Digitized by the Internet Archive in 2015 https://archive.org/details/shortessayonageuOOferg A SHORT ESSAY ON THE AGE AND USES OF THE BROCHS AND THE EUDE STONE MONUMENTS OP THE OEKNEY ISLANDS AND THE NORTH OF SCOTLAND. BY JAMES JERGUSSON, D.C.L, F.R.S., F.R.I.B.A., M.R.A.S., F.G.S., F.R.G.S., HON. MEMBER R.S.I. ETC., ETC. WILLIAM MULLAN AND SON, 34, PATEENOSTEE EOW, LONDON; 4, DONEGAL PLACE, BELFAST. 1877. DA m .05 BOSTM COLLME IfinVt CHESTHin M(U (A* 0216? APR 0 7 "90 PREFATOUY NOTE. The following short essay was originally written with the intention that it should appear in the pages of the CONTEMPORARY Review. When, however, it was set up in type, it was found to be longer than was compatible with the hmited space that can be allotted to the treatment of such a subject in a monthly periodical. The alternative was therefore presented to me, either of cutting it down so as to bring it within the required limit, or to let it appear as a separate publication. After several attempts, the first desideratum was found to be unattainable; the fact being that, though too long for the Review, the article is too short for the subject of which it treats. An argument such as it attempts ought not, in fact, to be broached at all, unless it can be treated with such fulness as to show that no essential points have been overlooked or evaded. Without this, every one familiar with the subject sees at a glance the breaches that have been left in the walls of the fortress, which is then stormed without difficulty, and an easy victory obtained for the adherents to the established faith. A third alternative was of course open. It was to distribute the type and forget that it had ever been set up. Tliis I was unwiUing to do ; for, though too short for the general reader, the argument is stated with sufficient fulness to enable the initiated to appreciate its value. If they think it worth while to do so, it can easily be written out at greater length with all the requisite illustrations and pieces jnstificatives, and in that state republished as a second edition. B iv PBEFA TOR Y NOTE. Meanwliilo, it may servo appropriately as an appendix to my work on Rude Stone Monuments." The word " Brochs " does not occur there, except incidentally in a note. When writing it, I looked on them as castles properly belonging to the History of Llilitary Architecture which I at one time intended should form a volume in the series of my Architectural histories. As there is now no chance of that intention being ever carried into effect, I am glad to have tliis opportunity of expressing my opinion of the age and uses of these towers, which I beheve to be certainly of Norwegian origin. JAMES FERGUSSON. December, 18 7G. A SHOET ESSAY ETC. ETC. THE publication of a translation of the Orkneyinga Saga," with an elaborate introduction by Mr. Joseph Anderson,* has recently had the effect of directing attention to the important group of antiquities that exist in the Orlmey Islands. From his position as curator of the museum of the Scottish Society of Antiquaries, and having been employed personally to superintend some important investigations in the north, Mr. Anderson has perhaps had better opportunities than almost any other living antiquary for making himself acquainted v^dth the facts of the case ; while his elaborate paper on the " Brochs," in the fifth volume of the " Archeeologia Scotica," is by far the most complete and exhaustive treatise that has appeared on that branch of the subject. His statements of facts may, therefore, be accepted without hesitation ; but whether the inferences he draws from them may be as implicitly relied upon remains to be seen. Still more recently the Orcadian group of antiquities have been the subject of an article in the Quarterly Rei'iew of July last ; and as that periodical is read by thousands, compared with the very few who have access to the Saga," or care to look into the Archaeologia," it would have been of the utmost importance that it should be written by some one capable of guiding rightly public opinion on this interesting subject. Unfortunately, however, it is only too evident that, in this case, the author, however competent he may be to speak with authority as to the " Saga," knoAvs but little of the science of archaeology in general, and still less of the authorities bearing on the special subject he proposes to discuss. What little he knows of Orcadian antiquities seems principally derived from the perusal of the work he takes on himself to * Edmonston and Douglas : Edinburgh, 1873. B 2 ORCADIAN ANTIQUITIES. criticize and condemn.* Having no independent means of judg- ing for himself, he is naturally aghast at the daring of one who has been guilty of the unpardonable offence of thinking and investigating for himself, and who, consequently, has arrived at conclusions, in some cases, not sanctioned by the orthodox world. Such essays as his — like works on controversial divinity — are no doubt extremely comforting to those who are of the same school, and were beforehand of the same opinion as the writer ; but seem singularly uninteresting and devoid of logical sequence to those who belong to another faith, or who have sufficient know- ledge of the subject to perceive how little their author is capable of suspecting there may be another side to the question. If the knowledge of the writer of the article had been only a little more profound, his criticism would have been very accept- able to the author he was reviewing. All he has to urge against his vieAvs is, they are novel, unlikely, improbable, impossible, and diff'er from those generally accepted.f If nothing more pertinent can be urged, we may wait with patience, convinced that the time is not far off" when their justness and correctness will be admitted. The truth of the matter is, a controversy of this sort is not to be settled by adjectives; any one can use them, but in no instance can they be accepted seriously as a substitute for facts. No one, in fact, ought to be allowed to sit in judgment and pronounce an opinion on such questions as these, who has not mastered the whole of the evidence bearing on the subject. This, however, is no easy matter. It is extremely voluminous, and the bulk of it scattered in fragments through some hundreds of volumes, of Transactions and Periodicals, written in some half-dozen of lan- guages, and many of them extremely difficult of access. Besides this, a very large portion of it is written by persons who are untrained for such investigations, and consequently incapable of recording correctly what they have imperfectly observed. From these, and fifty other causes, which any one who has ever attempted the investigation will easily call to mind, it requires not only long and patient labour to master the details of the sub- ject, but more than this, calm and sustained judicial attention to enable any one to arrive at trustwortliy conclusions regarding them. * Rude Stone Momimonts in all Countries, their Age and Uses. By Jas. Fergusson. Murray, 1872. f As the contention of the reviewer is that I stand utterly alone, and am -wrong in every essential particular but one, it may clear the discussion if that one is pointed out at once, as it will not then be necessary to refer to the article again, it being understood tliat all the rest is mere fault-finding of the ordinary