This section of the Birlinn site contains articles on the Birlinn’s design and construction.
The following notes were written in the late 1980s following fieldwork by Alison Gunn (now the Rev Alison Jaffrey) and the author. They have not since been altered significantly beyond formatting and have therefore been superseded in places by later publications, but are included in their entirety for completeness.
Hebridean Galleys and their History
THE HEBRIDEAN ENVIRONMENT
The climate of the West Coast of Scotland is, of course, windy and wet. That obscures a great deal of local variation. The Outer Hebrides are the windiest place in Europe, and the wind there is almost never absent. It has a steady, unremitting power that can keep all small craft in shelter for weeks. In the Inner Hebrides and the inland lochs there is in general much less wind, but especially in the lee of the mountains of Mull and Skye it is capable of greater destructive violence, when a gale is torn into vortices and white squalls by the peaks. Gales come most frequently from the South and West, although a major one may start in the SE and finish in the N.
The region has a reputation for mist which is not altogether justified, except in a few areas like the lower Firth of Lorne and around Ardnamurchan Point, where fast-running tides over a broken bottom bring cold water to the surface. In April and May there is often a period of east wind and hazy weather, but usually with a visibility of at least half a mile or so. Otherwise mist is not much more common than elsewhere around the British Isles, but visibility is often reduced by rain. The highest areas of the mainland, Skye and Mull attract the rain most, with 4-5m a year, but even in those parts of the Inner Hebrides where the rainfall is much less there are many days of thin soft rain that cuts the visibility as much as a downpour. The low islands – Tiree, Oronsay and Iona for example – are well known for escaping cloud and rain, and although records are not available the same is visibly true of the wider areas of sea amongst the islands.
In midsummer there is little real darkness in clear weather. Even though many important routes remain unlit today there is little difficulty in making night passages in reasonably open water. Visibility at night is very dependent on the light of the sky reflected in the sea, against which land, rock and other vessels show black (hence the preference of local fishermen for tanned sails to avoid being run down). In consequence it becomes very dark close under the land, where the hills are reflected in the sea. It would be possible for a galley to remain very difficult to see, just by keeping under the shadow of high ground.
The main seaways are the lochs running into the mainland and the larger islands, and the network of sounds between the islands. From the Firth of Lorne southwards, although not elsewhere, there is a strong NE-SW trend to both land and water. The longer lochs are of glacial origin, mostly deep and steep-sided with relatively few areas where depths are suitable for anchoring. The wind tends to funnel straight up or straight down them. In most cases they dry out for some distance at the head, and often moraines have partly blocked the entrance to leave it shallow and with fertile ground on either side. Sometimes the blockage has been complete, leaving a long freshwater loch accessible from the sea by a river or a short portage. The sounds, on the other hand, vary greatly in width and character. They are not usually as deep as the main lochs, and tend to have lower ground, many glens where the wind comes down, and frequent anchorages at the sides. The tidal range is nowhere excessive, varying from about fifteen feet in the N to only about four feet in the Sound of Jura. In general the flood tide sets north and the ebb south, mostly at about 2-3 knots springs. In a few bottlenecks on major seaways it runs much faster, in the region of 5 to 8 knots. Some of these are at the narrows at the mouths of the mainland lochs, the Falls of Lora at the mouth of L. Etive being an impressive example. More important are the narrowest parts of the sounds, particularly the Sound of Islay; Caol Rhea and Kyleakin between Skye and the mainland; and the Cuan Sound, Fladda Sound and the Dorus Mor on the inside route south from the Firth of Lorne. These are not ususally attempted by small boats against the tide. A headwind in these places is a serious obstacle because it throws up a short, violent sea when a favourable tide runs against it. There are numerous areas of small but vicious overfalls in the places mentioned above and elsewhere, like ‘Maclean’s dirks’ off Duart Castle.
Except for the inhospitable West coast of the Outer Hebrides, most Hebridean waters have some degree of shelter from the West. In heavy SW’ly weather, however, the Atlantic swell curls round and comes up many of the main seaways from the S., giving unpleasant conditions in the Firth of Lorne, the Minch and the Sounds of Sleat and Jura when the ebb is running against it. In these circumstances there are also dangerous tide-rips off the more exposed headlands, and these very frequently hold up vessels on passage up or down the coast. The most awkward today are the Mull of Kintyre, Calliach Point on the West of Mull, and Ardnamurchan Point. The passage round the Rhinns of Islay is worse than any of these, with a fast-running tide and heavy overfalls. It is little used now, but being on the direct route to Antrim it must have carried much traffic in the Middle Ages.
In comparison with, say, the Southern North Sea the West Coast of Scotland is certainly a deep-water area, but less so than West Norway. There are in fact a number of useful passages accessible only near high water to a boat of 1m or so draught. The most important example is Clachan Sound inside the Island of Seil, which avoids the relatively exposed and hazardous Sounds of Sheep and Fladda. It could also be useful on occasion to cross the drying channels between Colonsay and Oronsay and through the Uists, or to leave Tobermory Bay by the back door over the Doirlinn of Calve. Also, before the introduction of reliable charts and sailing directions many passages now taken for granted in vessels drawing 2m or so must have been hazardous without local knowledge. It would be very difficult for a conventional sailing ship without a local pilot to navigate the shallow sounds of Gigha, Iona and Raasay, or to penetrate from the Sound of Jura into the Firth of Lorne without risking the Gulf of Coirebhreacan.
Some of the Island coasts, without any shelter accessible to a boat of normal draught, are very suitable to one drawing only about 1m, especially if she can be hauled out on a sheltered beach if required rather than lying to her anchor. Perhaps the best examples are the islands of Tiree and Oronsay, fertile and rich in antiquities, lacking good anchorages but with plenty of sheltered nooks where a boat can be pulled up. The same could be said of much of the coastline of Islay, the west side of Kintyre and many other places. Indeed the whole geography of the region becomes different when deep water is not a requirement. This was as much the case in the 19th century as in the Middle Ages; the present day is the exception.
