This paper will review the shipbuilding history of the Iberian cork oak and, in passing, of other oaks. Not least it is intended to draw attention to the changes that have occurred since the period when Oliveira, and also Lavanha, spoke of its properties for shipbuilding. It has to be noted that much other evidence is less favourable towards cork oak. If some of the hypotheses put forward here can be verified by the archaeology of future shipwreck excavations, or otherwise, they may by extension and comparison throw an interesting light on for example English shipbuilding of that and earlier periods, based as it was on European oak.

Oliveira and Lavanha describe the dependence of Portuguese shipbuilding on the timbers of the cork oak tree, now generally a relatively small tree of unsuitable form for ships' keels, posts and frames, in the sizes and shapes required for a large ship in the sixteenth century. Equally, the frame timbers of a ship like the Mary Rose (1509) pose curious problems if we accept English forest history in conventional terms.

Perhaps one key point to remember is that all oak trees were a focal point of mediaeval and later European societies, dependent on timber for many purposes, and on tan-bark, fuel, and the fruits of these forest trees. The cork-oak is only clearly distinguished by its one special product, cork. It is not the intention to discuss in any detail the existence of hybrids, and related species such as the Q.lusitanica¹. The text will refer to the three main forms, cork oak (Q.suber), holm oak (Q.ilex), and European oaks (primarily Q.robur).

Another is that there is a substantial literature from the seventeenth to nineteenth centuries to make the point that a natural oak tree does not produce ship-timber in abundance; a point amply reinforced by simple observation of the form of most English oaks today - the hedgerow oak (traditional provider of ship-timber) as it now exists would rarely provide a single frame timber for the Mary Rose. Something about the management of oaks has changed, yet the literature contains little that can be verified².

The final introductory remark, indeed the starting point for these enquiries, is that during the latter part of the sixteenth century there was a marked change in the form of ships' frames: to about 1550 a typical north-west European midship section was effectively a flat floor and two quadrants - many of the frames were almost pure arcs of the same radii. By about 1570 in England at least, and subsequently more generally, the frame was a series of relatively straight sections linked by relatively sharply curved bends³. The unanswered question is why ? Timber supply has to be one of the possible answers. The documented form of Portuguese frame shapes presents similar problems - a high proportion of all major frame timbers in a ship were of the same arc. Unfortunately, what documentary records survive are essentially concerned with monetary values of timber, not matters of form or production.

The term ship-timber does not encompass all the timber required for a ship. Planking and masts are excluded; even to some extent the major pieces for keel, sternpost and deck-beams, required for any significant carpentry in buildings too. It is intended to refer to the special, usually heavy pieces required for the curved parts of the skeleton of the ship, and the crooks and knees used to reinforce all its joints, For a large ship only the trunk, and lower and leading branches would serve: the "top and lop" could be used for other purposes - often it seems harvested repeatedly, even to deliberately enhance the form of the tree. In essence, ship-timber is any piece requiring a combination of special size, geometry and grain pattern, often a circular arc of considerable length, to both form the shape of the hull and provide its strength. (Even plywood has a preferred direction of use). From the eighteenth century, encyclopaedias illustrated the ways a tree might be divided. Duhamel du Monceau above all explains the importance of this process in the supply of scarce pieces for large ships⁴. Such timbers would be distinguished from those needed for most buildings and carts by sheer size, and by the additional requirement to be resistant to both rot in water and ship-worms, especially as voyages to warmer waters increased. For practical purposes, the paper only relates to a small part of Portugal. English shipbuilding could only draw, economically, on woodland within about 25 kilometres of a shipbuilding site (or at least within that distance of water transport). At that range, the cost of the timber when bought on the open market had already doubled. There is no reason to suppose that Portuguese conditions were significantly different in this respect⁵. Much of the expanse of the Alentejo, now the home of the cork oak, is of marginal relevance here.