type. In his penultimate page (L")!)), he compliments me on being the first to show clearly the resemblance that exists between the African and European Dolmens, and the Buddhist Dagobas, espe- cially those of Ceylon. The compliment is all the more gratifying because unexpected. Till pointed out to me by the reviewer, I was not aware of any similarity between these two classes of monuments, beyond the fact that both their names commence with a D, while to me it appears, that a man who does not know the difference between a Dagoba and a Dolmen, does not know the A B C of his subject. ORCADIAN ANTIQUITIES, 7 In the huny-sknny of modem life, witli its tliousand-aii cl- one distractions, it is perhaps too much to expect that many will devote to it the time and attention requisite for such an investigation. Others, too, may think that the result to be attained is not worthy of the labour required to achieve it ; and for these and other reasons we can hardly expect it will often be attempted. Under these circumstances, it is fortunate that in the extreme north of these islands there exists a group of Rude Stone Monu- ments, so small and so easily accessible tliat any one can master their details with very little labour, and which have been so fre- quently described and so fully discussed that all that is requisite for forming an opinion regarding them is easily accessible. They form, at the same time, a group so complete in itself that their history may be accepted as, what lawyers would describe as, a representative case, on which issue may be joined and judgment accepted without much fear of the decision being upset on an appeal to a more extended argument, on the whole evidence bearing on all the points involved. Besides this, the discussion can hardly fail to be interesting as disclosing the nature of the authorities on which such subjects must be argued, together with a perception of the degree of certainty that can be arrived at from such evidence as is attainable. If it can be shown, with any reasonable degree of certainty, that all the Rude Stone Monu- ments of the Orcadian group belong to the Norwegian period, all a priori improbability is removed from the theory that Stonehenge, Avebury, and the greater number of the larger English tumuH are of post-Roman origin. If, on the other hand, the evidence is not sufficient to establish the comparatively modern origin of the Orcadian monuments, it will be difficult to find any other group upon whose age or use we, at present at least, can bring the same amount of evidence to bear. In order to discuss the history of the Orcadian rude stone monuments with as much clearness as is compatible mth the subject, it will be convenient to divide them into three sub- ordinate groups, and to examine each separately. First, the Brochs, or round towers, which are by far the most numerous, as well as the most characteristic examples ; Secondly, the tumuh, whether chambered or otherwise, and generally the sepulchral monuments of the islanders ; and Lastly, the stone circles, which, though the rarest, are by no means the least interesting monuments of the gimip. Though so numerous, the Brochs are all marvellously like to one another. The remains of at least 400 are Imown, and have been examined, and there may have been 1,000 at one time, or V. 8 ORCADIAN ANTIQUITIES, more.* Still tliey are so similar that no one has yet been able to point out a succession among them, or any characteristic by which they might be divided into groups. Indeed, I once heard an eminent Scotch antiquary, very familiar personally with their appearance, gravely maintaining that they were all erected at one time, and from one plan and specification. Though this is of course absurd enough, there is not, so far as I know, any example in any part of the world of so numerous a class of buildings which show so little difference in design or dimension. They are all circular, about 60 feet in diameter externally, and contain internally a court-yard 30 feet — more or less — in diameter. The walls of this inner court are practically perpendicular, while the external walls slope inwards at a considerable angle, so as to give the towers the form of truncated cones. The walls of the towers are consequently about 15 feet in thickness at the base, and in them, on the ground floor, are generally found two or three large apartments following the curve of the walls, which were apparently the living rooms of the inhabitants. Above these — still in the thickness of the walls — are generally to be found a series of low apartments, divided, by horizontal slabs, into berths like those in our passenger steamboats, and used apparently for the same purpose ; but higher up, where the walls get thinner, they could only be used as store-places or cupboards for the custody of provisions or valuables. All these apartments were hghted from the inteiior by openings looking into the court- yard, which — at least originally — never appears to have been roofed. In addition to these there is always a staircase — also in the thickness of the walls — leading from the basement to the top of the building, and giving access to these various apartments. In none of the Brochs is there any opening externally, except the doorway. That is always on the level of the ground, low and narrow, and leading by a passage of about the same section as the doorway, but 15 feet long, to the inteiior court of the build- ing. There were always apparently two doors in this passage, and between the outer and inner either one or two guard-chambers, wliich formed very efficient defence against any one trying to penetrate by this entrance to the interior. The height of these towers is the one element which is un- certain. None unfoiiunately are quite complete. That at Mousa, * I am sorry to observe my friend Mr. Anderson prejudging the question by apply- ing the name " Pictish towers " to these erections, as if the matter were quite settled, ■which however is, to say the least of it, still siihjndice. Others call them borgs, burgs, or burghs, which may or may not be correct designations ; but as these terms are fre- quently applied, at least in combination with other names, to other fortified places, which are not Brochs, their employment may be misleading. Consequently, throughout this paper the term Broch, and that only, is employed. 'SMiatever its provenance, it ■was never applied to any other building than these Orcadian round toAversj and if confined to them only, its use can never be misleading. ORCADIAN ANTIQUITIES. 