USES OF WOOD AND TIMBER
Reconstructing the Woods from Buildings, Hurdles and Ships
I have also consulted the available museums, only to find that many devote far more space to the war-gear of Polynesian tribes or the internal structures of obscure marine worms, than they can spare for the native arts and peasant crafts of our own people.
H.L. EDLIN, WOODLAND CRAFTS OF BRITAIN, 1949
Any writer on trees feels obliged to relate the peculiar uses to which each species is supposed to have been put. Herbert Edlin, besides Trees, Woods and Man, wrote Woodland Crafts of Britain, an invaluable record of ‘woodland’ crafts as they survived into the 1940s. As a student of craftsmen at work, he was rivalled only by Helen Fitzrandolph and Doriel Hay, Oxford dons of the 1920s. However, it lacks time-depth: this is one of those books that give the impression that ‘traditional’ practices are all of the same remote antiquity. It also lacks sources for most of the information. Another useful study, based on written records and limited to the period since 1750, is by E.J.T. Collins.1
Since Edlin’s time, although the crafts themselves have further declined, much more material has come from archaeological excavations and the study of antique furniture and standing buildings. (Built-in furniture, made by local carpenters, is especially instructive.2) Woodworkers such as Richard Darrah have been studying the crafts to the point of being able to replicate ancient artefacts. Museums have responded to Edlin’s stricture, though objects are still displayed without saying what they are made of. The time has come for an authoritative book on ancient and early-modern woodworking: its materials, methods and products. This chapter is concerned, not with makers and manufacture, but with what timber and wooden artefacts reveal about trees, woodland and woodland management.
EARLY WOODMANSHIP
Woodland management goes back at least as long as agriculture. Whether it goes back further is uncertain: the Mesolithic excavation at Star Carr produced evidence for carpentry and reedbed management by burning, but so far not for woodmanship. There is, however, a scrap of archaeological evidence in the form of buried multi-stemmed alder stools.3
The obvious motive is producing trees of manageable size. It is not too difficult to fell a tree, and experimenters have claimed that stone tools compare favourably even with chainsaws, but what happens next? Reducing a great tree to pieces that can be carried away and used is no light task, especially if one has not invented saws. Most human cultures use the smallest tree that will do the job, leaving the converting and extracting of huge forest trees until machines have been invented.
It is a surprise to go to Cartagena, Colombia – one of the best preserved of old Spanish towns – and find so much seventeenth-century timber surviving within 8 degrees of the equator. Much of this is not rainforest giants but small trees, one log per timber. Even slaves hated to waste effort cutting up big trees.
Other motives for woodmanship might have been producing a continuous supply of rods, poles and firewood, and producing leaves for feeding livestock. Conservation is unlikely to have been a motive until later: at first woodland was very abundant, though much of it would have consisted of trees too big to be useful.
Somerset trackways
Early woodmanship is known from ‘trackways’ preserved in peat, especially in the Somerset Levels. From Neolithic to Anglo-Saxon times people lived on fen islands and made walkways across the soft peat between, which have been engulfed and preserved by the growth of the peat. They are made of underwood, branchwood and cleft oak planks used in many different ways. Some materials probably grew on the peat and some were brought from the upland.
The famous Sweet Track, of early Neolithic date (about 3900 BC), is made of large poles and small timber. It has the greatest variety of species and the most elaborate carpentry. Some of the ash, oak and hazel poles apparently came from coppice-woods.4
Walkways of woven wattle hurdles are known from the middle Neolithic to the Iron Age. The earliest, Walton Heath and Rowland’s Tracks, are made of rods of hazel and a few ash. Like a modern wattle hurdle, they have straight sails (stout rods) interwoven with long flexible ethers to form a panel about 4 feet 3 inches by 8 feet (1.3 × 2.4 metres). The rods were rent from the stool by a blow with a stone axe. They are all used in the round (there being no metal tools needed to split them). Hence they are very uniform in size, but not in age. It was not the practice, as now, to fell a whole stool at once and split the thicker rods; instead, rods were drawn from each stool one by one as they reached the right thickness. Many rods had their tops cut off two to four years before being felled. Topping and felling were done in summer (the last annual ring is incomplete), which suggests that the main product of the wood was leaves.5
CONVERTING TIMBER AND WOOD
Timber (or wood) may be used either as the round log, or scappled into a square or rectangular section, or split or sawn into two or four or smaller divisions. Figure 199 shows various types of subdivision of the log and the shapes into which they warp as the timber dries out and shrinks.
Scappling
Many cultures use poles and small timbers in the round, lashed together with fibres. Carpenters’ joints and nailed joins, however, are best worked on flat surfaces.
In scappling a log, the excess may be hewn away with an adze or a special tool called a broad-axe. Or an ordinary axe can be used by the ‘notch-and-chop’ method, common in medieval England and modern Australia. Every 2 or 3 feet (60–90 centimetres) a notch is cut down to the required level; the hewer then turns the axe round and splits away the unwanted timber, leaving the bottoms of the notches still just visible. Saws were used only if enough timber was removed to yield a useful offcut.
The flat faces may extend to the corners of the beam; but usually in England the corners are left rounded (waney), often with the bark left on. Waney edges indicate whether the beam is from a whole log, and the diameter of the original log. They are the stuff of tree-ring dating: the outermost annual ring on a waney edge gives the year that the tree was felled.
A medieval carpenter would normally fell the trees and work them while still green, saving the labour and wear on tools of working seasoned oak. As a log dries out it shrinks: very little along its length, more in the radial direction, and still more in the circumferential. With some trees the difference may be taken up by the timber stretching circumferentially: some eucalypts are notoriously unpredictable to work because they have internal stresses lying in wait for the unwary carpenter. Oak usually cracks along the line of least resistance. An oak beam made from a whole log typically displays four cracks called shakes, from the pith to the middle of each face (Fig. 199); these wax and wane along the length of the beam as the pith wanders towards one face or another. If the tree is allowed to dry out before working the cracks will be distributed at random around the circumference.6
Cleaving
Before the invention of saws, people might split a log lengthways. This is the earliest known form of carpentry, going back to the Mesolithic hunter-gatherers of early postglacial Yorkshire.7 It calls for co-operation from the tree, especially for big trees in which it is difficult to influence the direction of the cleft as it progresses.
Woodmen make thatching-broaches or hurdle-ethers by cleaving rods of hazel or sallow, less often ash, using a metal edge tool called a froe; the workpiece is gripped in a frame called a brake. If the split runs off course it can be corrected by bending the thicker side.
The Pacific Coast Indians of North America, superb carpenters and architects, are said to have had a method of splitting huge planks off standing Douglas fir and other conifers. The Japanese, down to the seventeenth century, would split big trunks of hinoki (Chamæcyparis obtusa) and sugi (Cryptomeria japonica) into standard-sized pieces, small enough to carry, which were the usual stuff of the timber trade for middle- and upper-class buildings.8 Both technologies demand conifers with a very regular wood structure in which a split, once begun, proceeds along a predetermined plane, even if it is not along the radius of the tree.