With such a title, it is necessary to review what Oliveira and others do tell us of the cork-oak. Firstly, Oliveira’s Livro da Fábrica das Naus, chapter 2, written about 1570-80⁶. Interpolations are marked by [..]; most quotations are heavily abbreviated:

Oliveira’s Ars Náutica¹¹ adds a few details, as for many other important topics:

In his earlier Arte da Guerra do Mar of 1555¹² Oliveira says:

Lavanha, writing on naval architecture about 1614¹⁴, uses similar material and sources, and adds little but a more polished text; except that planking can not be made from this tree, the "great naus of our long navigations" are singled out; and that Malabar teak and angelim are referred to in glowing terms as combining all naval requirements in one timber. Pliny also noted teak (Bk XVI Ch LXXX).

The Livro Náutico, more an inventory than a discursive text, records the use of cork oak and pine in the ships¹⁵.

Oliveira and Lavanha manifestly draw heavily on classical authors such as Pliny and Vitruvius; and unfortunately on their or their interpreters' errors. Oliveira actually acknowledges the point in Livro da Fábrica das Naus. At the same time they purport to advise on identification of useful timbers in new lands (and many timbers of Madeira were "unknown in the kingdom" - Dias Leite). What Pliny's Natural History actually has to say about cork oak is rather limited, not least by the fact that Iberia is not mentioned, and the trees he is describing may be expected to differ from place to place - Oliveira and Lavanha are both careful to point this out as a generality; while they are not averse to relying on the Italian writers for specific cases, as for cork oaks. This is an oddity, because the Italian and Greek cork oaks attracted few compliments. Lawson, a later English writer comments that "I admire and praise Pliny, Aristotle, Virgil, Cicero and many others for wit and judgement in this kinde, and leave them to their times, manner and several countries"¹⁶. Pliny says of oaks:

Theophrastus (Enquiry into Plants, I) has a different account, based on the case of Tyrrhenia, and may conceal critical evidence, in addition to confusion of species: "...a tree with a distinct trunk and few branches [which description, compared with the photographs of Vieira Natividade, strongly suggests that the trees Theophrastus knew were already systematically pruned], and is fairly tall and of vigorous growth. The wood is strong... The tree is not evergreen, but deciduous. It has always¹⁷ an acorn-like fruit like that of the holm-oak" (Bk III Ch XVII).

Evelyn, writing in 1664¹⁸, says that cork will grow "in the coldest parts of Biscany and in the north of New England: why should we despair ?". The uses of the bark were well known but "the timber is else inconsiderable".

Haddington tried to grow cork oaks in Scotland, as they had been "much praised of late"; he had been told that they "grow large enough to be fit for ship timber - but whether they'll do so in this country, I shall not pretend to say.....I have raised a great many of them from the seed, but they are ill to transplant, and the vermin are very fond of the acorn... I am not fond of them"¹⁹.

Not all Portuguese sources are as enthusiastic about the merits and significance of cork oak. Jerónimo Dias Leite in his work on Madeira²⁰, writing about an indeterminate period perhaps nearer 1420 than 1450, implicitly provides conflicting evidence. The literal interpretation is that before timber from Madeira was available, the Portuguese could not build larger ships (defined as having topmasts, which might only mean as small as 50 tons), while authors of his own period claim that Portuguese shipbuilding was founded on cork oak and pinhos bravo and manso, all ideal for their ships (up to more than 1,000 tons). None of these timbers were available from Madeira. This interpretation is simply incompatible with the specialised shipbuilding texts of Oliveira and Lavanha.

No oaks grew as natives on Madeira, and this account is actually restricted to straight timbers; saw-mills are of limited value for ship-timber, and hard timbers are not generally the best for either plank or mast. It seems more likely that the water-powered mills and water transport to the Ribeiras were actually the critical factors; coupled perhaps with the reduced cost of importing very large masts, otherwise only available from northern Europe. It is precisely the large ship-timber that presents and remains the problem: it did not, apparently, come from Madeira, and so was already available. In theory, so was plank and mast - if only from Leiria, planted from the early fourteenth century specifically for shipbuilding use. The explicit link between the supply of timber from Madeira and the ability to build larger ships in Portugal is probably hyperbole - and the larger ships still relatively small.