TYPES OF EVIDENCE ON WEST HIGHLAND GALLEYS
Contemporary Stone Carvings. Over 80 carvings of galleys and smaller boats have been recorded, mostly on grave-slabs and mostly in Argyll. At a rough estimate, considering the number of grave-slabs now worn out of recognition and the general standard of preservation in the graveyards where they are found, this may represent rather less than half of the number originally carved. At any rate the percentage surviving is sufficient to ensure that the survivors are fairly representative. Because they mostly lie at ground level in the open air, the grave-slabs tend to become covered with turf and not all that have been recorded in the past are visible at present, but it is probable that clearing turf in may places would reveal more. Although the stone used is hard and durable the preservation of fine detail is very variable and more than half of the stones, although they clearly show a galley of some sort, are too worn for any useful information to be obtained. Stones preserved below the turf tend to be in better condition than those exposed to the weather. The Rodel carving, which is on a softer stone than usual, would not have retained its beautiful detail without the protection afforded by its indoor location. This carving is a unique resource and its protection is a priority for the future of the subject.
Records of some sort, i.e. photographs and drawings, were obtained of all but about 10 of the known carvings, largely from my own fieldwork. Published photographs were used where suitable. However published drawings are less accurate in representing hull form than other details, and are generally not usable. The drawings by White are particularly unreliable in this respect. Photographs held by the Royal Commission in Ancient and Historical Monuments for Scotland are excellent but the published versions in their inventories have in many cases been reduced too far for details to be confirmed, and few of the originals are now accessible. The photographs of latex casts used to record the Islay stones are quite satisfactory. Photographic recording of stone carvings in the open air is an art that I have never mastered, and in most cases detailed drawings and measurements done at the site were the preferred form of recording in my own fieldwork. Rubbings were avoided to prevent damage to the stones. At a late stage a more effective method was devised: a sheet of clear perspex was placed over the carving and used as a support to trace the outline onto clear polythene film using a permanent marker.
The dimensions and details of hull form, construction and rig have been computerised for all the carvings on which data were available and some simple statistical analysis has been done on them.
Graffiti on Stone. These are fewer in number and variable in quality, but useful because they are unconstrained by a formal design or style. They have been recorded similarly to the carvings and are considered more useful for estimating hull proportions.
Viking-age and Present-day Scandinavian Parallels. These have to be the main source of evidence on construction. Experience with the rigs of modern and replica boats in Scandinavia, especially at Roskilde, is also very valuable: but the West Highland carvings are in fact a much better source on rigs than anything available from early Scandinavia and show some differences in detail – they seem to be rather simpler.
Oral Tradition. There is a considerable number of Gaelic folksongs, and some more formal bardic material that are contemporary and mention galleys. This material gives a clear impression of the galley as a symbol of wealth and power, and a few minor details on sailing and beaching. The Birlinn Chlainn Raghnaill of Alasdair MacMhaighistair Alasdair may be included here and, although problematic due to its very late date, it is the best source we have on how the boats were crewed, rowed and sailed. The undated ‘Advice of Donald Gorm’ is cryptic and of rather doubtful reliability, particularly with respect to dimensions, but it gives interesting and plausible details of hull form.
Accounts. A set of accounts for the construction of a birlinn is given by MacLeod but is very late. There may be some Campbell records but if so they are even later.
Charters. A considerable number of charters to land in the West Highlands require the upkeep of a galley or smaller boat. This appears to have been a formalised system related to land valuation and should, when analysed, give an idea of the economic basis required to build, maintain and crew a galley.
Coastal Field Recording. Examination of the siting and harbour structures of most of the 50 or so castles on the West Coast of Scotland has shown clearly that they were sited and arranged primarily for warfare involving boats, and as well as being a novel insight into castle archaeology this gives useful on the way in which the boats were used and beached.
Place-names. A list of place-names involving the Gaelic words for galley, birlinn etc has been constructed and gives complementary information to the field recording of castles.
DIMENSIONS AND PROPORTIONS
The galleys shown on the grave-slabs vary from 14 to 34 oars (strictly speaking the smaller ones would be birlinns rather than galleys). We may reckon the length as about 1m for each pair of oars plus 4-5m for the ends. In attempting to decide how accurately the galleys were carved, we have to consider the proportions separately from the details of construction, rig and decoration. The main question is whether they were as light, flexible and beautifully constructed as the earlier Viking ships, or whether they were relatively short, deep and heavy as the carvings suggest at face value. Even within the Viking tradition there is a great deal of variation, for example between the solidly built knarrs and the light, narrow snekkr.
To put this in quantitative terms; the Gokstad ship and the faering found with her are more or less average in lightness within the Norse tradition and close to each end of the size range that interests us.
| Length overall (L) | Displacement (D) | 3Ö D/L | |
| Gokstad ship | 23.3m | 20t | 0.12 |
| Gokstad faering | 6.6m | 0.3t | 0.10 |
Interpolating between these extremes, we can estimate the displacement of galleys of intermediate size if they were built to the same standards;
| Length | Displacement | |
| Galley of 24 oars | 17m | 8.5t |
| Birlinn of 14 oars | 12.5m | 3.4t |
| Eight-oared boat | 10m | 1.3t |
It must be emphasised that these displacement figures are very light indeed. For example the Shetland boats, which are far from crude, have about double the displacement on the same length. The scantlings of the thirteenth-century English galleys sound much heavier still, suggesting perhaps 4 times the above displacement. A modern Scottish fishing boat has about 6 times these figures.
Evidence from the carvings.