Cleaving most pines, oaks or other of the world’s crooked, knotty trees is difficult, especially for people who have to make do with wooden wedges. It is an unrewarding task to propagate a split (Fig. 87) along the length of a long log if the grain is spiral or there are knots in the way. Nevertheless cleavable oaks were not uncommon in Anglo-Saxon England, to judge by their use in carpentry.9
In the Middle Ages there were enough cleavable oaks for the regular manufacture of cleft objects such as park pales, barrel-staves, clapboards (where not imported), laths, tile-battens and the thousands of shingles used instead of tiles on castle roofs and church spires. Edlin describes the Lake District craft of making spelk baskets from thin laths cleft from oak timber, a small-scale use in which the craftsman could be choosy about which logs to use.
Sawing
The Bronze Age invention of saws gradually spread from furniture making to coarser carpentry and finally to tree-felling: saws were once made that would get through a giant Californian sequoia. Saws can save labour and wastage and will go through knots and crooked grain (Fig. 88); but they are expensive and need regular doctoring; landowners disapproved of them because they made less noise than axes when trees walked off at night.
The Ancient Egyptians sawed lengthwise by lashing the timber to a post,10 but European practice uses a two-man saw. The accounts of any medieval building project include payments to sawyers at so much a square foot. At first one end of the log was propped up on a trestle, but in the fourteenth century the sawpit was invented, in which the log lies over a pit like a churchyard grave; the top sawyer stands on the log and the bottom sawyer in the pit. Each method produced characteristic changes in the direction of the saw-scratches as the saw reached the support and the log had to be moved along. When the sawyers got near the end they often split the last foot or so.
Mechanical saws, driven by water wheels, were invented probably in France in c.1200; they spread to much of Europe and then Scotland and America, but in England sawyers did not tolerate them until the eighteenth century.11 An early sawmill produces a pattern of parallel scratches, getting closer together on encountering a knot with its greater resistance.
The chainsaw, originally a surgical instrument,12 came into use for felling in the 1950s. It produces distinctive semicircular patterns on the sawn face.
CHOICE OF TREE SPECIES
Writers often claim that particular trees are necessary for specific uses, and even exist because of such uses: as if someone wanting to make chairs might plant beech trees, and come back and make the chairs when the trees had grown. There may occasionally be something in this ‘orchard’ ethos: the Royal Gunpowder Factory at Waltham Abbey had alder plantations (which still survive) because the wood structure of alder charcoal results in the particular intimate mixture of carbon, sulphur and potassium nitrate that gives gunpowder its bang. More often industries adapt themselves to the available trees or migrate to where suitable trees grow. To some extent trees are interchangeable: different cultures use the same tree for different purposes and different trees for the same purpose.
Fitzrandolph & Hay and Edlin describe uses of specific trees in the 1920s and 1940s. Significant properties include strength at right angles to the grain (in which oak is weak and elm strong); splittability (chestnut best, followed by oak); and ‘woolliness’, resistance to abrasion (poplar best).
Furniture
Furniture timbers depend more on fashion and custom than practicality. Almost any species has been used, even alder. Upper-class furniture of tropical hardwoods had its imitators in pine or beech, coloured with dragon’s blood – or, when the dragon-trees of the Canary Islands were not supplying it, with kino, the exudation from damaged eucalypts that is said to have been Australia’s first export to Europe.
Modern ‘Windsor’ chairs are factory-made of beech; those of the eighteenth century were made in local workshops of whatever was available, such as ash for the bow and sallow or birch for the legs; but the seat was always elm, whose high transverse tensile strength allowed it to be less than an inch thick. Beech, although otherwise an inferior timber, works well with woodworking machinery; it took over gradually as the craft became industrialised around High Wycombe. The last component to be made in beech was the seat, sometimes with disastrous results. America, however, has its own tradition of Windsor chairs and other furniture, in which the kind of timber does not matter: seats are made thicker to compensate for a weak material, and since American chairs are painted nobody notices the timber.
There are occasional medieval references to beech furniture, but it remained local until the nineteenth century. Much vernacular furniture, especially big table tops and built-in settles, is elm.
Wheels
George Sturt’s The Wheelwright’s Shop, published in 1923, is a classic account, by a practitioner, of one woodworking craft in its highly developed nineteenth-century form, which included making vehicles as well as wheels.
The spoked wheel was invented in the Iron Age and probably reached its highest development then. Two refinements, the shrunk-fitted iron tyre and the one-piece bent rim, were later lost: the tyre was reinvented in the eighteenth century, but even Sturt never recovered the bent rim. Wheelwrighting has two branches, dished wheels and flat wheels; in Britain the flat wheel is known chiefly in spinning wheels, but in Asia and Sardinia it is used for load-bearing wheels.
Sturt never made the nave, spokes, felloes and axle of a wheel of anything other than elm, oak, ash and beech. Other cultures use other timbers, although elm is generally used for the nave where obtainable: its high tensile strength in all directions is useful for a structure that has mortices cut in it all round. But I have heard of an Iron Age wheel with an ash nave; the Japanese use the elm-relative keyaki. Wheelwrights in Australia soon learnt to adapt their craft to different species of eucalyptus instead of European timbers; they made flat wheels on a heroic scale, such as 12-foot (3½-metre) wheels on the wains, drawn by herds of oxen, that moved logs of the giant, heavy karri eucalyptus in southwest Australia.
Turnery
Wooden cups, known as mazers, were made in their millions in Anglo-Saxon and medieval England. Coppergate in York is not the street of policemen, but of cuppers or cupmakers: its excavators found great quantities of spoilt cups and cup-cores of the tenth century. Cups were normally made by splitting a short log, about two and a half times the diameter of the mazer, into four billets, which were then made into cups on a lathe. Maple was specifically used for cups, but other species such as ash would serve. About 80 mazers survive above ground; almost all are ‘fine’ mazers with a ripple or bird’s-eye grain, banded with silver-gilt.13
Mazers were probably the commonest product of the turner’s art, which ran also to tool handles, chair-legs, balusters, reels and much else.
Bark
For most of history the bark of oak was a by-product for tanning leather; no other native tree has nearly as much tannin. It must be removed in May when the cambium is beginning the new annual ring. Medieval accounts include sales of bark when oaks were felled, but it was not very important: until the seventeenth century bark was commonly left on if the timber was felled at the wrong time of year to strip it.