An alternative interpretation might be that the larger ships were soon built (albeit partly with timber from Madeira) precisely to transport the quantity (pela muita) of great timbers of Madeira (and thus larger may equally mean longer). This would explain away much of the difficulty, and is a doubly tempting hypothesis if we consider the extraordinary prominence given to João II's "sala de madeira", constructed for ceremonies in Évora in 1490, which attracts more attention in Resende's Chronica²¹ than all shipbuilding and ordnance, in a reign supposedly noted for these latter. The date of this individual structure is too late to explain Dias Leite' text, but the timbers referred to were brought from Lisbon at short notice, so were probably imported as a norm: "... the king ordered a new hall of timber to be made with great ingenuity and artifice, and a great thing... of length three hundred palmos [probably 66m], breadth 75 palmos, and height 72 palmos. It was fitted out with walls on great and strong masts that were brought from Lisbon at great expense, and between the masts by walls and screens, and fitted out above with slender masts and other timbers, and covered with cleft planking [trincado: clinker] and caulked and tarred as a wooden nau, so that not a drop of rain water could get into it". Rui de Pina²² suggests a relatively temporary structure, a vast tent, but on the same great masts.

At the same time, it is interesting to note that many of the species identified as available from Madeira from an early period are not widely recognised as shipbuilding timbers (and significantly were not cited by Oliveira or Lavanha either). The barbusano, pau branco and til were used, and perhaps the cedar (though this is a juniper, not the famed Cedar of Lebanon). To what extent this is simply the result of over-exploitation of mature specimens is unclear, and not resolved by commentators on the timbers of Madeira - the matter will now await archaeology. This writer has the impression that the famous vinhaticos, loureiros, etc were actually used not for shipbuilding, but for house building and furniture. There are not even any pines in the lists of natives; ordered to be planted only from 1515²³.

The other early comment on timbers of Madeira is from Cadamosto, dated 1455: "... sawmills continually working timber and plank of all kinds for the supply of all Portugal and elsewhere. Of this timber two kinds are esteemed; one cedar, which is very odorous and resembles cypress. Very beautiful tables... chests and other furnishings are made from this. The other kind is yew...."²⁴. Again, there is no suggestion of ship-timber.

Correia offers a parallel - three and four-storey buildings also followed the introduction of timber from Madeira²⁵. Yet Damião de Gois speaks in 1554²⁶ of the buildings at that time "with all interior walls and arcades, covered throughout with timbers from Samarcia" - probably cleft-oak panelling from the Dantzig area. Thus the contradictions abound, but there is no clear evidence to support the usual reading of Dias Leite. Even the statement that there were previously nothing but caravelas and barineis is contradicted by other chronicles: Affonso III had fleets of navios grossos and galleys used in the capture of Faro, and against the Moors generally, around 1270²⁷.

We may insert at this point the interpretation of Pimentel Barata²⁸ that the Portuguese used cork oak exclusively for ship-timber. This is apparently an act of faith based on the preferences of Oliveira and Lavanha. Apart from the reality that ships were built with what was available; and the admissions of Oliveira and Lavanha to that effect, it is impossible to reconcile the statement with the facts of economic history. These clearly reveal that all sorts of timber were imported into Lisbon and Portugal in the sixteenth century, including European oak: the reality clashes with the quasi-classical commentaries, in their turn.