As with any medieval representation we cannot be certain that either details or proportions are correct or contemporary, without other evidence in support. Whenever human figures are shown aboard a galley they are to a far larger scale than the rest of the boat. This makes it quite clear that the carvers were not unduly fussy about realistic proportions. The carvings of the galleys are normally fitted into a rectangular space, limited horizontally by the edge of the stone and set within a composition where the main lines are vertical. This would possibly encourage some foreshortening, with an exaggeration of the vertical lines of stem and stern; at least, that is more likely than the reverse.
The D:L ratio (depth amidships : length overall) is the simplest criterion by which to describe the proportions of the carvings.
| Depth/Length ratio | |
| Galleys, mean of 12 published examples | 0.18 |
| Galleys carved emblazoned on shields, mean of 2 | 0.26 |
| Creagan and Kilchattan graffiti of galleys, mean of 3 | 0.10 |
| Norse ships: Gokstad | 0.09 |
| Oseberg | 0.08 |
| carved on Gotland stones, mean of 2 | 0.16 |
| Bayeux tapestry, mean of 5 | 0.14 |
| Skuldelev I (cargo vessel) | 0.13 |
| Bremen cog | 0.21 |
| Ships on C13 English burgh seals, mean of 3 | 0.20 |
These figures show that the galleys as shown on the grave-slabs were much shorter and deeper than surviving Scandinavian vessels, even the Skuldelev I boat; but that foreshortening was by no means uncommon in pictorial representations of other kinds. In the Kilchattan and Creagan graffiti, which were not fitted into a restricted space, the galleys appear much longer and lower. To summarise, the carvings suggest relatively short, heavy hulls, but their evidence is of little value on this question.
Evidence on construction
The construction of the galleys will be discussed in a later section. At least some of the galleys had winged stems and oak planking. Oak, although a dense timber, is well suited to a light displacement clinker-built boat because of its resilience to splitting across the width of the planks. This is particularly necessary in the Scandinavian style of construction with widely-spaced frames.
These constructional details imply high-quality construction and exceptionally skilled selection and seasoning of oak in at least some of the galleys. The necessary skills are not likely to have flourished without being based in a tradition of some length. But this evidence only suggests that West Highland builders were capable of building to Viking standards of lightweight construction, not that they actually did so; the skill could have gone into strength and longevity instead, in boats of much heavier displacement.
Requirements for hauling out and portaging.
Folksongs and the siting of contemporary castles imply that the galleys were often – perhaps generally – hauled out rather than kept afloat. The beaches and slipways in question are mostly quite steep. The displacement figures suggested earlier may be compared with about 10t for the first class luggers working out of the port of Deal (up to 20t with a full load of fish). These were the largest British craft in recent times to be regularly launched and recovered from a beach (E.J. March. Inshore Craft of Britain in the Days of Sail and Oar. Vol. 2. Davis & Charles,1970). They were drawn up greased ways by quite large and well-organised teams using fixed windlasses above high-water mark. It is correspondingly more difficult to envisage the routine hauling-out of very much heavier vessels than this at all stages of the tide, although it could certainly be done. Much the same considerations apply to the portages from East to West Loch Tarbert (Kintyre), and elsewhere. If it could be achieved, light displacement would be a considerable advantage in the handling of a large galley ashore, although less important in the smaller sizes.
The requirement for speed.
The speed of a galley under sail would be less affected by her displacement than speed under oars, because the greater the displacement, the greater the power to carry sail. The effect of heavy displacement would be particularly slight sailing to windward (in smooth water, at least), a deeper-bodied vessel having a better grip of the water. How the displacement relates to the lightness of construction depends on the amount of internal ballast carried, which is uncertain and may have changed according to requirements. One thing that has emerged from Scandanavian experience with replicas is that, if built as lightly as the originals, they must be quite heavily ballasted to go to windward. This raises the question why they were built so lightly in the first place! But it should not be assumed that ballasting was a permanent thing as in a modern sailing vessel. It would not take very long to throw ballast stones over the side when speed under oars, or off the wind, was needed in an emergency. And in most parts of the West coast shingle beaches where ballast can be loaded are not difficult to find. We certainly cannot exclude the possibility that the Scottish galleys – especially the larger ones – were more crudely constructed than earlier Viking boats but carried less ballast, giving them similar displacement when trimmed for windward work under sail. There would be little difference in windward performance or seaworthiness, although a light boat with the ballast stowed near amidships will pitch less, go faster and take less water aboard in rough conditions. Light displacement becomes an advantage off the wind under sail and is much more important under oars. Speed and weatherliness under sail clearly mattered, as they must have been improved by the straight keel and hollow garboards even though these entailed some cost in rowing performance.
However all that we know about the ways in which the galleys were used suggests that speed under oars was crucial. There is some documentary evidence that battles were fought under oars only as in the Viking era (Orkneyinga Saga; MacInnes 1974), and a number of stories of desperate chases, like the tale of the Weaver (J. MacPherson. Told by the Coddy. Johnston & Bacon, 1960), together with place-names like Rudh’na Bhristeadh Ramh (promontary of the broken oars, where a hard-pressed small boat lured her larger pursuer into a narrow gap). In the later period, when galleys were matched against conventional naval ships and armed merchant vessels, they would have needed their speed even more. Their only chance of success would be surprise, appearing at high speed out of a dark night and pulling alongside before any cannon could be brought to bear upon them.
The effect of displacement on rowing performance depends on the speed range being considered, or more precisely on V/Ö L. Roughly speaking, expressing the speed V in knots and length L in feet, around V/Ö L = 1.0 and below hull resistance depends on wetted area, and at speeds much above this, on wavemaking. The Gokstad faering replica has a maximum speed of over 7 knt (V/Ö L = 1.75), and a boat as light as that could be rowed for some distance at speeds of V/Ö L = 1.0-1.2. In the larger sizes, however, there is much less power available relative to the displacement, most of the weight being oak rather than oarsmen. Thus a 24-oar galley would, if on Viking proportions, have just under 12 times the power of a 4-oared boat (assuming the latter was rowed as a double scull and the galley’s oars were not double-banked)- and 28 times the displacement. Relatively speaking, therefore, speeds would be lower in the larger sizes. The 24-oar galley might be capable of 1.2-1.3 /L (about 9 knt) on a sprint or somewhat under 1,0 /L on a longer row.