In later centuries bark increased in value faster than timber. By the early nineteenth century mountainous stacks of bark came from every wood and hedge in the country, so valuable that thousands of acres of western oakwoods were managed solely as coppice; timber was foregone for a greater yield of bark. Those who think of the Navy as the great consumer of oak-trees for ships should contemplate the Army’s consumption of oak-trees for boots and saddles. After 1860 the value of bark sharply declined as other sources of tannin came into use, such as valonia, the giant acorn-cups of Quercus macrolepis from the Ægean. Bark-gathering lingered late enough to be described as a curiosity by Thomas Hardy and photographed by Edlin.
FAGGOTS AND FUEL
1 hundred wood faggots 6s. … ½ hundred furze faggots 3s. 4d.
BILL FOR BURNING ARCHBISHOP CRANMER, 155314
At most times fuel has been the chief product of woodland, in terms of tons per acre per year. Logs, faggots and charcoal would come from underwood (the residue after the more exacting trades had taken their pick) and the branches of timber trees. Timber itself would be more valuable for other purposes, and would incur the extra expense of chopping it into burnable sizes.
Faggots were bundles of rods, compressed in some sort of vice and tied with bands of twisted hazel (Fig. 89). (In Japan they are still made and are tied up with wisteria.) They came in various sizes, defined by market regulations, from the little ostrey faggot (a firelighter) through kids and bavins to the two-band faggot. In 1461–2 Canterbury Cathedral sold from Short Wood (the long wood in the Blean now called Church Wood) courtfagot, kechynfagot, halfagot and salefagot, besides logs in the form of belet, talwode, ostwode and orwe, rods, bark and charcoal.15
Faggots were used in brewing, baking (until the 1970s), brickmaking and land-drains (faggots laid in a trench). Sea defences included (and still include) long faggots placed to trap silt in salt-marsh creeks, or interwoven between upright stakes; these were significant products from woods near low-lying coasts. Late-medieval Calais was an English outpost: keeping out the sea and the French (it was doubtful which would get Calais first) called for large quantities of timber and underwood, most of which was brought from England.16
Logs came in various sizes of billets, shides and talwood. A consumer-protection statute (1542) laid down the permitted sizes, normally 4 feet (1.2 metres) long and of various girths from 20 to 44 inches (50 to 110 centimetres) which were supposed to have identifying marks cut in them. At Cambridge we find 40,000 billets, valued at £36, confiscated for being under size. In 1523 a licence was issued to export four million ‘byllet’ from England to France.17
How much notice was taken of the species of firewood? Almost any tree will do. Elm, despite common belief, is excellent fuel if well dried, but in my experience ash does not live up to its reputation. Woodmen need marketing skills when it comes to disposing of alder or sallow.
For charcoal the species matters more. Some processes, like smelting, call for specific sizes of charcoal; some trees, like pine, produce a soft charcoal that easily abrades into unusable dust. Oak makes a good hard charcoal. The Japanese use charcoal as a domestic and cooking fuel, and are particular about its quality: it has to be made of oak (even the species of oak matters) and must not be excessively charked. Good Japanese charcoal is almost as hard as the original wood. It is made in a semi-permanent kiln near a water source, used for damping down the fire. Charcoal sites are often to be seen in mountain woods.
TIMBER AND WOOD IN BUILDINGS
The choice of whether to build in timber or some other material is not merely utilitarian. The distribution of timber-walled buildings in Britain is not related to the distribution either of woodland or of stone. The Weald (with superabundant timber) has a strong timber-framing tradition, but so have Essex (with average supplies) and Cheshire (with little). In the Middle Ages, timber and stone were both often brought from a distance. To generalise, where there is not much timber building it will be urban, as in Devon, Dorset, Gloucestershire (east of the Severn) or Scotland. Medieval Norwich had mud houses, none of which appears to survive; these were superseded by timber-framing, especially in that characteristically urban form with a timber façade (often of very thin timbers) on a brick or stone rear structure.
Timber was an architectural medium, not merely structural; it could be a status symbol. Wattle-and-daub was designed to expose the timbers at least inside, and in middle- or high-status buildings on the outside as well. Timbers often have traces of original paint.
Species of timber
Oak, though it was expensive, is the commonest timber in standing buildings. This is partly a matter of survival: less durable timbers, and buildings containing them, will have disappeared more often. Until recently, other species were thought to be a mark of late date and were under-recorded. Elm occurs from the thirteenth century onwards, often in regions without much woodland, although sometimes in well-wooded areas such as northwest Kent. Ash and even aspen and black poplar survive from the Middle Ages, and are not usually in the last stages of decay: if kept dry they are quite capable of lasting 700 years. Pine occurs especially in Scotland and Cambridgeshire; it was always imported in England and usually in Scotland. Beech is curiously absent as a structural timber, although it functions perfectly well in France.
In underwater carpentry oak is not so obviously prevalent – either because the species was thought to matter less, or because waterlogging preserves non-oak species.
Timber-framing
Timber-framing in the medieval manner goes back at least to the Roman period. In Roman London, above-ground timbers are sometimes preserved through reuse in waterfront structures. The carpentry and the wattle-and-daub infill were similar in concept to medieval structures, although details differ and joints were rarely pegged. The timbers came from whole oak logs scappled from trees 25 to 65 years old. If the small amount of material is representative, the regular production of conveniently small oaks was as much a feature of Roman woodmanship as of medieval.18
Medieval timber-framing involved large numbers of small oaks. Normally each timber represents one log, scappled leaving waney corners. Bigger components come from bigger trees. Any house represents 150–300 such trees, a minority of which were bigger and used for such things as tie-beams and principal posts. Components such as floorboards and curved braces were sawn several from a log.19
Medieval sawyers could (if they had to) saw a great tree lengthwise into ordinary-sized timbers. At Gloucester in the thirteenth century, Henry III gave the Dominican Friars 82 huge oaks from the Forests of Dean and Gillingham (Dorset). The roof rafters, which still exist, were sawn four or six out of a tree.20 Some similar reason probably explains why Great Livermere, in the woodless Breckland, has a church with a fourteenth-century roof (now hidden) sawn with great skill from big, crooked oaks. However, outsize trees – and their transport – were normally too costly to waste in this way.
There was a change to using bigger trees and sawing them into smaller components; in eastern England this can be dated to the late sixteenth century. Probably the growth of oaks had been gaining on felling. Another possible factor is the falling cost of sawing, as better saws were invented and sawyers’ wages fell.
An analysis of a small two-and-a-half-storey house in the Weald of Sussex, dated to the 1560s, reveals at most 39 trees, all between 1 and 2 feet (30 and 60 centimetres) in diameter (and thus big trees by medieval standards), sawn into 280 components. Very few of the trees were more than 14 feet (4½ metres) in useful length; they are interpreted as hedgerow oaks.21
There are regional differences. In northern England ordinary timbers appear to be more often sawn out of big trees than in the middle and south. In the western half of England, for unknown reasons, cruck construction prevails. Every 10 to 15 feet (3 to 5 metres) a house is spanned by two great curved timbers running from the base of the walls (stone or timber-framed) to the apex of the roof. Often they are black poplar (which grows into that shape); sometimes they are the two halves of one tree. These support a frame made of lesser timbers.
