. Fungal spores are also found in sections taken from it and indicate that the tablet was attacked by a saprophytic fungus either before burial or following excavation.
Other leaf tablets
Two leaf tablets have not been positively identified. Dr. Turner describes Tablet No. 22 (250) as being cut from a diffuse-porous hard wood, which possesses uniseriate, homogeneous rays and multiseriate, slightly heterogeneous rays up to 4 cells in width. The pits appeared large and alternate, but no endplates were found, and so little was examined that the absence of spiral thickening could not be taken as conclusive.
Some features of Tablet No. 39 (299) are illustrated in PL. XIV, 8-10. A uniseriate ray is shown in cross-section but in side-view it appeared that more than one layer of cells was present in the medullary rays, when the plane of focus was altered. The rays are homogeneous to very slightly heterogeneous. A significant characteristic of this wood is the presence of large elliptical pits, shown in PL. XIV, 10, which are about three times the size of those of birch or alder. Endplates were not found in any sections taken from this tablet. Without observations of the vessel endplates and of greater detail in the medullary rays, the wood cannot be firmly identified. However, the presence of large pits suggests that Tablet No. 39 (299) may be made from willow (Salix sp.).
Tablets No. 117 and 113 were examined by Dr. Turner, who
identified the wood as either larch (Larix)
or spruce (Picea) from the
presence of resin ducts and the pitting of the tracheids. These
tablets are more than 7 mm thick, with the centre hollowed out to