When the length of the boat is fixed, wavemaking resistance is directly proportional to the displacement, so that increasing the displacement by a factor of 4 – or even 2 – would have a very serious effect on this component of the resistance. This would considerably reduce the speed of most galleys near their maximum, and that of small boats most of the time. With the length fixed, the wetted area increases with the displacement but the exact relationship depends on the hull form. In boats of this type, roughly speaking the wetted area scales with ÖD, so that heavy construction is less of a disadvantage at low speeds. The conclusion is that if light construction could be achieved, the resulting benefit in rowing performance would make a galley considerably better suited to her assumed purpose, particularly in the small to medium sizes.
To summarise, we can say that the evidence is far from conclusive but tends to favour relatively light displacement for the galleys or at least the smaller boats related to them, for which speed under oars was more important relative to speed under sail. This suggests that they are foreshortened in the carvings. It seems likely that the skill to achieve light construction existed, and that light displacement would let them do their job better. However to answer this question finally the best solution would be to find and excavate a wreck.
HULL FORM AND RIG
Where possible the ideas in this section have been based on photographs or on examination of the carvings themselves. While it is not generally possible to distinguish galleys of different dates within the late Medieval period, there is enough information to establish some of the changes in hull shape that led into the beginning of this period and occurred at its end. The rig, on the other hand, remained essentially the same from the Viking era almost to modern times.
Predecessors to the Galleys, 13th Century or earlier.
There is a small but interesting group of representations of what appear to be Viking craft from the period of the Norse-Gaelic Kingdom of the Isles. These are;
1. The seal of Angus Mor of Islay, dated to 1292.
2. The galley on the Maughold cross-shaft on the Isle of Man.
3. A boat on an ‘Early Christian’ cross-shaft now Iona, which is thought on petrological grounds to have originated on Man.
All these lack a stern rudder, and may be assumed to have the rudder on the quarter in normal Norse fashion. They have a characteristic broken sheer, lowered in the way of the oars, which again was a common Norse feature at this period and earlier. The stem and sternpost are rather high and vertical in each case, although in (1) and (3) the overhangs are quite long with a pronounced knuckle just above the waterline forward and aft. With this last exception their general appearance is entirely Norse.
In addition there are some West Highland boats which may be tentatively assigned to the same early period from the lack of a stern rudder. These are;
4. The boats shown in the Kilchattan graffiti.
5. The 6- or 8-oared boat of which the Eigg stems were intended to form a part. These again show strong Norse affinities.
Because of the way in which the main group of galleys are carved within rectangular frames on the grave-slabs, it is often difficult to discern whether a stern rudder was intended or not. However clear stern rudders are normally accompanied by a square after ending to the deadwood in those examples that have been examined directly or from photographs. The profiles of nos 1-5 are smoothly rounded at this point. This feature is easier to see on the grave-slabs than the stern rudder itself. If a square-ended deadwood is assumed to go with a stern rudder in general (which is natural enough), then it is clearly missing in only one of main group of carvings;
6. A single carving on Iona dated by inscription to the mid-fourteenth century.
Thus the transition to stern rudders, a widespread development in Europe at the end of the 13th Century but perhaps a little later here, may be taken as marking the beginning of the main period.
The Main period.
It will be assumed that the galleys were light, fast and constructed to a high standard in the Norse fashion. The evidence relating to this has been discussed in the previous section. The consequences if they were deeper and heavier will be mentioned later. Most of the galleys fall into what will be called the Iona type. There are other variants that seem to occur consistently enough to be considered as additional types, but whether they differed in date or function, for example, is not fully clear. Note that this tentative classification of galleys into ‘types’ is independent of Steer & Bannerman’s classification of the carvings into ’schools’.
The Iona Type
The Norse ancestry is obvious immediately, whatever the proportions of the hull. The main difference is the stern-hung rudder. To get it as deep as possible the keel is quite straight, with no rocker, unlike the earlier boats and the normal Scandanavian design. As mentioned above the sternpost meets the keel at a sharp angle, and is either lightly curved or has a slight angle around the rudder head. The stem and sternpost are perhaps extended somewhat in the carvings to fill the space available, but the sternpost was clearly high enough to cause some problems in the arrangement of the tiller; the rudder head is only about half-way up it. There is no indication in any of the carvings of how this problem was solved. The rake of stem and sternpost is fairly similar and less than in most Viking or modern Shetland boats, although greater than in the North Norwegian fishing boats. The toe where the stem meets the keel is a smooth curve.
These features can all be found individually in other Northern countries when stern rudders were added to an essentially Viking hull form, but the exact combination of features characteristic of the Scottish galleys is not found elsewhere. In particular, more recent Norwegian, Faeroese and Shetland boats have a rounded heel to the sternpost and the rudder profile is curved to follow it. Danish, Swedish and a few South Norwegian boats have sharp angles in profile where both stem and sternpost meet the keel. The features of the Scottish galleys add up to a profile with more area of deadwood aft, and a deeper forefoot than in the Viking ships. Any gain in manoeverability from the stern-hung rudder would be nullified, and indeed the galleys must have been quite difficult to turn under oars even with the help of the rudder. The gain would be in sailing performance to windward and close reaching. In a shallow draught boat the forefoot, and to a lesser extent the after deadwood if there is much drag to the keel, present most of the resistance to leeway. The straight keel would also be a help when a galley was being pulled up, especially if greased ways or tracks for rollers were being used. A boat with much rocker on the keel tends to wriggle around and is difficult to control when being hauled out.