In the Middle Ages, and down to the Oak Change of c.1900, oaks grew readily from seed in existing woods: any long run of woodland accounts records the felling of hundreds of small oaks, and replacement could not have been difficult (see Fig. 76). The woods being coppiced, browsing animals could not have favoured the regeneration of oaks, contrary to Francis Vera’s theory of wildwood (p.79f).
An example: Valley Farm, Flatford, East Bergholt
This is a hall-house in Suffolk, nearly typical of thousands that still exist and perhaps typical of hundreds of thousands that once existed. (It belongs to the National Trust, but is not generally open to the public.)
The house is in three parts (Fig. 90). The middle was a great hall open to the roof, heated by a fire smouldering in the middle of the floor. Both ends of the house are two-storeyed. The ‘service’ end has two rooms downstairs, the buttery and pantry, supposedly for storing drink and food, with a solar (a bedroom) upstairs. At the other end is a parlour (the sort of ‘best room’ that would later be used for weddings and funerals) with a second solar.
This is a middle-class house: the owner is thought to have been a modest farmer who was also a fuller, one of the small processing businesses that clustered around the towns of the cloth industry. In size it is in the middle of the range; there are many hall-houses both larger and smaller in East Bergholt. The few decorative details suggest a date of around 1400. (The timbers, as often, contain too few annual rings for a dendrochronological date.)
Alterations to the house were almost as standardised as the original plan. In the late fifteenth century a chimney was inserted in the hall, a brick tower projecting through a hole in the roof. A little later the hall was divided into two storeys by inserting a floor of massive timbers, in the makeshift manner of inserted floors.
The original fabric is identifiable by the soot-blackened timbers, the soot dating from the first hundred years before the chimney was added. Besides the hall, one of the service rooms is sooted inside, and so is the service solar; a fire was somehow contrived on the floor of an upstairs room.
Not all the timber is oak. Elm is used for some of the big timbers; ash appears in the service end. Generally these less durable timbers are used in places well away from the weather, although one wall top-plate is ash (and has rotted in one place and been fished by bolting on another timber with an early type of bolt). The original service-end stairs are still in use, with treads partly of elm, enclosed in a cage of black poplar boards.
In houses of this period one can often work out how many trees were used. A house of this size would normally contain 200–250 oaks, most of them about 9 inches (23 centimetres) in diameter, with a few bigger, and about 20 feet (6 metres) long. It would represent one year’s growth of timber trees on something like 150 acres (61 ha) of woodland; or (to put it another way) the timber component of a 50-acre (20 ha) wood could produce one house of this size about every three years for ever.22 In Valley Farm trees of even smaller size form the rafters of the roof (Fig. 91). They are sooted on worked faces, but where the corners are waney the soot deposit is patchy, showing that the bark was left on and slowly curled up and fell off. (A few rafters still have soot-covered bark attached.)
Much of the timber here, however, did not come straight from the tree. Many components are short, crooked, knotty bits of oak, oddments left over from some more important job. Others contain empty mortices, sooted inside, showing that they are reused. This was not just a matter of salvaging pieces from whatever building was on the site before: East Bergholt evidently had a recycling business, re-scappling second-hand timbers to new dimensions.
The use of oddments, recycled timber and non-oak looks like a concession to economy: the householder was building a little beyond his means. Economies are mainly in inconspicuous places, except for some elm timbers at the (socially) high end of the hall. These, however, are painted red (beneath the soot), so would not have looked different.
Lower-class houses are scaled-down hall-houses, sometimes lacking the parlour. In towns (like Hadleigh, Suffolk) they may be built in terraces.23 The trees used were often small, little more than coppice poles. They often include timbers other than oak, for example being all elm (even where oak was available). However, a little oak is usually present: perhaps its distinctive smell enhanced the status of even the humblest house.
Great barns
Among the biggest and oldest timber structures built from ordinary trees are aisled barns (often miscalled ‘tithe barns’). They are the barns of big, usually monastic or collegiate, farms. They may have stone or brick outer walls and huge timber roofs covered with tiles or tilestones; in east and southeast England they are wholly timber-framed. They were not just workaday sheds, but were built by architects (whose names are occasionally known) and were meant to be impressive.
The Knights Templars, soldier-monks, built two giant barns at Cressing Temple in Essex, the Barley Barn in c.1210 and the Wheat Barn in c.1270 (Fig. 93). Each is built round 12 great posts in two rows. As originally built they contained some 480 oaks per barn, mostly less than 10 inches (25 centimetres) in diameter. In the Barley Barn the trees are less completely scappled, leaving waney corners, giving an impression of rugged, massive solidity – a Durham Cathedral among barns. In the Wheat Barn slightly smaller trees were used and more was removed in scappling, giving a more elegant appearance – a Wells Cathedral among barns. The Knights had about 110 acres (44 ha) of woodland, which might have yielded one barn’s worth of oaks every five years.24
The rafters of the Barley Barn (originally about 300) were about 21 feet (6½ metres) long and made the best of a tree trunk that was not much longer. The crooked shape and pattern of branch-scars shows that there was only moderate competition from surrounding underwood. The Wheat Barn rafters were from rather smaller trees, drawn up straight by greater competition, such as might arise if oaks grew in a well-grown lime coppice. Much of the roof of the Barley Barn was replaced later in the Middle Ages, using oaks grown with more competition still.
The problem in both barns was getting 12 trees big enough for the posts. The trees, probably hedgerow oaks, were not especially big by modern standards – in the Barley Barn some 2 feet (60 centimetres) in diameter, in the Wheat Barn a little smaller – but in each barn the four best oaks were used for the conspicuous central posts, and the worst, waney and even rotten, oaks were put in dark corners. The wood likely to have supplied the other timbers was grubbed in the eighteenth century. A fragment of a contiguous wood survives (Lanham’s Wood, Rivenhall) – a lime and hornbeam wood, now containing several oaks big enough for barn posts.
Around Canterbury, despite thousands of acres of woodland in the Blean, post-sized oaks were even scarcer. The barn of St Augustine’s Abbey at Little-bourne had originally 22 posts, ‘curved like snakes’, as Alexander Wheaten put it. Faversham Abbey’s two barns have 29 surviving posts, three of them forked at the top of the tree.25
Timber quality and Baltic oak
Nearly all medieval structures, from barns to cottages, were built of what would now be very poor-quality oaks. Good-quality timber was normally imported in the form of ‘Baltic oak’ boards. (Those who think that the future of woodland is in growing ‘quality’ oak should reflect that this has not been done here before.)
Accurately cut boards from huge, slow-grown, straight-grained oaks came in bulk from what is now Germany, Poland and Lithuania. The trade began (as far as is known) around 1170 and ended c.1610. ‘Wainscot’ boards were not too expensive to use for ordinary purposes.