Evidence from economic and forest history is presented by Leonor Freire Costa²⁹, who documents the continuing abundance of cork oak along great stretches of the Ribatejo during most of the sixteenth century, explicitly supplying the Ribeira in Lisbon. This was more or less protected by edict from use as fuel, and the supply crisis only arose towards the end of the century. There is one telling phrase in a document of D.João III for 1546, intended to preserve the cork oak for shipbuilding. What was prohibited was the cutting of cork oak "at its foot" (seu corte pelo pé) for use as charcoal, which was evidently happening commonly. This qualified phrase could be interpreted as evidence that it was perfectly legitimate to cut it otherwise, as part of regular pruning and management, for fuel and/or to improve its condition for some purpose, whether ship-timber, acorn crop, or cork.

Vieira Natividade is cited as describing the Alentejo as the primary cork oak area, and Freire Costa notes the possibility that its relative absence from the documents was due to the lack of water transport to exploit the timber. The pine forests of Leiria, reputedly planted specifically for shipbuilding purposes by D.Dinis in the early fourteenth century, should have been providing mature pines throughout the period of our interest, and indeed attract later comments about their indispensibility for the continuance of the India shipping.

This account broadly serves to confirm the statements of Oliveira and Lavanha - with one conspicuous exception. The schedules presented (for the period 1525-1540) of timber imported to the Ribeira, or used on specific shipbuilding tasks, flatly contradicts any notion of native sufficiency in timber and/or that European oak, carvalho, was not used in Portuguese shipbuilding. Timber and plank of carvalho are almost as frequent as those of sovaro and pines. We find European oak for knees, as heavy planking, and as planking for the bottom of the naus; it might be local, or from Flanders or Galicia. As Lavanha stated³⁰, "we are forced to use what grows in this land": another way of saying that ships have always been built with whatever timber was available to the builder. In Portugal, it is quite clear that the balance of availability and price commonly led to the use of European oak, as everywhere in Northern Europe. Its alleged defects did not actually prevent other European ships working in tropical waters.

As a sidelight on this issue, there is a record that in 1506 three ships were completed in Holland for the Portuguese crown, specifically for the Calicut voyage³¹. This is interesting on two counts. The live-work of their hulls would probably have been built substantially of European oak; and beyond that, it is by no means certain that they would even have been carvel-built. And this after half a century of experience of European timbers and building methods in tropical waters.

Incidentally, this topic produces a persistent feature of several recent Portuguese works: an unsourced insistence that northern European oak was a normal and suitable provider of timber for masts³². Nothing could be further from the truth, in general: except that two small examples have recently been found in Basque wrecks.

It is not safe to assume that the current distribution of cork oaks reflects that at the period under consideration. The present centre of economic production is the Alentejo, and as Freire Costa demonstrates from the records of supplies to shipbuilding there was a heavy dependence on the Ribatejo for cork oak for shipbuilding (and fuel), which would in time have depleted those forests. However, Amorim Girão³³ presents qualitative toponomical evidence for the former distribution of cork oak, which is the very inverse of the present situation. The greatest density of place-names associated with cork oaks is actually in the northern regions, not the Alentejo. Quite exceptional specimens are still found widely in the north. He also reported vestiges of unexploited ancient stands in the extreme north, and drew attention to similar vestiges of ancient cork oak forests in the coastal belt from Peniche to Ovar in 1940, where forest cover had long been primarily of pine.

Reis Gois, writing in 1955³⁴ described the processes by which a massive increase in cork oak stands has occurred south of the Tejo, in this century. The older natural cork oaks had often been removed as prejudicial to agriculture, with a general policy favouring cereal production. However, the low fertility of the soils was quite inadequate, and the attempt was often abandoned. The ploughed land actually favoured the germination of evergreen oaks within a dense new scrub, and when after a dozen years of stunted growth they finally over-reached the scrub, they grew vigorously and came to dominate. Many areas have also been artificially planted, mostly by seeding, and especially since 1935, with coverage dramatically greater than at the turn of the century.