The sheer is flat in the way of the oar-ports, to keep the bow and stroke oars close to the water, but rises sharply in the ends. Provided any shifting ballast or supernumary crew were carried amidships the galleys would be quite dry in a head sea, and being so nearly symmetrical fore-and-aft would be controllable and well-behaved in a following sea as well. There are limits to the seakeeping abilities of any undecked boat in very heavy weather, but the galleys look as if they would go right to these limits if strongly crewed and well handled. (There is no certain evidence that they were indeed undecked, but it seems rather likely).
If the displacement were heavier and the construction rougher than envisaged above, there would be some loss of speed under oars, particularly over short sprints. Hauling out would be more difficult, particularly with a large galley on a steep beach. There would probably be a greater tendency to take water aboard, particularly to windward in a sea, even if the freeboard were increased all round with the extra windage that would entail. In other words there were good reasons for choosing light, flexible construction if it could be achieved, but the argument for this is strongest in the small sizes where the very exacting demands of this kind of construction, particularly in quality of timber, would be easiest to satisfy.
The Kilmory ‘Type’
A number of otherwise convential galleys have the sheer slightly hogged or convex amidships, although the bow and stern are still as high as usual. The oar-ports follow the sheer closely when shown. The stones at Kilmory Knap are good examples of this variant. The hogged sheer was not recognised by Drummond or White, who substituted a normal sheer, fairly flat amidships; hence its absence from the Inventory drawings. It is not clear how it should be interpreted. If the boat were heeled away from the observer the sheer would look hogged in this way when in fact it was not. However one stone on Oronsay shows a galley on the port tack, seen from the starboard side – and the sheer is still hogged. So perhaps it was a genuine characteristic of some of the galleys. It must be said, howeever, that no other difference from the Iona type could be identified apart from the sheerline itself, either visually or by analysis of the dimensions.
One possible function for the hogged sheerline would be to enable the vessel to be rowed from the lee side on a close reach or when working to windward under sail. Obviously at any angle of heel the oar handles on this side would be lowered, but the problem becomes much more severe as the oar-ports approach the water. Rowing from the lee side would obviously provide some extra power, would make tacking much more certain, and might reduce leeway if a good proportion of the arc of the stroke were forward of the oar-ports. It would only be possible if the galley carried a certain amount of lee helm when pinned in tightly on the wind. This question is discussed further in the section on the rig.
The Rodel Type.
There is one complex of features that are often found together and almost certainly represents a distinct type. Examples of carvings showing these features can be found, alongside other carvings attributable to the conventional Iona type, in the work of three of the main schools of carving (Iona, Knapdale and Oronsay) tentatively identified by Steer and Bannerman. So we seem to be looking at a real difference in the nature of the galleys represented, rather than a different style in their representation on stone. Their distinctive features are most common, however, in carvings attributed to the Oronsay carving school – and of course the Rodel carving itself – which belong to the last period of monumental sculpture in the West Highlands, and a late date therefore seems reasonable. The Rodel type has a rather shorter and more vertical stern than usual; a rudder with a disinctly separate blade in profile; and a considerably longer, smoother curve to both the stem and the sheerline in profile forward, so that the forward overhang is much longer than the stern. A hogged sheer is never found in this type. The sail area is sometimes drawn noticeably large, whether or not this was true of the boats themselves. The mast is generally stepped around the centre of the overall hull length, but this is distinctly forward of the centre of the waterline. The purpose of these features is again uncertain, although they would probably improve the manoeverability under oars. The balance on the helm would also be affected, with lee helm being more likely.
None of the carvings except the Rodel one gives a clear idea of the midship section. The Rodel boat has a certain amount of hollow in the garboards like the Gokstad ship and the modern Shetlanders. The hollow garboards are carried into the ends around the waterline but the ends are particularly full and buoyant above this, and the sections above water in the ends are more rounded than those of other vessels in the Norse tradition. The verses “The Advice of Donald Gorm” can be interpreted in a way that echoes these features, which would combine reasonable speed and dryness to windward under sail with limited draught, at a small cost in rowing performance.
Small boats of the Loch Awe Type
A small group of good carvings showing boats apparently with 6-7 pairs of oarsmen are located around Loch Awe. These differ distinctly from all the other carvings. There is no sign of a stern rudder and the keel is clearly rockered as in the early galleys that lack stern rudders. There are no oar-ports so presumably the oars lie on the gunwale. It is tempting to refer these to an early period but there is no inscriptional evidence on which to do so. It seems more likely that they are contemporary with the other main groups of galleys but retained earlier features because of their smaller size and the consequent emphasis on oars rather than sail. The preferential retention of characteristics from an earlier age in small craft is a recognised phenomenon in other maritime traditions. There is a worn carving of a very small (2- or 4-oar) boat on the church tower at Rodel, more or less contemporary with the Rodel galley, but with apparently similar characteristics to the Loch Awe boats.
Possible Irish Galleys.
There are two odd-looking carvings on Islay that do not correspond to any of the types above but have some similarity to the graffito at Dunluce in Antrim. Could these be Irish visitors?
Rig.
The main features of the rig are remarkably constant. There is a single square sail, on a stout mast stepped amidships or, occasionally, distinctly further forward. The mast is supported by a variable number of shrouds and a forestay, which in one case seems to finish in a tackle or something of the sort near the hull, perhaps to allow the mast to lower aft. Sometimes a backstay – or perhaps it is the fall of the main halliard – leads down aft of the shrouds. The shrouds and forestay appear to be spliced over the hounds, and the mast is sometimes thickened there as if for a halliard sheave. Two braces are attached a little inboard of the ends of the yard, and come down well aft. It is not clear how the yard is slung, nor how the sail is bent onto it. It may sometimes be in two pieces fished together. The sheets are never shown, nor is any form of bowline, although the tall stem could take a bowline without difficulty. Where the sail is shown set, its curves are moderately realistic although often twisted to accommodate the yard lying fore-and aft in profile. The luff is never shown distorted as it would be by a single bowline.