9 wliicL. is the most perfect of those remaining, is upwards of 40 feet in height,* and with the parapet probably reached 45 feet. One in Glenelg was 30 feet 6 inches high Avhen Pennant visited it in 1776, but he states that 7^ feet of height had just been removed.t The Dun of DornadiUa is now 30 feet, but must once have been much higher.J Taking it altogether, we shall not perhaps be far wrong in assuming that the three principal dimensions of these buildings were, or were intended to be, 60, 45, and 30 feet. There are variations of course, and such as should be pointed out in a special treatise on the subject ; but all that is Avanted here is to state their general characteristics so that they can be recog- nized, and by which they can be distinguished from buildings of a similar nature. This is easily done, for they are all so much alike that, though the theory of my friend just alluded to is untenable, we may certainly assume, without fear of error, that they were all erected by one people, for one purpose, and Avithin a very limited space of time, say two, or at most three centuiies from the earliest to the last. The one question therefore that remains is, who were this people ? Fortunately, in this case, there are only two claimants. Either they were erected by the Picts or Celtic races who inhabited these islands from the earliest times to which history and tra- dition ascend ; or they were the work of the Norwegians Avho settled on the islands in or before the eighth century after Christ, and finally conquered and extirpated the Celtic inhabitants till, as the good Bishop Tulloch said in 1449, " that of their posteritie there remained nocht."§ So far as we can at present judge, the extirpation must have been indeed complete, for in no part of Scotland is there less trace of the blood or language of the Celts than in these islands. Whatever may have been the casein the days of Monkbarns, the modem school of antiquarians has settled ethnographically, with sufficient precision, that the Picts are a branch of that great race of Celts that occupied Scotland and Ireland, and all at least of the western part of England, at the time when the Romans held sway in these isles. The geograpliical distribution of the Picts has also been made unexpectedly clear by a map published by Dr. Stuart in the first volume of his beautiful work on the "Sculptured Stones of Scotland." || The evidence he produces in his two introduc- tions to these volumes satisfies him, and, I presume, must convince every one else, that the sculptured stones are the works of the Picts ; but so little is the value of architectural remains, as an ethno- graphical index, known or appreciated in this country, that it * ArchcBologia Scotica, vol, v. pi. 22, t Pennant's Tour in Scotland, vol. ii. p. 190. % Loc. cit. p. 194. § Arclioeologia, vol. xxxiv. p. 89» I| Vol. i. p. xxxi. 10 ORCADIAN ANTIQUITIES. does not seem to have occurred to him that, while constructing a map of the locahties where these sculptured stones are foimd, he was, in fact, compiling, from the only really authentic source of information, a map of the distribution and relative density of the Pictish population in Scotland. So it is, however. Their piincipal seats are shown to have been between the Tay and the Don, but they extend southward to the Forth, and northward, though sparsely and sporadically, to the Shetland Islands. There they existed, however, so far as we can gather from such glimpses of their liistory as remain, only as a wretched scattered race of fishermen. They may also have been shepherds, but hardly ever lising to the dignity of an agiicultural people, and certainly never indulging in maiitime pursuits. Like all the various branches of the Celtic race, not only in these islands but on the Continent of Europe, the Picts eagerly embraced Christianity whenever and wherever it was announced to them, and seem never to have wasted a thought on any previous religion, if they had any, or to have expressed regret at the loss of any superstition or symbol they may once have cherished. The Saxons — using that name in its T^idest sense to include the Scan din avians — on the other hand long resisted any attempt at conversion. They absolutely refused in their own coimtry to adopt the new faith till it was thrust down their throats at the point of the sword by Cliarlemagiie in the nmth century, and in England they remamed unconverted till they found it impossible to govern by far the most numerous body of their subjects, without conforming to the faith they had adopted. But even then they clung to their old gods and their old fonns of faith ^uth a tenacity that proves how constrained their conversion really was, and the moment the Reformation afforded them a chance they threw aside at once one-half of what the Celtic races think and always thought were essential parts of Christianity, and asserted the independence of then' own private judgment. That, however, is not a question it is necessaiy to discuss here, but it is indispensable, in order to understand what follows, to point out that the Celt and the Saxon were then, as they are now, antagonistic races, ha^^ng different faiths, striving after diverse auns, and working by distinct and generally opposing methods. All this ^411 become clearer as we proceed. Meanwliile, how- ever, there are certain things that clearly belong to the Picts or Celts, and to which the Scandina^-ians lay no claim, which it may be convenient to point out before proceeding further. Through- out the islands, and occasionally on the mainland, there are numerous cells or diminutive chapels of the i-udest possible con- ORCADIAN ANTIQUITIES. 11 stmction.* These are so nearly identical with those found in Irc4and, especially on the west coast, that there can be but little (if any) doubt that they were erected under the guidance of the Irish missionaries for their Pictish disciples shortly after their con- version by St. Columba. It is not easy to determine how low they extend in date, but there is certainly nothing to show that they attained to the architectural level or modern date of such a church as that in Egilsey,t built probably in the twelfth century. Notwithstanding the temptation at first sight to assign an Irish origin to the round tower that adorns the west end of that church, a closer inspection makes it clear that, like the Norfolk churches with circular west towers, the whole of its affinities are purely Scandinavian.^ There are, besides these, a series of Picts' houses — using the term " house " in its English, not its antiquarian sense — which, if they cannot be shown to extend as far back as St. Columba's time, were certainly inhabited by the Celts up to a very recent period.§ The best known example of these is that at Skara, in Skail Bay, recently excavated by Mr. Watts,|| and though according to the Danish system it is certainly prehistoric, I quite agree with Mr. Petrie, in whose company I visited it in 1865, " that it is evident that the rude nature of the implements, apart from other facts and circumstances, cannot be accepted as a proof of their great anti- quity ;"^T in other words, that it probably was inhabited within the last few centuries. But whether this is so or not. Captain Thomas has conclusively shown that the Celtic population lived in under- ground bothan and beehive houses as rude as those of any North American savage down nearly to the present day, and that all the conditions which would produce the harbour mound at Keiss,** and all the evidence of the lowest possible state of civilization in which a community of men can exist, have prevailed in these parts till very recent times. The question now before us is. Did these Celts or Picts at any period between the departure of the Romans and their subjection to the Norwegians ever attain to such a stage of civilization as v/ould enable them to erect such a tower as that at Mousa, such a sepulchre as the Maeshow, or such a circle as that at Stennis? The writer of the article in the Quarterly Revieio probably * Tho best account of these will bo found in Muir's " Characleristics of Old Church Architecture in the Mainland and Western Islands of Scotland." Edinburgh, 18G1. Further information on tho subject will be found in Anderson's " Orkneyinga Saga," pp. xci. et seqq., and Archcoologia Scotica, vol. v. pp. 109 et seqq. See also same volume, pp. 237 et seqq. t Orkneyinga Saga, pp. xci. ot seq. X History of Architecture, vol. ii. pp. 181 et seq. § See Captain Thomas's Paper in Proceedings of Society of Scottish Antiquaries, vol. iii. pp. 134 et seqq. II Proceedings Scot. Ant. vol. vii. p. 201 et seqq. i Loc. cit. vol. vii. pp. 217. ** Loc. cit. vol. viii. pp. 192 et seq. B 3 12 ORCADIAN ANTIQUITIES, represented correctly the general feeling on this subject, when he stated (p. 147), that " it is admitted on all hands that the Brochs " — he calls them Borgs — " are p re-No vwegian." ]\[r. Anderson takes the same view, but contends with more precision for their being all erected by the Picts, not earlier than the fifth, and not later than the ninth century."