Baltic oak is stable, soft and easy to carve, and light in weight (p.242f). It was never used in floorboards, probably because it abrades easily. It is encountered in church doors, carved tracery and some linenfold panelling.26 In church screens the frame timbers are of English oak; the painted panels and carved tracery are usually of Baltic oak, though in Welsh screens I have seen even the most intricate carving done in fast-grown, hard ‘local’ oak.
Outsize trees
Cathedral roofs and royal palaces called for oaks bigger than barn posts. Lincoln Cathedral, planned in the twelfth century, called for tiebeams 46 feet (14 metres) long. Any oak much more than 2 feet (60 centimetres) in diameter or 30 feet (9 metres) in useable length would have been an outsize tree. These were very expensive per cubic foot and were hauled long distances, for instance from Northumberland to Walsingham in Norfolk. Such trees were probably at their scarcest in the fourteenth century, when great building projects such as the Lantern of Ely Cathedral or Westminster Hall were in progress.
The ordinary manor possessed one such timber, the post on which the windmill revolved. Such posts did not normally grow in the local woods, but had to be fetched from a special source. At Hatfield Broadoak (Essex) in 1328 the blacksmith’s perquisite was the second-best oak in the Forest every year, which was worth 13s. 4d., about ten times the value of an ordinarily big oak-tree.27
By the sixteenth century, oaks had got bigger, and trees of this size were not so exceptional. One was inserted into the medieval Old Court of my Cambridge college to carry the weight of an upstairs chimney. In 1503 two carpenters competed for the contract to replace a great beam in the church roof at Bassingbourn (Cambridgeshire). The timber cost £2 3s. 8d., a huge sum for a single tree, even including workmanship; it was fetched 16 miles (26 kilometres) at a cost of 9s. 5d.; it came from Datchworth (Hertfordshire) and was probably a non-woodland tree.28 The greatest such project was the replacement of the giant roof of Old St Paul’s Cathedral, London, after a fire in 1561. This called for 25 tiebeams 50 feet (15 metres) long and 16 inches (40 centimetres) square, and 2,440 rafters of 40 feet (12 metres) by 12 inches (30 centimetres). These were found, and the roofs prefabricated, at Welbeck Abbey (Nottinghamshire) and Guisborough (Yorkshire).29
Outsize timbers are sometimes elm. The keel of a ship – the biggest component in the hull, and one of the most important – would be a huge elm-tree. Two examples, 337 years apart, are the Mary Rose (1512) and the Jhelum (see below); each has a keel about 90 feet (27 metres) long, scarfed out of three great overlapping elms.
Wattle-and-daub
All-wattle structures: An entire one-storey building can be made of wood – of stout rods set in the ground, interwoven with flexible rods (ethers) to form a wall. The Iron Age roundhouses, replicated at Butser Hill, Hampshire, were 30 feet (10 metres) in diameter, with a conical thatched roof. Whole towns could be made of wattle, as in Viking Dublin and medieval Aberdeen which imply coppice-woods on a huge scale. ‘Caber houses’, sometimes stiffened with a few cruck frames, survived late enough to be photographed in the more wooded parts of the Scottish Highlands.30 I have found wattle-built outhouses, and the occasional dwelling, in the remote province of Grevená in the Pindus Mountains, Greece. Most early hurdlework involves whole rods. Often the sails are double.
Hurdlework was used as formwork to support concrete and rubble vaults or door and window openings while the mortar set. Usually the formwork was removed when no longer needed, or rotted away, leaving an impression on the underside of the vault. In medieval Ireland anyone building a vault had first to construct a wattle ‘vault’ on which to set it up. There are examples in Anglo-Saxon churches in England.
Hurdlework was used to form minor waterfronts, as excavated in Reading and many other places. Major quays, built to withstand the battering of ships rather than boats, were made of timber, new or reused. Hurdlework fences are depicted in many medieval pictures of gardens, but never survive; nor do the wattlework bodies of carts known from medieval drawings.
In timber framing: The commonest infill between timbers is wattle-and-daub (Fig. 92). Alternative materials include stone, interwoven oak laths (in Welsh Border barns and high-status work elsewhere), mud (in Normandy), special bricks (in medieval York and Norwich), half-baked brick (in New England), or air trapped between layers of lath and plaster.
Usually oak staves are fixed horizontally into small mortices cut in the upright timbers; stout underwood rods (whole or split) are set upright, passing in front of the top and bottom staves and behind the middle one, and tied on with string. The whole was then covered with daub, made of clayey subsoil with chopped straw added to prevent it from cracking; it was usually given a plaster finish. This is the commonest method for the tall narrow panels customary in medieval East Anglia, Essex and the east Midlands. In somewhat higher-status work cleft oak laths can replace the rods.
There are regional variations. Square wattle panels, favoured in west and south England, commonly have three vertical staves. Roman London had wide upright panels, each with about six horizontal staves, interwoven with thin vertical rods. Of medieval London, because of the Great Fire, little survives above ground: reused timbers from waterlogged excavations indicate horizontal staves, set in continuous grooves in the uprights, interwoven with vertical rods.
Wattlework, like some other aspects of woodmanship, was not much used in colonial America, but was introduced later into Australia. The name ‘wattle’, meaning species of Acacia, is said to derive from their use for this purpose for a few years in pioneering times.
Significance of wattle: For hurdlework, using interwoven rods, hazel or sallow (or osiers, as in basketry) are preferred because of their flexibility. Otherwise, the species did not much matter. For wattle-and-daub infill (excluding oak laths) in Eastern England I have found sallow to be the commonest, followed by hazel, ash and aspen, with small percentages of elm, maple, oak, lime, birch and willow.31
Wattlework on any but the smallest scale implies coppice-woods or coppiced hedges. Authors who wrote about ‘building huts from branches of trees’ could not have tried it themselves. Without regular cutting the supply of suitable-sized straight rods soon gets exhausted. Although wattle infill can last as long as the timbers, wattlework set in the ground would need renewing every ten years or so, calling for a continuous supply.
Wattle in standing buildings or excavated structures is probably a smaller sample than timber of the produce of the local woods and hedges. It is unlikely to be a random sample: the wattler would probably have taken first pick of the underwood, before less specialised users whose work never survives; the bulk of underwood would have gone for fuel. I usually find seven or eight annual rings in underwood rods, this being the length of the coppice rotation.
Why were sallow and aspen so abundantly used? Since they are among the least durable species, there was probably originally even more of them than survives. Aspen occurs as timber as well as underwood: being a weak competitor, it would be encouraged by short-rotation coppicing. Growth rate may have been decisive. The wattler, building a child- and burglar-proof partition, would choose the stoutest rods irrespective of species. Sallow and aspen, the fastest growers, would be more likely to reach the required size within a seven-year cycle.