In the great majority of galleys where the sail is not set, it is still aloft, furled below the yard. Going out to the ends of the yard to stow the sail would an extremely precarious job on a light, narrow birlinn in a seaway. Most probably the yard was dropped when the sail was to be stowed, and then hoisted again out of the way when rowing or at anchor, if the mast was not lowered then. The windage would be considerable, however. Alternatively the sail may have been handled from below with brails, like a spritsail barge. There is nothing in the carvings to contradict their use, except that the Rodel boat appears to have its sail cut with the cloths in a strange arrangement like bricks in a wall (the seams have been interpreted as reef points, but this is unconvincing). Straightforward vertical cloths would best suit the use of brails. With the exception of the Rodel boat, on the above improbable interpretation, and the Kilchattan graffiti, which are hard to interpret, there is nothing to indicate how the sail was reefed. There are indications that the sailcloth was locally woven wool, although flax was also available.
A number of the galleys are shown with a tall flagstaff, a little inboard of either the bow or the stern. When stepped forward it normally has a pronounced forward rake and is close to the position suggested for a bearing-out spar or fore girt (beitass) in Norse vessels and used in luggers a century ago (March, op. cit.). However it is never shown clearly attached to the luff of the sail or doing anything else apart from flying a large pennant. If it was not used as a fore girt, the forward rake must have been to clear something, probably when going about, but we cannot tell what without knowing how the sail was handled when tacking.
The mast position of some of the galleys – all of variant B, some of the main group and one of variant A – is far enough forward to suggest that these vessels might carry a certain amount of lee helm when going to windward, with the yard nearly fore-and aft and the luff, in theory the working part of the sail, well forward. Perhaps whether this happened in practice would depend on how tight the luff was kept, and hence whether a bowline or fore girt was used. With a slack luff the flow moves aft, the leech hooks and the centre of effort is brought considerably further aft, preventing any tendency to lee helm. The possibility of rowing from the lee side has already been mentioned. Even if the balance on the helm were neutral, rowing with some lee oars might induce a little weather helm and thus make the hull more efficient hydrodynamically. Scandanavian experience with reconstructions of Medieval ships can possibly answer this question. So far as I know, the main lesson learned is that balance on the helm in these vessels can be controlled in a particularly sensitive way by the fore-and aft distribution of crew weight.
CONSTRUCTION
The MacLeod record quoted by Grant (1975) suggests that the planking, as well as presumably the frames and centerline timbers, were of oak. This record is late, but there is not much other evidence within the Highlands. The detailed and valuable information on the construction of galleys on the Clyde, for the Crown and for lowland magnates, has been set out by Gunn (MA Thesis, Glasgow University, 1986) The question is whether they were of the Highland type that they were presumably built to oppose, or related to the very different contemporary galleys built in England and the Continent. One piece of evidence, the sending of a messenger from Ayr to Arran for help, suggests that their affinities may have been with the Highland type. Arran was certainly Gaelic-speaking at that date, though within the Clyde Estuary. Both oak and pine were purchased for the construction of the lowland Scots galleys, but it is not clear which was used for their planking. While oak is a denser timber than the alternatives, it is probably more suitable for very light, thin planking since it is very resistant to splitting. Its superiority is reputed to be still greater when cleft rather than sawn. Elm was used more recently for the same reasons in the Cornish gigs and other light English craft, but the Wych elm of Western Scotland is inferior, not very common, and rarely available in long, clear lengths.
The Rodel boat has the winged stern and sternpost commonly used in Scandanavia at an earlier date and still to be found in a crude form in the Southern Baltic. The Eigg stern and sternpost are of this type. It is not clear whether or not it was general in the West Highlands. Even the best of the Argyll stones are uncertain in their detailed depiction of the planking. Close examination suggests that nos 015, 025 and 038 may have been intended to show the winged type of stem, and nos 053 and 055 the conventional type. The winged stem has considerable advantages if it can be built. It allows the nails in the hood ends to be riveted over rooves. In modern clinker construction (with a conventional stem) these are fastened with brass screws, and still sometimes pull away with age or when rainwater freezes in the bilges. Riveted fastenings would be stronger and longer-lasting than screws, and far superior to ordinary blind nails or trenails, the only contemporary alternative.
However the winged stem of a vessel the size of the Rodel boat would have to be worked from a massive piece of oak, even if the wings were not very thick when finished. It would require a level of skill in selecting and seasoning the timber that probably could not be found today, even if a suitable tree were located. One is reminded of the problems encountered in making up the stem and sternpost for the relatively tiny Greenwich faering replica, when shakes appeared all over the first attempt in oak. Perhaps bog seasoning was the key. At any rate we are clearly dealing with shipwrights of considerable virtuosity, who knew their materials well.
Whether or not winged stems and oak planking were universal, they imply a high standard of workmanship when they do occur. This and the corresponding skill in procuring materials are unlikely to have arisen out of a vacuum. It is more likely that over an extended time within our period, at at least some centres, there was a flourishing tradition of high-quality construction, and owners who did not stint expense. It does not follow, of course, that all the galleys were built to that standard. Even if only some were, it would help to explain why they still catch the imagination of a local population that has always appreciated boats.
Sources of timber.
1) Local Sources.
For the moment it is assumed that oak was used. West Highland pine is in any case not very suitable for planking. It has a very open habit of growth that produces reasonable crooks, like a hardwood, but little long, straight timber. Pine would certainly be available from the Baltic, but only rather short boards were being imported into Lowland Scotland where we have records, probably due to the difficulty of shipping long lengths. If yew were used as tradition suggests, it is hard to see what for. Yew grows on limestone, principally on Lismore and Bernara. It is desperately gnarled, contorted stuff.