* On the other hand I believe it can bo proved with as much certainty as such a question is capable of attaining, that they were all erected by the Norwegians ; the bulk of them in the ninth, tenth, and eleventh centuries. One or two may be earlier, and a few also later, but the mass, if I am not mistaken, belong to the true Yiking period. This, therefore, is a clear and distinct issue, on which judg- ment may be pronounced, without ambiguity, and I believe without hesitation. Brochs. The following table, compiled by Mr. Anderson, may be taken as fairly representing our present knowledge of the number and distribution of the Brochs : — Shetland 75 Broug'ht over ... 328 Orkneys 70 Inverness — Islands ... 41 Caithness 70 Forfarshire 2 Sutherland GO Perthshire 1 Koss — Mainland 10 Stirlingshire 1 Island of Lewis... 28 Berwickshire 1 Inverness — Mainland . . . 6 Total ... 374 Carried over 328 This total, he adds, is exclusive of a large part of the west coast of Ross-shire, of the whole of the mainland of Argyle, and the whole of the outer Hebrides south of Harris, and the islands south of Skye."t Consequently, though it may no doubt be true that some remains are enumerated as Brochs which have no title to that name, there are certainly more than 400 — it may be 500 or 600 at least — of these curious towers still existing in the islands and mainland of Scotland. Assuming the prehistoric or Celtic theory to be the correct one, the first thing that must strike any one on examining this list is that we have here 400 or 500 fortified residences of the older and inferior race, while no one has yet been able to point out even the site of a single residence, fortified or unfortified, of their Scandinavian conquerors, erected during the first three centuries of their occupation. This is, it is true, only a negative * Arclia3ologia Scotica, toI. v. p. 140. t Ibi.l. p. lOS. ORCADIAN ANTIQUITIES. 13 fact and cannot consequently be considered final, but it is so unlike what has generally been found to happen elsewhere, and is so curious and unexpected, that it ought at least to make the ad- vocates of their extreme antiquity pause before proclaiming it too loudly. The next inference, however, to be derived from the study of this table, is more direct and more important, and so far as I can judge nearly final. All the known Brochs, with the exception of the five at the end of the list, are situated in those parts of Scotland which are known to have been possessed by the Norwegians. Unfortunately we have no very detailed map of the possessions of these sea rovers ; that by Mr. Skene is the latest and certainly the best, as embodying all the known written information, and is compiled by one of the most competent scholars of the day.* It is, however, on too small a scale to be quite satisfactory for our purpose. On the other hand Mr. Anderson's paper is accompanied by a map of the distribution of Brochs, north of the ]\Ioray Firth. The southern sheet of this map has never been compiled, and for the western isles the numbers only, and not the sites, are given.f So far as they go, these two maps are absolutely identical, and when completed Mr. Anderson's map of the Brochs will be as perfect an illustration of the geographical and ethnological distribution of the Norwegians as Dr. Stuart's map of the sculptured stones is of that of the Pictish people. % The five exceptional Brochs cannot be considered as in any way invalidating this conclusion. If indeed there is any truth in the old maxim, Exceptio prohat regulam, they rather confirm it. The two that are found in Forfar — Pictland proper — are situated one on each side of Dundee, close to the Firth of Tay, and near the battle-field of Carnoustie, where Malcolm II. (1000, 1033) defeated the Danes, who may have been Norwegians ; but whether or not, these Brochs from their situation look much more like the points dajyjjui of a sea-faring people invading the country, than the castles of the inhabitants, whoever they were. * Celtic Scotland, by W. J. Skene. Edinburgh, 1876. Map, p. 39G. t Arcliseologia Scotica, vol. v. p. 198. j The curious part of the business is that the nature and importance of architectural or sculptural remains for illustrating questionsof political geography or ethnology is so little understood or appreciated in this countiy, that these two distinguished antiquaries -svero hardly aware of the service they were rendering when they compiled these maps. In his two introductions, Dr. Stuart proves, in a manner that will hardly bo disputed, that the sculptured stones were one and all the works of the Picts ; but it does not seem to have occurred to him as a natural consequence, that where a sculptured stone now is found, a Pict must previously have existed. While so unconscious is Mr. Anderson of the science of ai'chitectural ethnology, that he will prol)ably be very much astonished to bo told that he has compiled the best geographical and ethnographical map of Scandi- navian Scotland, from the best, and in many instances the only, available materials for tlie purpose ! For thirty years I have been trying to persuade my countrymen to take up this subject. Had they done so, many of the problems that puzzle and perplex antiquarians, would never have arisen, or would long ago have bien settled. 14 ORCADIAN ANTIQUITIES. The next pair are situated on the top of two hills about eight miles apart, astride on the Forth, above Stirling, and again their situation seems only intelligible on the same grounds as apply to the last pair. The fifth is so wholly exceptional as hardly to re- quire notice, but its name with its history* audits outworks are all so avowedly Saxon that there seems no good reason for doubting that the central keep or Brocli may also owe its existence to Saxon rather than to Celtic hands. But it is hardly worth while arguing the question here. It is enough to show that something may be said in favour of the Saxon, Danish, or Norwegian origin of these five Brochs. But deducting them from both sides of the question, the great fact remains, that at least ninety-nine out of every hundred of the kno^^m Brochs are situated in the localities known to have been possessed by the Norwegians, and only one per cent., by whomever they were built, witliin the limits of the Pictish kingdom. I cannot of course judge how this coincidence of distribution may strike other people, but to me it