Log building
There is another way to build in timber: by laying horizontal logs one on top of the other, notched where they overlap at the corners. It occurs in the Alps, Scandinavia and eastern Europe, transgressing national and tribal boundaries. It is unknown in Britain, but quickly got into European North America – it is uncertain how – which developed its own traditions (Fig. 94).32
The timbers are normally smallish conifers, whose straightness helps in this type of construction. The bottom log may be oak to resist rot. In the mountains of Norway I have seen a hut ingeniously built of curved rowan logs (there being no other tree) with the curve of each log matched to its fellows. America more often uses oak for the whole edifice.
Log building is sophisticated and lends itself to elaborate architecture, even churches with onion domes. I call the components ‘logs’, which in some workaday structures they are, but usually they are half logs, or scappled to a square section, or worked to an accurate cylinder or a more complex shape. There are different traditions in the jointing of corners, in door and window openings and gables, and in the method of fixing each log to the one above and below.33 In American high-class log cabins the ‘logs’ are often fractions of a log, widely set with mud or moss chinking between, and have axe-marks giving a decorative texture.
THE JAPANESE PARALLEL
It is instructive to compare Europe with how an independent civilisation used its trees. In Japan, monumental buildings – the equivalent of cathedrals – are all built of timber, for the Japanese never invented mortared stone construction.
The world’s oldest reasonably complete timber structure above ground is the Five-Storey Pagoda of the Horyu Temple in the countryside near Nara, Japan.fn1 This skyscraper tower (Fig. 95) is built from moderate-sized trees of the exceedingly rot-resistant, cypress-like hinoki (Chamæcyparis obtusa). It has been dated from tree rings to around 670 AD. Like all pagodas it is built round a central mast, a giant hinoki trunk nearly 3 feetfn2 (90 centimetres) in diameter and apparently 80 feet (24 metres) high, which mysteriously is dated around 594.34 This venerable structure, wherever the timbers are protected from rain, looks deceptively new; the round posts and great doors retain their faded red paint overlying the distinctive tool-marks produced by a small convex adze.
In the eighth century the Emperor Shomu built a capital at Nara on a scale of grandeur that only Nero, Louis XIV and Peter the Great could rival in Europe. His palace – nearly a mile (1½ kilometres) square – has been excavated. In its hundreds of buildings most of the timbers were hinoki, which was evidently not expensive and was used in workaday as well as monumental structures. Where could all this large hinoki have come from? What set its growth in motion, some 200 years before? As with oak in England, there is a discrepancy with its present ecology. Although much grown in plantations, as a wild tree hinoki is not common and tends to be a light-demanding pioneer of disturbed ground. Had it been encouraged by a period of renewed erosion in the mountains?
In Japanese monumental carpentry there are never waney edges where they will be seen. (Most tree-ring dates are therefore approximate, for the last rings are missing.) Hinoki trees used as columns do not follow the shape of the log, but are worked to exact (usually tapering) cylinders. Sometimes they are full of knots, indicating that the top length of a tree was being used. Such lower-quality timber may be put in less conspicuous places, although paint would originally have hidden the defects. The giant hall of the Todai Temple in Nara, as built c.1200, used about 80 hinoki posts each 4 feet (1.2 metres) in diameter and about 60 feet (18 metres) long, but this was (and remains) the biggest timber building that the world has seen. Its gatehouse, with similar posts, survives (Fig. 97).
From the fourteenth century onwards keyaki (Zelkova serrata) or kusonoki (Cinnamomum camphora) were used, either because hinoki was getting scarce or because red paint was becoming unfashionable and these trees have a beautiful grain. The present giant hall of the Todai, dating from 1709, has only 64 columns a little smaller than the originals; they are not single trees, but are built (like a medieval ship’s mast) of pine beams fastened with huge nails and iron bands to a keyaki core. (Where did all that keyaki come from?) The Senjokaku Temple on the sacred island of Miyazaki near Hiroshima, built by Hideyoshi Toyotomi, was left unfinished at that potentate’s death in 1596; unfinished it remains, revealing roof carpentry of huge, rough, crooked pine logs, not meant to be seen (Fig. 96).
Japanese vernacular houses
Rural houses, in contrast, are made of small timber, large coppice poles and bamboo, finished to similar standards to those of medieval England. Timbers are of small oaks, pines and the less durable sugi conifer (Cryptomeria japonica). Waney edges and crooked shapes are tolerated, indeed appreciated for their beauty. One enters the ‘service’ end of the house. The step where one leaves one’s shoes, and the post in the middle of it, are often keyaki with its distinctive grain. In the floor of the soot-blackened main room is a square irori pit on which a charcoal fire smoulders, reminiscent of a medieval English hall. A ladder ascends to the soot-blackened loft and the underside of the thatch (Fig. 98).
Urban houses, which are rarely ancient, tend to be of sawn sugi. This would have been needed suddenly in huge quantities whenever there was a great fire, like the one that consumed two-thirds of Kyoto in 1864.
Japanese wattle-and-daub, which is still regularly made, is normally of split bamboo, though underwood is sometimes used.
Japanese log-built strongrooms
These (azekura) are rare, but a handful of them are among the oldest timber structures in the world. The Shoso-in, a three-storey log structure of the 750s, held the treasures and curiosities of the Todai Temple for over a thousand years. They are apparently derived from granaries built on stilts. The ‘logs’ are hinoki timbers cleft to a hexagonal section with three wide faces and three narrow. Of the four buildings that I have seen, three got at least six ‘logs’ split out of a single length of a huge tree; the fourth derived most of its timbers from small knotty trees.
SHIPBUILDING
Most of the oak used in the [Lowestoft] yards was obtained within a radius of about fifteen to twenty miles … when buying standing trees Mr William Parker, the wood yard manager and draughtsman, would get on his cycle, go out into the country, visit the various sources of timber for sale and decide what to buy …
Ted Frost, old shipwright, 1985
Shipbuilding is the stuff of myth. It is supposed to have consumed more trees than anything else, devouring the ancient forests (or, alternatively, preserving ancient woods). In reality, one can read long in woodland history without meeting it: the records of Hayley Wood, the Bradfield Woods and Hatfield Forest never mention it.
Let us begin with practical works, such as Ted Frost’s account of building steam drifters at Lowestoft in the 1910s. The shipyards produced about ten vessels a year, about 90 feet (27 metres) long and 20 feet (6 metres) wide, measuring about 45 tons. Even in meagrely wooded northeast Suffolk there was no difficulty in getting timber.35
Let us visit shipyards building vessels of similar construction in Turkey (Fig. 99), where I have calculated that one ship of similar size represents about 3 acres (1.2 ha) of pine savanna at 50 years’ growth. The present industry around Bozburun turns out some 30 ships a year, that is, 1,500 ships in 50 years, and would be in equilibrium with the growth of pines on about 4,500 acres or 18 square kilometres of pine savanna, much less than the actual area of pineries available.36 This is quite intensive shipbuilding by ancient standards, and the calculation (however rough) does not support scholars’ obsession with shipbuilding as the great consumer of trees.