The planking and keel would therefore need fairly long, clean oak. The Eigg boat was intended to have much narrower strakes than was normal in Norway, and this may reflect the availability of timber here. Without seeing a full midship section of a galley it is not clear how much curve would be in the strakes, and hence how wide would be the timber from which they were cut. If the topsides had plenty of flare, as seems possible, the curve would not be great, but the hogged sheer must have made for awkward planking. Unfortunately none of the carvings is detailed enough to show the butts in the strakes. No doubt the high, curving ends were worked separately, at least near the sheer. However the galleys were planked, much heavier oak would be required for the stem, stern, and the mast step if Viking fashion.
Oak of the necessary dimensions could not be obtained in the West Highlands today. Of course it may all have been felled for building ships. The difficulties in procuring suitable oak in the lowlands are clear enough. The longest and cleanest trees in the surviving Highland oakwoods are mostly in steep, sheltered glens and ravines where they have a long way to grow to reach the light. There is a small amount of surprisingly tall oak in sites like this in Knapdale, and the dimensions of the timber become steadily greater eastwards within Argyll. One of the two claimants for the title of Britain’s tallest tree (an American species, naturally) is in the deep glen of Cairndow by the head of Loch Fyne. Oaks of large girth might be harder to find. Again the best prospects are eastwards.
From Dean Munro and the Bleau atlas, which is surprisingly reliable in its depiction of woodland in Lowland Scotland, the 16th-century distribution of woodland seems to have been fairly similar to today’s in the West Highlands. Obviously there was more of it, but not a lot. The main areas that have disappeared were along the E coasts of Jura and Mull, which would be oak, probably not growing to great size; and in parts of Wester Ross and Sutherland, which would be mainly birch and pine except in the S (D.N. MacVean. Ch.5 in J.H. Burnett, The Vegetation of Scotland. Oliver & Boyd 1964; M.E. Ball, Proc. Royal Soc. Edinb. 83B: 319-339, 1983). Good oak remains in a few places in Kintail. The building of a galley is recorded in the Kishorn area (MacInnes, 1974), where there would certainly be some oak; but pine would be available too a little inland, and the best surviving ashwood in the Highlands is close at hand. However it is likely that Knapdale, Lorne and the country around the upper half of L. Fyne had the most extensive and most valuable oakwoods then as now. That is Campbell land, of course, and long has been.
Import of timber from Ireland.
This has been plausibly suggested by MacInnes (1974). Certainly oak was coming from Wexford to the Clyde at a slightly later period. We do not know how it would be brought, if the dimensions were large. It would not be an easy tow in these waters. Perhaps it was selected, felled and roughed out by either Highland shipwrights or local agents before transport. But if so, why did the Irish themselves build no galleys at that time?
Imports from Norway.
The construction of warships in Norway, on the scale of the larger West Highland Galleys, had ceased by the 14th Century. However before the end of our period, there are records of small boats being shipped in knocked-down form from Norway to Shetland (A.G.Osler. The Shetland Boat. I. South Mainland and Fair Isle. NMM Monograph no 58. Greenwich (1983). There is not much sign of direct trade with the Hebrides at that time, but of course it could have taken place via Shetland. At a later period – the late 18th Century – we have records of 4- and 6-oared boats from the Trondheim and Christiansand areas to N and NE Ireland (M.McCaughan. Irish vernacular boats and their European connections. Ulster Folklife 24: 97-108 (1978). A mention at the same period by Knox (J.Knox. The Highlands and Hebrides in 1786. Reprinted by James Thin, Edinburgh, (1975), p.92) makes it clear that similar boats were also arriving then in the Hebrides. So there is no doubt about the Norwegian connection, but no evidence that fighting vessels the size of galleys came from Norway after the Viking period.
Oars and Spars.
Here we have little to go on apart from modern experience with locally available materials. For spars, the best combination of lightness and strength would be obtained with Norway spruce, which may have been worth importing. Pine is adequate but heavier. For oars, spruce is again very suitable, but a rather thin, whippy oar can also be built from ash with little increase in weight. The stiffness of oars is a matter of rowing style and personal preference. Ash oars would be less readily demolished than spruce under battle conditions.
Some references relevant to galleys (up to 1987 only)
R. Simper. Scottish Sail, a Forgotten Era. David & Charles, Newton Abbot (1974)
B. & O. Faeroyvik. Inshore Craft of Norway. Conway 1979.
A.G. Osler. The Shetland Boat. South Mainland and Fair Isle. NMM Maritime Monographs & Reports no. 58 (1983).
W. Rudolph. Inshore Fishing Craft of the Southern Baltic. NMM Maritime Monographs & Reports no. 14 (1974).
O.H. Wicksteed. The Hardanger Faering. NMM Maritime Monographs & Reports no. 32 (1978).
O.T.P. Roberts. The Cots of Rosslare Harbour and Wexford. Mariner’s Mirror 71: 13-34 (1985).
M. McCaughan. Irish vernacular boats and their European connections. Ulster Folklife 24: 1-22 (1978).
E.J. March. Sailing Drifters. David & Charles (1969).
E.J. March. Inshore Craft of Britain in the Days of Sail and Oar. Vol. 2. David & Charles (1970).
S. McGrail (ed.) Aspects of Maritime Archaeology and Ethnography. NMM 1984.
J.T. Tinniswood. English Galleys 1272-1377. Mariner’s Mirror 35: 276 (1949).
R.W. Unger. The Ship in the Medieval Economy 600-1600. Croom Helm, London (1980)
R. Williams. The Lords of the Isles. Chatto & Windus (1984).
G.V. Scammell. The World Encompassed. The First European Maritime Empires c.800-1650. Menthuen (1981).