appears that, in any other science at least than archasology, it would be considered as nearly conclusive as to their Norwegian origin. Their local is, however, nearly as important as their geographical distribution. The seventy-five Brochs in the Shetland Islands are all, Avith scarcely an exception, situated on the shores of the sea. The same is true of those in the Orkneys, with the exception of eight or ten, of which more hereafter. So I believe are those in the Hebrides and generally in the islands ; but as we have no map of them, this statement must be taken for what it is worth. On the mainland of Scotland the case is somewhat different. They may all be said to be in connexion with the sea, but they are situated in straths and glens frequently at very considerable distances from it, but in almost all cases with what may be caUed repeating towers so situated as to keep up, in a military sense, their communication with the ocean. Taking the case of the islands first, it is evident the Brochs were placed on the sea-shore for one of two purposes — either it was to defend the islands against the attacks of some enemy approaching from the sea, or it was that they might be used as a base of operations, peaceful or warlike, by some people whose business was on the great w^aters. It could hardly be for the first purpose that they were erected ; for even if they were armed with the long-ranging rifled ordnance of the present day they could not defend the islands against even the keels of the Norwegians. They are too far apart, and situated * Proceedings of Royal Society of Scottish Antiquaries, vol. viii.pp. 41 et seq. Trans- actions, ditto, vol. V. p. 1G4. ORCADIAN ANTIQUITIES. 15 generally on promontories singularly ill-suited for this purpose. Besides this, the Broclis, though admirably designed as safes for passive resistance, are, of all fortifications known, the least adapted for active defence. Their active power is limited to a range of some 50 feet or yards from their own walls. Beyond that they are harmless as the rocks on which they stand. If the Brochs had been designed by the peaceful inhabitants of the islands as a defence against attacks from the sea, they would all have been drawn as far inland as possible. Their builders would have tried to block up the entrance of the fiords, sought out inland lakes with islands or inaccessible morasses, and there placed their towers of safety. To plant them on every headland was to court the attack it is presumed they were erected to prevent. If, however, we take the opposite view of the matter, and try to realize the conditions of a Viking's life, it will be seen at once how admirably adapted the situations of the Brochs are for his purposes. The first necessity of his existence was a place where he could leave his family and treasures in safety when he was absent on business, and the next was that he could communicate safely and easily with his treasures without having either to fight his way through hordes of savage Celts or be interrupted by the interference of hostile neighbours. If we adopt this view, the position of every Broch on the islands is easily and perfectly exphcable, while no other hypothesis that I am aware of at all accounts for their situation. Nor does it seem difficult to account for their position on the mainland, for when they were erected there, the Viking seems to have passed into the condition of a Jarl, and territorial ambition seems to have taken the place of the desire for merely maritime dominion. But here again their position seems wholly unintelligible, except on the theory that they belonged to a people using the sea as a base for their warhke operations. If they had been built by the Celts as a defence against the Norwegians, they would have avoided the sea- shore as they would perdition. They knew very Avell that a Broch had no power of preventing a landing anywhere beyond 50 yards of its own walls, and no fortification was ever less provided with means to enable its garrison to sally forth and interrupt an enemy's communications. For all purposes of active or offensive warfare the Brochs are absolutely useless ; for passive resistance they are as admirable as anything yet invented. They are con- sequently admirably adapted for the purposes of an invading power whose purpose it was to establish places of security with the smallest possible garrison in the interior — places where, if attacked, they could remain in security till they were succoured either from the sea, or from the neighbouring garrisons, with whom they had always abundant means of communication. 16 OBCA DIAN A NTIQ UITIES. It may be difficult to make this all clear merely by words, but if any one will look at the Brocks themselves, or study Mr. Anderson's map of them, he can hardly come to any other con- clusion than that they belonged to a sea-faring people, but at the same time to a people who did not use the sea for fishing purposes. They are not fishermen's huts, nor placed in situa- tions where they congregate or where they could haul up their boats. Nor do they belong to a people who used the sea for com- mercial purposes. They are neither marts nor exchanges. They are in fact nothing but the fortified nests of a race of sea rovers ; and whether these were the Celts or the Scandinavians does not seem difficult to determine. If the geographical and local distribution of the Brochs makes out a very strong presumptive case for their Norwegian origin, their architectural arrangements confirm this view in a still more satisfactory manner. If we take into consideration the conditions of Vildng life, his object in settling among a hostile population on a foreign shore, the necessities of the business or traffic he was engaged in, and the class of materials that were available for his purposes, I defy the whole Institute of British Architects to produce a design so admirably adapted to meet all these exigencies as that the builders of the Brochs hit upon. Their circular form was mainly due to the natm'e of the self-faced sandstone -^^dth which they were erected, and to the fact that no mortar was employed in their construction. AYhether this arose from the paucity of hmestone in the districts where the Brochs are found, or from ignorance on the part of their builders, is by no means clear. But when mortar is not used, it is evident that a circular form is the strongest, an angular one the weakest, that can be employed. It is no wonder, therefore, that the former was so universally adopted. They do not, of course, exist elsewhere, simply because no similar conditions of life nor exactly similar materials were found in any other country ; but, hke all objects of true architecture, they tell the tale of their origin with a truthfulness that admits of no mistake. Their purpose would probably never have been mistaken — by an architect at least — had it not been, as Petrie says, that " there is scarcely one that does not afibrd clear proof of subsequent additions by later inhabitants;"* and Mr. Anderson, in his account of his excavations at Yarhus, and throughout his paper, insists on this " secondary occupation," as he calls it, as a proof that "they were in the first instance built by the Celts, and subsequently occupied by the Norwegians.'"