Timber did not have to be local: shipbuilders are the best placed of craftsmen to import their materials. As ancient Athens or early-modern Holland bear witness, a country can be a naval superpower without growing its own ship timber.
Shipbuilding was not a constant consumer of trees. Medieval cargo ships were mostly about the size of a Lowestoft or Bozburun ship. Between 1295 and 1348 the cost of timber and plank amounted to some 30 per cent of the cost of building a ship; in the fifteenth century this fell to 17 per cent. The fifteenth century began an escalation in both merchant and naval shipping that continued at an exponential rate until the mid-nineteenth. Cargo ships came to exceed 300 and then 1,000 tons. Specialised battleships were built, such as Henry V’s mighty Grace Dieu of 1,500 tons, now buried in the mud of the River Hamble.37 Ships went to other continents and shortened their lives in tropical waters being eaten by shipworms. They were built to carry great guns or to take cargoes of railway locomotives through a Cape Horn gale. The quantity of English-owned shipping increased from 50,000 tons in 1572 to a million in 1788, and continued to increase until the mid-nineteenth century.38
There were setbacks: the Mary Rose was less than half the size of the Grace Dieu a century before, and the largest ships on either side in the Armada (1588) were little bigger. However, by the late eighteenth century, although the biggest ships were much bigger, the Navy had more than a hundred battleships as big as the single Grace Dieu or bigger. Had there been the slightest difficulty in finding timber for the ships that defeated the Armada, it would have been utterly impossible to build the fleet that defeated Napoleon.
Medieval ships usually had one huge mast, built of many oak timbers around a pine core, held together with rope belts and great iron hoops. The towerlike mast of the Grace Dieu was at least 190 feet (60 metres) high. From the late fifteenth century ships acquired multiple masts and topmasts, and masts began to be made from single giant conifers.39 The provision of masts was more critical than of hull timbers; they were brought from America in special ships.
Complaints about shortages of timber came from the Navy: commercial shipbuilders, although they built far more shipping, seldom had any difficulty. The Navy was short of money rather than short of trees: it needed large sizes and special shapes of timber, but without paying extra for them. And there is a strange discrepancy between the amount of timber going into naval dockyards and the amount coming out as warships. The hull of a regular 74-gun ship ‘needed’ at least 3,000 loads of timber at 50 cubic feet to the load. This would be a stack of solid timber bigger than the entire volume of the hull, leaving no room for officers, men, guns, powder, shot, salt beef, or room to swing a cat. A rough calculation, based on measured sections, shows that the actual timber content of the ship would be at most 500 loads. Maybe as much again was lost as sawdust and offcuts: what happened to the other 2,000 loads? It was not there to build the ships, but to pay the men. Until 1801 shipwrights kept the ‘chips’ as a lawful perquisite; some chips were very large; and Deptford still has well-built timber-framed houses made from them. When the Navy Board offered gold instead, the men went on strike.40 (There was a similar discrepancy, and a similarly privileged labour force, among Portuguese naval shipbuilding in Brazil.41)
While complaining of shortages, the Navy only very slowly took to using foreign timber, species other than oak, or iron components, or moving the dockyards to the Empire as the Spaniards and Portuguese did. John Ramsbottom, the mycologist, stressed how bad ship design and poor seasoning resulted in dry and wet rot, wasting timber in repairs. The limiting factor was probably not the growth of timber, but the seasoning: ‘had the nation faced another 20 years of war the domestic timber supply would have performed far better than it had between 1793 and 1814.’42
Other countries had curious statutes regarding shipbuilding timber. In Spain there was ‘an ancient law by which the king was proprietor of every tree which his officers judged fit for any purpose of naval construction’. In the territory of Venice anyone who let his trees reach shipbuilding size risked having his woodland banned to the State by sound of trumpet; the trees and their successors were confiscated, and he might have to provide free labour to transport them. I cannot say whether this really discouraged the growth of big trees.43
Roughly half the timber shipping ever built in Britain was between 1800 and 1860. Only at this time did the price of oak timber (but even more of oak bark) rise enough to affect nearly every wood and hedge in Britain. This market, lasting only 60 years, began the dissociation between trees and local uses. Even so, lack of timber was not a constraint on shipbuilding. An example from the multitude of cargo ships is the Jhelum, 430 tons, built at Liverpool in 1849 and now stranded in the Falkland Islands. The hull timbers seem to be still all English oak, although much of the planking was foreign and there were many iron components.44
‘Old ships’ timbers’ in timber-framed buildings are somewhat of a canard. Land and ship carpentry are not interchangeable: ship components were more easily cannibalised into other ships. There are a few authentic examples of ships’ timbers and masts being used in late carpentry on land, especially in Norfolk and the Channel Islands, where shipwrecks would be a source. M. Sinclair tells me that the straight timbers out of barges were used in houses along the Severn. Bits of boats are commonly met with reused in waterfronts.
Dugouts
What can you do with a big tree other than make it into a dugout boat? Logboats were used in Europe from the Mesolithic into the nineteenth century, and are still in use in many tropical countries. I have travelled in a dugout boat, 3 feet 6 inches (107 centimetres) in the beam, hewn from a tree of somewhat greater diameter.
Logboats are usually flat-bottomed like a punt to be more stable. The sides may be heated to make them flexible, and then forced apart to give more capacity. Logboats can have a mast (with integral step), but one of their limitations is the difficulty of providing a keel. Logs, moreover, tend to crack, especially at the ends, making fissures that are difficult to caulk.
Dugout boats give a clue to the largest trees available. Those excavated in Britain range in date from late Neolithic to medieval. Nearly all have been identified as oak, which is not an obviously suitable tree: it is hard, heavy, liable to split, and liable to weak spots and holes where dead branches have rotted. In other countries more suitable trees have been used, such as lime; in early-modern Greece sea-going logboats were made of giant white poplar trees. Maybe oak logboats are better known (and logboats better known than other types of boat) because they are more often preserved and recovered.
The biggest single timber known from any period in Britain was the Brigg Logboat, made in the Bronze Age from an oak log 48 feet (15 metres) long and 5½ ft (1.7 metres) in diameter at the small end (without sapwood). (The Hasholme Logboat, Iron Age and slightly smaller, is on display in Hull Museum.) In Scotland, although some logboats were 4 feet (1.2 metres) in diameter, most were less than 2 feet (60 centimetres), and at 60 feet (20 metres) long or more were disproportionately slender.45
Dugouts also include coffins, early church chests and many other kinds of container.