The eighteenth century offers contradictory information: notably French, and most reliably of all, Duhamel du Monceau in 1764³⁵:

Let us now consider the figure of a modern cork-oak tree³⁶. It ideally has a straight trunk of 2-3 metres length, and two to four limbs equally spaced and growing at about 45 degrees to the horizontal. These limbs are short and straight, and then in turn divide into branches of a size not likely to be much use in the construction of large ships. This form is achieved (and equally, only maintained) by rigorous pruning of the individual trees at regular intervals. There is no long straight trunk for keel or beam. The range of ship-timber that such a tree can provide is extremely limited: no compass timber for frames, but there is a selection of potential crooks, and knee-timber from the combination of trunk and one key limb, by cutting largely to waste. There is even a remarkable similarity in form between the profile of these trees and the floor timbers drawn by Oliveira, Fernandes or Lavanha, sometimes with a surprising sharp transition at the bilge, sharper than the expected tangential sweep of other traditions for large ships.

We must note however that these modern trees are not naturally grown, but the product of intense selection over two centuries, and regular and severe pruning. They are not natural forest trees, but fruit trees, yielding a regular and valuable crop - indeed three crops; the prunings are themselves firewood, and the acorns as valuable as the timber. In the related case of the azinho - holm oak, the balance is shifted in favour of the acorn crop across Iberia, but the same basic point holds true. They are tended as any other fruit trees.

If we now consider the trees known to Oliveira, it is probable that they were managed rather differently. The firewood and the acorns were probably far more critical to the economy: until around 1800 there was no industrial utilization for bottle stoppers or other manufactured products. There was of course an ancient cork trade for footwear and floats. Indeed D.João II seized the monopoly of trade to Flanders from 1484 to 1498, and with the proceeds paid for the copper with which a new generation of bronze artillery was cast³⁷. This move was arguably as significant for Portuguese expansion as the contribution of cork oak timber to Portuguese shipbuilding.

The trees were relics of ancient forests or montadas. They would have included far less regular examples, displaying the ravages of wind, lightning, insect attack and disease. Trees with damaged leaders or major branches will alone of "natural" oaks produce the compass timbers for large ships of which Oliveira speaks. We might however note that the frames of the ships he knew did not, as far as texts tell us, exceed one palmo (256mm) in section. This must represent the use of branches for frames, and the idealised modern cork oak would not produce such compass timbers.

It is well documented for the case of Spain especially that the cork oaks were widely sacrificed to produce tan-bark for short-term profit. The period of most intense activity extends through the nineteenth century, until chemical substitutes became available³⁸. The critical difference here is that the cork was a secondary product, and the stripping of the cambium for its high tannin concentration actually killed the tree. There is some suggestion that over much of southern Portugal the bark was exported at this time. If so, it helps to explain why older trees disappeared over large areas: past their prime for cork production, and so felled for the tan-bark, and charcoal, or short-term gain. Rackham even suggests that in most current cork oak areas the trees are first generation - less than 150 years old³⁹. A parallel explanation is that many of the plantations date from the nineteenth century, when demand for cork increased dramatically - production increased four-fold between 1885 and 1924⁴⁰.

Vieira Natividade cites historic data for the air-dried density of cork oak⁴¹ which is at variance with modern data, for whatever reason - there are many variables behind such data. This suggests densities far higher than typical European oak, and often heavier than water. However, modern data records typical figures as low as 0.75 for cork oak grown in an undisturbed state; 0.82 where commercial cork extraction and pruning are practised on around a 10-year cycle (there being a direct relationship between the "crisis of cork removal" and density in cork oak, contrary to the almost universal inverse relationship in other oaks)⁴². Does this suggest intense pruning and at shorter intervals for firewood or otherwise in the past ? An explanation has to be found to complete the history of cork oak for our purposes.