! Practically the theory amounts to this, that the * Proceedings Soc. Scot. Ant., vol. vii. p. G5. t Transactions Scot. Mit. yol. y. p. 13i and loi et seqq. ORCADIAN ANTIQUITIES. 17 people who could build ships so strong as to withstand the tempests of the northern seas — fleets so numerous as to be the terror of the islands and of a great part of the continent of Europe — were so helpless when they got on shore, that they were like hermit-crabs, who are content to dwell in the deserted shells of dead mollusks, and all they could do was to make some slight addi- tions to their shells, which, however, the crabs could not. If the additions had differed in kind, and had they been built Avith mortar, or any essential change been observable in their mode of building, there might have been some ground for tliis theory ; but all the variations are mere questions of degree, never of kind; and nothing that has yet been brought to light tends in any way to upset the hypothesis that the same people that built the Broclis built also the additions to them. The fact is, the Broclis under- went the same process of transformation that the Peel towers have been subjected to in every part of Scotland. These are probably quite as numerous as the Broclis, and were originally as uniform in design. They are square towers, three stories in height, with one apartment on each floor, and some sleeping accommodation in the roof. There is generally an external staircase at one angle, and turrets at the other three. How our forefathers lived, or rather pigged, in these towers, is a mystery — a Broch was a far preferable residence; — but as security of property and modern forms of civiliza- tion advanced, wherever these towers continued to be inhabited, — which has happened in some hundreds of cases, — wings were thrown out to afford additional accommodation for the family, outhouses were added, and the rooms of the old tower sub- divided, till it lost all its character of a fortalice, and became the picturesque and commodious dwelling of the modern laird, who, however, was in most cases the lineal descendant of the original tower builder. Precisely the same thing happened in the Orkneys when more peaceful times converted the Viking into a Udaller. He required not only more accommodation, but of a different class from that which satisfied his warlike ancestor. The upper part of the Broch was removed as no longer required. The court was subdivided, and, in some instances at least, roofed, or at least partially so, and outside, drinking-halls and other necessary appliances added, but in the same style and with the same materials. It is in fact a case of "continuous," and not of " secon- dary " occupation ; and so far as any evidence noAv available bears on the question, it goes to prove that those who built the Broclis built also the additions. Or, to put it the other way, those, who- ever they may have been, that built the additions built also the Brochs ; and as it is generally admitted that the additions are the work of the Norwegians, it follows necessarily, in so far as this piece of evidence goes, that the Brochs are theirs also. 18 ORCADIAN ANTIQUITIES, The historical is another point of view from which it is requisite to examine the question of the antiquity of these Brochs before their age can be considered settled, but as in all instances of this sort it is the least satisfactory of any. The Broch builders, hke the heapers up of tumuh or the erectors of stone circles, were not literary, and have left little or no record of their doings in this class. They built but did not write, and modern antiquaries who write but do not build are naturally impatient with their stupidity, and refuse to believe anything for which a written voucher cannot be produced. Unfortunately, for our purpose, the Sagas had many more impor- tant events to record than such every-day occurrences as the building of Brochs. One only instance has come across me in my reading, though there may be many more. It is said that Sigurd in 883 built a Broch at Maerha3fui,* wherever that maybe, but it is very imlikely he would have done so had it been a long- exploded Celtic form and if his brother Jarls had not been in the habit of doing the same thing. But if this is the only piece of direct evidence, the indirect is full and satisfactory. We have several accounts of Brochs being occupied and besieged by the Norwegians, and generally success- fully defended, but no record of any other building — tower or castle — being either built or besieged down to the feudal times. Two of these sieges were of the Broch at Mousa ; one of the most perfect of those now standing. The first was about the year 900,t the second in 1150 A.D.i: In both these instances it answered perfectly the pui'poses for which it was erected, as it proved itself to be impregnable to any mode of attack which the Norwegians could bring against it. The assumption usually made that it was ruinous and deserted on these occasions seems to be wholly gratuitous. It must have been at all events habitable and defensible ; but the great fact that the Norwegians had in these islands round towers which two hundred and fifty years' experience proved to be impregnable and to answer all their purposes, renders it very unlikely that they would go further and seek other models. On the last occasion on which we know of a Broch being besieged the defence was not successful. About the year 1200 Harold laid siege to the Broch at Scrabster in Caithness, in which Bishop John and his followers had taken refuge. The tower was stormed, all the garrison slain, and the bishop blinded and his tongme cut out.§ With the results of this outrage we have nothmg to do here, but the fact of a bishop taldng refuge in a Broch * Skene, Celtic Scotland, pp. 33G, 403. t Orkneyinga Saga, Introduction, cxi. Proceedings Soc. Ant. Scot., vol. iii. p. 187. X Ibid. p. 161. Lubbock's Prehistoric Times, p. 552. § Ibid. Introduction, xliii. Saga l'J7. ORCADIAN ANTIQUITIES, 19 as late as the year 1200, and holding it for some time, against superior forces, shows, so far as it goes, that they were still in use, and no other place of security was then available. It may also be remarked that all the sieges recorded, are of Norwegians by Norwe- gians ; never of Celts by Norwegians, nor of Norwegians by Celts. I do not pretend to be so familiar with the Sagas as to be able to say, of my own knowledge, that there may not be in them passages which may have a different interpretation. But I believe I am correct in saying that no one has yet pointed them out in any writing I have access to ; and till that is done it seems justifi- able to assert that all the contemporary writings tend to prove both directly and indirectly that the Brochs were built by the Norwegians, and were the only fortifications they knew in Scot- land and its isles down to the thirteenth century. If the facts above adduced are to be depended upon, every- thing tends so clearly to prove that the Brochs are of Norwegian origin that it would require some very strong evidence on the opposite side to establish a contrary conclusion. This, however, the advocates of their prehistoric origin fancy they have obtained in history of the Broch at Okstrow excavated by Mr. Leask, the farmer of Broadhouse, on whose lands it was situated, and de- scribed by Mr. Petrie in the fifth volume of the "Scottish Archaeo- logia," p. 76. Its