QUESTIONS TO ASK OF TIMBERS IN BUILDINGS
(AND ALSO EXCAVATIONS AND SHIPS)
TIMBER AND WOOD IDENTIFICATION
This requires a book to itself. Identification of samples of wood by cutting sections and examining the cellular structure under the microscope is well understood. Most of the commoner timbers can be identified in a clean transverse section seen with a ×20 lens.46 But the student is often presented with radial or tangential sections, or with bark, or the surface under the bark, or charcoal, each of which has its own characteristics, usually subtle and difficult to put into words. Diagnosis often has to be attempted through a layer of dirt, soot, or paint, or the ancient lichens that grew on bark (themselves of interest as indicators of acid rain or lack of it).
There is no substitute for experience: for examining logs of known trees and building up a body of knowledge and of reference samples. Here are a few diagnostic characteristics most often met with in ancient timbers.
Oak
Oak in cross section is unmistakeable. It is ring-porous: each annual ring begins with two or three rows of big vessels visible to the naked eye, followed by a darker mass of late-wood with light-coloured streaks (‘flames’) running through it in a radial direction. Every few millimetres the annual rings are interrupted by a conspicuous, light-coloured ray that runs from the middle of the tree towards the outside.
The outermost inch or so of an oak-tree consists of sapwood, tissue that performed most of the tree’s vital functions. It is much lighter in colour than heartwood and is more easily attacked by insects.
Oak is very prone to differential shrinkage. Most oak beams that contain the pith show at least two radial cracks. Contrary to common opinion, ancient oak does not go ‘black with age’ but with soot. Old oak is very variable in colour.
Baltic oak, although probably of the same species, is different from British oak. It usually occurs as planks cut in an almost exact radial plane, thus displaying the rays as a pattern (‘figure’) on the face, often treated as decoration. The annual rings are very narrow (typically less than 1.5 millimetres (more than 15 rings to the inch)) and the grain very straight and lacking knots. It is more stable than British oak and does not warp, and being slow grown is light in weight (Table 15).
Under the microscope oak is very distinctive. The late John Fletcher, dendrochronologist, claimed to be able to distinguish pedunculate from sessile oak, but his work was based on a narrow range of planted trees and has not been confirmed. Since the two species often grow intermingled, anyone investigating must collect samples of timber, leaves and acorns all from the same tree so that there shall be no doubt of its identity.
Elm
Elm in cross section is ring-porous. The annual rings have within and parallel to them a series of banded structures (as if there were ‘monthly rings’), which are diagnostic of elm and related trees (e.g. mulberry and Japanese keyaki). Rays are many and just visible to the short-sighted naked eye. Elm is tough, stretches as it dries, and does not form many shakes.
Different elms have variations on this structure. From a whole log, preferably with bark on, the big, fast-grown, straight trunks of English elm can be distinguished from the slow-grown, crooked, often bossy stems of East Anglian elms. Chair-seats – massive, fast-grown, with annual rings often ½ inch (10 millimetres) or more – are likely to be English elm.
The under-bark surface of elm may have bark-beetle galleries – big ones of Scolytus scolytus or small ones of S. multistriatus. These were formed within a year of the tree dying. The brood-gallery, from which the others branch off, was formed in a vertical direction, and is normally at right angles to the length of a horizontal timber. If it is parallel to the length of a timber, this raises the suspicion that the tree died upright – of Elm Disease?
There may be direct evidence of past Elm Disease. Vessels invaded by Ceratocystis are visible as dark streaks along the length of the log. Usually they occur in one annual ring, or several if there have been repeated attacks. Close inspection may reveal white structures, tyloses, formed by the tree inside individual vessels to block the advance of the fungus. An infected ring, followed by a sudden decrease in ring width and then gradual recovery, indicates that the tree was damaged and recovered.
Ash
Ash is ring-porous: the annual ring typically begins with two rows of big vessels, followed by smaller ones scattered in the late-wood; there are many narrow rays. It too can have bark-beetle galleries, not unlike those of elm, but made by a different beetle. It is generally pale or greyish in colour. When encountered as rods, the combination of branches in opposite pairs and smooth bark is distinctive.
Conifers
Pine is much the commonest conifer among ancient timbers. Distinction between conifers is a microscopic matter; however, all pines have resin-canals penetrating the wood structure. Any pre-1700 timber that exudes resin is likely to be pine.
‘Scots pine’ (often imported or fossil) has pronounced annual rings, but no other structures visible to the naked eye. Within each ring, the darker-coloured early wood begins suddenly and gradually fades into the late-wood.
Cypress, occasionally met with among ancient chests, is most easily recognised by its distinctive, non-resinous scent. It came from Crete, and the short, broad, oddly shaped planks speak of the ancient gnarled mountain cypresses.
Poplars
Black poplar is not uncommon among timber and planks and in the half-cylinder lids of medieval chests. The body of the chest is often pine, although I know a huge chest dug out of a massive, contorted poplar log. Its ‘woolly’ texture on worked faces and its tendency to woodworm hide the inconspicuous annual rings. A yellowish colour is diagnostic.
Aspen, met with both as timber and wood, is similar; microscopists will not commit themselves beyond a diagnosis of ‘poplar’. However, the bark of aspen, with its diamond-shaped lenticels, is very distinctive. (The phloem under the bark is easily confused with lime, both being in the form of sheet-like layers of fibres.) Whole-log timbers of black poplar may show rugged bark and bosses on the trunk.
Beech
Beech is most often encountered in late furniture. Like most broadleaved trees it is diffuse-porous, with annual rings distinguishable only by colour. The rays are well visible in transverse section; in tangential section they appear as a myriad of little brown flecks, darker than the rest of the wood.
Chestnut
In England chestnut is a phantom timber: I have yet to find any supposed ancient ‘chestnut’ that is not oak,fn3 though I am familiar with it in southern Europe. It should be looked for in regions where the tree has long been prevalent, especially among wattle rods. It is ring-porous, with tracts of visible vessels extending into the late-wood, and barely visible rays. It looks very like oak and is similar in colour range. I suspect the misidentifications are by people who see only the tangential section, in which the strong rays that are diagnostic of oak can be difficult to detect.
Hazel
Hazel appears only in the form of wood. The bark is the most recognisable part, dark brown and shiny, with raised rectangular lenticels at intervals.
Sallow
Sallow wood is one of the commonest constituents of wattle. If, as usual, it is Salix caprea, it is recognisable by short, raised ridges under the bark. Older stems develop diamond-shaped lenticels.
Footnotes
fn1 The oldest functional timbers in the world are probably the olive-wood blocks that join the columns of the Parthenon in Athens (fourth century BC) and prevent them sliding apart in earthquakes.
fn2 The Japanese foot (shaku) is nearly the same as the English foot.
fn3 At Wingham church in Kent the octagonal posts of the nave are supposed to be chestnut, but the only one that I could identify as such is a Victorian replacement.