Any regular pruning to yield local firewood (and in the process re-invigorate growth and acorn yield) would also improve the size and quality of the branches that were left, even if the rate of growth of the tree as a whole was reduced in the short term by reduction of the green canopy. (It seems remarkably difficult to find any quantification of the rationale for pruning any trees, but the classical pantheon is full of Priapic gods - Evelyn refers to an un-named, untraced, goddess of pruning - with pruning shears, centred on pruning the vine). At the same time, stress on the root system would be reduced, producing a healthier and perhaps longer-lived tree. Loss of part of the limbs and canopy in a dry year, as a result of a tree growing beyond its root system, especially in poor ground, would introduce decay at an earlier age, so that the tree never grew the scantlings desired. This sort of argument is a major feature of numerous English texts from 1600-1831, proposing better ways to grow ship-timber. What is not clear is how widespread the practices were at any time, or how deliberate.

Old oak trees are notoriously liable to rot in the heartwood. A giant such as the Major Oak in Sherwood Forest (Robin Hood's tree in myth, some 800 years old), may have a girth of nine metres, but it is totally hollow. The tendency starts before the tree reaches 200 years. The economic balance has long been recognised: leaving the tree to mature and yield larger, more valuable timbers brings with it increased risk that the heartwood will rot, and render the whole tree into so much firewood - or jeopardise the longevity of the ship into which it is incorporated. Evelyn expressed this as "a timber-tree is a Merchant Adventurer, you shall never know what he is worth, till he be dead"⁴³. Nineteenth century English sources are perfectly clear as to the marked difference in longevity between large ships, dependent on large timber, and smaller ships, utilising much younger and healthier trees. Regular and certainly any careless pruning will increase the tendency to rot, by allowing rot and infection into the tree. Pruning branches above 50mm diameter, or old trees, is not recommended; removing the epicormic buds is preferred, but impracticable in forests. Michie sums up many of the issues in what is a vast professional literature: "If well directed, pruning is one of the most useful, and if ill directed, it is one of the most mischievous, operations that can take place in forests"⁴⁴. One seventeenth century writer describing the practice and merits of pruning forest trees put it as:

"Reader if thy faith hold out, read on;

but if you find you can't believe, be gone".

There are also numerous descriptions extant, from the mid-eighteenth century on, of actively training forest trees to required forms for ship-timber in England, France and Italy, but space precludes any details here.

This study of cork oaks and their use in shipbuilding has inevitably suffered from the fact that it is now well over a century since their use ceased in significant shipbuilding, and a quarter of a millenium since the last trees to be used were actively nurtured. (Any study of current boatbuilding must be conducted on a local basis, not from afar). Severe difficulties with the source texts have been discussed and new hypotheses put forward for their re-interpretation. The particular history sought is one of rural history and essentially craft skills, invariably under-recorded.

It is however conspicuously the case that cork oak, a forest tree, a key structural component in major shipbuilding, has been shown to be affected by and even dependent upon regular pruning; and that there are complex interactions in this history with other economic uses of the trees, and with transport issues, so that we can be certain that the balance differed from one period to another, and also from one location to another, perhaps only ten miles away.

Portugal undoubtedly relied to a significant extent on imported timber, despite polemics against northern timber and fastenings - the link of particular interest may prove to be that to the oak forests of north-west Spain.

The northern oaks had many features in common with the evergreen oaks, not just as individual trees, but in their forest environments and their range of products in timber, fuel, bark and food. The balance differs, but the evidence adduced here provides a fascinating parallel to English and French forestry at least, for which the written evidence has a different emphasis, and commentaries substantially, even systematically, ignore practices that are fundamental in areas traditionally supplying shipbuilders. The issues of pruning and training especially of oak trees for shipbuilding use, and the use of roots, are clearly worthy of further investigation based on the new evidence that is at last emerging from underwater archaeology.

Dr. Inácio Guerreiro, W.G.L.Randles, and the staff of the Direcção-Geral das Florestas, have all kindly assisted with the provision of source materials for this paper, but are in no way responsible for any errors.