evidence is at first sight, it must be confessed, a little startling, but when carefully examined it turns out to be one of those moleliills which a pleader magnifies into a mountain, but which a judge brushes on one side as either irrelevant or immaterial. Its appearance is that of a Broch whose walls are standing only to the height of 5 to 10 feet, buried in a mound about 15 feet in height. In tliis mound were found a number of cists containing human remains, showing evidence of interment by cremation. These, Mr. Anderson thinks, may be as late as the tenth century,* and he is probably right in this ; but even assuming this to be so, they are found superimposed on the remains of a ruined Broch, and the usual theory is that the Broch was deserted and ruined so long ago that not only its existence had been for- gotten, but that the earth had grown and raised itself 5 to 10 feet above the top of its still standing walls, thus tempting the Norwegians to use the mound so raised as a convenient burying- place. If this is so, it is evident the Broch may be of any con- ceivable antiquity, for the mound does not seem to have been composed of the stones that formed the upper part of the Broch, which would hardly have formed a tempting place for being hoylaid," but was composed of earth ; and how long it would have taken the earth to heap itself up in that fashion it is * Archccologia Scotica, vol. v. p. 178. 20 ORCADIAN ANTIQUITIES. impossible to say.* As so mucli importance Las been attaelied to it, it is very much to be regretted that Mr. Petrie did not excavate this monnd himself. The greater part of it must have been ]-c- moved before he saw it, as a considerable portion of the wall of the Broch itself had been carted away down to the foundation, and the interior must have been entirely excavated, as he doubts the existence of a wall which he was told stood across the court when the excavations were commenced. All we know of it is consequently at second hand ; and though Mr. Leask may no doubt be trustworthy and intelligent, he had no previous expeiience in works of this sort, and the observations of uneducated eyes are seldom to be depended upon. From the section, however, as we have it,t it is evident that those who used the mound as a burying-place must have known of the existence of the Broch, as some of the cists are placed within a few inches of the tops of its walls ; and if so, it is very unlikely the Norwegians would have chosen the ruins of the habitation of a despised Celt as the resting-place of their honoured dead. It is equally to be regretted that Mr. Petrie did not see " the piece of silver like the head of a walking-stick, 2 J inches long by I of an mch in diameter," which was found at the very bottom of the well (sepulchral pit ?) in this Broch. It could hardly have been prehistoric, but much more probably belonged to a Yiking. The " fragment of Samian ware with holes bored in it for mend- ing " could hardly have formed part of the furniture of a Celtic Broch, and if found in one of the cists — we are not told where it was found — would at first sight appear to argue a greater antiquity than could well belong to a Viking's grave. The first idea that occurred to me to get over this difiiculty was that the Jarl's wife had been affficted with a mania for old china, hke many a countess since her day, and had insisted on her pet specimen of cracked Samian being buried with her. A more prosaic explanation, however, is to be found in the fact that Mr. Loranje, attached to the museum at Christiania, has recently, among some hundreds of tombs he has explored in Norway, found eighty chambered tumuli whose dates he was able to ascertain from the Boman or quasi-Roman remains they contamed. From their similarity with corresponding objects in the south of Europe, he fixes their dates as ranging from the third to the seventh cen- turies.J There are, however, two elements of uncertainty here : * I have before this had occasion to remark on the curious property matter had in prehistoric times of divesting itself of that gravity it always is assumed to possess in the historic period, and to climb into places which in modern days it could not have reached without the assistance of man, or some similar agency. (Rude Stone Monuments, p. 338.) But the earth, of its own accord, heaping up this mound of Okstrow, surpasses any other example I know of. t Archaeologia Scotica, vol. v. p. 7f). X Compte Rendu du Congres International, etc., a Stockholm, en 1871, p. G44. ORCADIAN ANTIQUITIES, 21 first, that the Southern objets d'art " were immediately copied in the North, hke "objets de Paris" at Birmingham; and, secondly, that they were buried as soon as made. We may certainly allow a century or two for these two constituents, except in the case of such as may have been imported from the South; but, be this as it may, if there are at least eighty tombs in Norway containing Roman or Byzantine remains, we need not be astonished at finding a fragment of Samian ware in a Norwegian tomb in the Orkneys, whatever its age may be. It would be tedious, as it would be out of place here, to go through all the theories that have been or may be raised re- garding this Broch at Okstrow, and I may consequently, perhaps, be allowed to state that after carefully weighing all the evidence bearing on the subject, the theory that appears to me best to meet all the circumstances of the case is that the walls of this Broch — if Broch it is — never were carried higher than we now find them. When they reached that point the builder may have died, and the old pagan Viking may have been buried like Joab, "in his own house in the wilderness,"* and his family and depen- dents buried around and over him. All this reasoning proceeds on the assumption that the buried building is a true Broch, but this appears to me, to say the least of it, extremely doubtful. Its dimensions are quite exceptional. Its court-yard is 45 feet in diameter, or one-third more than usual; its walls only 12 feet thick, and, if tapering as these walls generally do, totally unfit for the purposes to which the walls of Brochs were usually applied. I believe with the venerable Nilsson that the tombs of the Scandi- navians were " a copy, a development, or an adaptation of a dwelling-house," and " that they buried the house with its owner, and the grave was literally the dwelling of the dead."t I am, indeed, fast coming to the conviction that many of the Brochs which have hitherto been supposed to have been erected for the defence of the living were originally only intended as the resting- places of the dead, and that this Broch at Okstrow is one of the latter class. But of this more hereafter. Meanwhile there may be other theories which may account as well as this for the observed facts, or better. But of one thing I feel certain, that the prehistoric theory neither accords with the recorded facts, nor with any reasonable view of the whole circumstances of the case. The Broch that was found buried in a shell mound or Kitchen Midden, in the harbour of Keiss in Caithness, is another on which great stress has been laid to prove their remote antiquity.t If we are to adopt the doctrines now fashionable among antiquaries, * 1 Kings ii. 34, t ^ir John Lubbock's Prehistoric Times, 3rd ed. pp. 134, 135. X Prehistoric Remains in Caithness, by Samuel Laing, M.P. : London, 18G6. See also Proceedings Scot. Ant. vol. viii. pp. 192 et scqq. 22 ORCADIAN ANTIQUITIES. and believe that all middens must be prehistoric which do not contain implements of bronze or iron, or cannot prove from their contents that those who heaped them up were men in an advanced stage of civiUzation, of course this Broch — if Broch it is — must be very old. But middens very similar in their contents to this one have been aggregated in New Zealand or Terra del Fuego, and on the west coast of Ameiica north of Vancouver's Island, up to within the limits of the last century. There is nothing what- ever to show that a settlement of rude Celtic fishermen may not have existed in the Bay of Keiss in as low a state of civilization as either the Xew Zealanders or North Ameiicans down to within the last 200 or 300 years, or like the dwellers in the beehive houses and bothan in the Hebrides at the present day, whoso stage of civilization, if Captain Thomas is to be beheved, is httle, if at all, more advanced than that of the New Zealanders or North American savages.