H. Marwick. Leiding in the West. Proceedings of the Orkney Antiquarian Society 13: 1934-5.
H. Marwick. Naval defence in Norse Scotland. Scottish Historical Review 28: 1-11 (1949).
N.E.H. Nicolaisen. Norse settlement in the Northern and Western Isles. Scottish Historical Review 48: 6-17 (1969).
A. McKerral. What was a davach? Proceedings of the Society of Antiquaries of Scotland 82: 49-52 (1947-8)
A. McKerral. The lesser land and administrative divisions in Celtic Scotland. Proceedings of the Society of Antiquaries of Scotland 85: 52-64 (1950-1).
J.P.W. Philipsen. The Utrecht ship. Mariner’s Mirror 51: 35-46 (1965).
J.R. Baldwin. (ed.) Scandanavian Shetland, an Ongoing Tradition. Edinburgh (1978).
A. Campbell. Argyllshire Galleys, some Typical Examples from Tombs, Slabs and Crosses. Privately printed, 1906.
A. Fenton & H. Pallson (eds). The Northern and Western Isles in the Viking World. Edinburgh (1984).
S. McGrail (ed.) The Archeology of Medieval Ships and Harbours in Northern Europe. NMM Archaeological Series no. 5 / BAR Int. Ser no 66 (1979).
O. Hasslof. Sources of maritime history and methods of research. Mariner’s Mirror 52: 127-144 (1966).
P. Johnstone. Bronze Age Sea Trial. Antiquity 46: 269-274 (1972).
C.E. Dove. The first British navy. Antiquity 45: 15-20 (1971).
A.C. Evans & V. Fenwick. The Graveney boat. Antiquity 45: 89-96 (1971).
O. Crumlin-Pedersen. The Gredstedbro ship. The remains of a late Iron Age vessel found in 1945 in South Jutland. Acta Archaeologia 39: 262-267 (1968).
O. Olsen & O. Crumlin-Pedersen. The Skuldelev ships. II. A report of the final underwater excavation in 1959 and the salvaging operation in 1962. Acta Archaeologia 38: 73-174 (1967).
H.H. Brindley. Graffiti in Upper Deal Church. Mariner’s Mirror 20: 235-237 (1934).
S. McGrail (ed.) Sources and Techniques in Boat Archaeology. NMM Archaeology Series no 1. / BAR Supp. Ser. 29: (1977).
C.O. Cederlund (ed.) Postmedieval Boat and Ship Archaeology. BAR International Series no. 256 (1985).
E.V. Wright. The North Ferriby Boats. NMM Maritime Monographs & Reports no 23 (1976).
O. Crumlin-Pedersen & M. Vinner (eds) Sailing into the Past. Proceedings of the International Seminar on Replicas of Ancient and Medieval Vessels, Roskilde. Viking Ship Museum, (1986).
O. Crumlin-Pedersen. From Viking Ships to Hanseatic Cogs. NMM (1983).
V. Fenwick. The Graveney Boat. BAR British Series no. 53 (1978).
S. McGrail. The Gokstad Faering. NMM Maritime Monographs & Reports no 11 (1974).
J. MacInnes. West Highland sea power in the Middle Ages. Transactions of the Gaelic Society of Inverness 518-555 (1974).
A.R. Cross. Tombstones in the Isle of Iona. Mariner’s Mirror 36: 270-1 (1950).
K. Steer & J. Bannerman. Late Medieval Monumental Sculpture in the West Highlands. HMSO, Edinburgh, pp. 180-4 (1977)
Royal Commission on the Ancient and Historical Monuments of Scotland. Argyll: and Inventory of the Ancient Monuments. HMSO. Volume 1. Kintyre (1971). Volume 2. Lorn (1975). Volume 3. Mull, Tiree, Coll and North Argyll. (1980). Volume 4. Iona (1982). Volume 5. Islay, Jura, Colonsay and Oronsay. (1984).
E. Bray. The Discovery of the Hebrides. Voyagers to the Western Isles 1745-1883. Collins, London (1986).
S. Vadstrup. Imme Saerimmer. Gokstad-faeringen som Seniorprojekt. Danske Spejderkorps, Frederiksvaerk (1978).
A.B. Taylor. Alexander Lindsay: a Rutter of the Scottish Seas circa 1540. NMM Maritime Monographs & Reports no. (1980).
W.C. MacKenzie. The Highlands and Isles of Scotland: a historical survey. Moray, Edinburgh 1949
W.C. MacKenzie. History of the Outer Hebrides. Paisley (1903).
G.A. Hayes-McCoy. Scots Mercenary Forces in Ireland (1565-1603). Dublin (1937).
J. Dawson. The fifth Earl of Argyle, Gaelic lordship and political power in sixteenth-century Scotland. Scottish Historical Review 67: 1-27 (1988).
W. Watson. Bardachd Ghaidhlig. pp. 259-262. Inverness (1976).
S. McGrail. Ancient Boats in Northwest Europe: the archaeology of water transport to AD 15. London (1987).
S. McGrail. Sewn Plank Boats. Oxford (1985).
B. Greenhill (ed.) Aspects of the History of Wooden Shipbuilding. NMM Maritime Monographs & Reports no 1 (1970).
O.L. Filgueiras (ed.). Local Boats. BAR International Series no. 438 (1988).
S. McGrail (ed.). Woodworking Techniques before AD 1500. NMM Archaeological Series no 23 / BAR International Series no. 129 (1982).
J. Coates & J. Morrison. Authenticity in the replica Athenian trieres. Antiquity 61: 87-90 (1987).
J. Coates & J. Morrison. The Athenian Trireme. Cambridge University Press (1986).
Please also click on the links below.
Evidence contributed to Colin Mudie for the design of Aileach