* The beginning of a shell mound maybe 1,000 or 10,000 years ago, but till it can be shown Avhen it ceased to grow, or when the inhabitants of the place became too civilized to continue it, no argument can be based on its age. In this instance, at all events, the evidence is infinitely stronger which goes to prove that the Broch was erected within the last 1,000 years than any that can be adduced to show that the inhabitants of Keiss had reached a higher state of civilization than their midden indicates withm the last two or three centuries. If all this is as clear and intelhgible as it is above represented to be, the question may naturally be asked how it came ever to be doubted ? The answer, however, is easy. It never was, and my behef is, never Avould have been, had not the empirical Danish system of the three ages become the fashionable creed of this country. According to it every building that cannot produce a written certificate of age, attested by contemporary mtnesses, must belong to one of the three prehistoric ages. The Brochs cannot luckily be carried back to the stone age, as "no flint arrow-point, flint celt, pohshed stone axe, or perforated stone hammer has yet been found in any one."t Both bronze and iron unplements have, however, been found, but not in such a manner as to enable these systemists to make up their minds whether they belong to the bronze or iron ages. All, therefore, that could be done was to declare them prehistoric and leave it to accident to deter- mine hereafter to which of these ages they should be assigned.^ * Proceedings Soc. Scot. Ant. vol. iii. pp. 127 et seqq. and 13-i et seqq. t Orkneyinga Saga. Introduction, p. ex. X No one, of course, denies that savages all the world over nsed stone implementa before they became acquainted with the use of metals. It has also been ascertained, as a historical fact — however it may be accounted for — that in some countries of Europe at least, men used bronze for certain purposes before they employed iron. But it by no ORCADIAN ANTIQUITIES. 23 From such trifling, it is a pleasure to turn to the honest and distinct utterance of such a man as Sir Walter Scott. His intimate knowledge of the antiquities of his country and strong common sense enabled him to see at a glance what their true origin was. After describing in detail the Broch at Mousa, which he had per- sonally inspected with great care, he adds, Such is the general aspect of that very early rude period when the Northmen swept the seas and brought to their rude homes, such as I have described them, the plunder of polished nations."* Most of his contem- poraries acquiesced in this decision, but it is now singularly un- palatable to their successors. They consequently wrote to Worsaae and Munch, the two principal apostles of the new creed, to ask them if any such towers existed in Scandinavia, and on receiving their answer that nothing of the sort was to be found there, they were satisfied that they were not built by the Norwegians, and must consequently be prehistoric.t Had the Brochs been either Temples or Tombs there might have been some relevance in the question and answer. Where sentiment or tradition alone guide the architect in making a design, it is generally very difficult to discriminate between what is due to race or religion, what to imitation of forms existing in any country, though belonging to alien tribes, and how much is due to constructive necessities. It requires in these cases great care and discrimination to feel quite sure of the conclusion arrived at. But when use and con- venience only are thought of, there is only one idea that pervades the whole design, and that is common sense. Whether their design grew up gradually through a long series of years, or sprung at once with ]\Iinerva-like completeness from the brain of some Scandinavian Vauban, there is nothing in the arrange- ments of the Brochs that was not designed with direct reference to the places in which they were situated and the uses to which they were to be applied, and it is consequently useless to look for them in places where these conditions did not exist. The probabilities are that in his own country the Viking lived in picturesque towns or villages on the shores of some far- inland-stretching fiord ; his abode a one or two-storied log-house, covered externally with quaint carvings, and internally with all the means follows from this, that any nation abandoned the use of stone as soon as ten, or it may be a hundred, of the richest inhabitants were enabled to purchase bronze s■v^•ords of foreign manufacture. Nor does it follow, that long after iron was introduced, they did not continue to use bronze weapons and brooches for all festal and ornamental purposes. In the countries regarding which we are at present wiiting, it is extremely doubtful if any individual in the Orkneys was rich enough to buy a bronze sword, and they certainly could not manufacture one, before the advent of the Norwegians, and nearly certain that the aboriginal Celts continued to use stone implements — passing over the bronze age — for centuries after their masters were as familiar with the uses of iron as we now are. * Note to chapter xviii., Ivanhoe. t Proceedings Scot. Ant. Yii. p. 6G. Worsaae, Danes and Northmen, p. 233. 24 ORCADIAN ANTIQUITIES, appliances of a cheerful, if not of a civilized life. He had nothing to fear from attacks of any enemy coming from the sea, and the government at home must have been sufficiently organized and powerful to render life and property secure. AYithout this they never could have accumulated the stores or organized the labour required to build and maintain the fleets with which they obtained and kept the dominion of the seas, and in such circumstances a Broch would have been as much out of place as the knight in armour is in a lord mayor s show. If the Norwegians required castles or strongholds of any sort, they most probably were square towers, like those subsequently erected in the south of Scotland, and elsewhere. The nature of their building stone was different from that found in the Orkneys ; and they had limestone, and knew apparently how to use it for mortar. When this is the case, a rectangular form is not only more economical, but much more convenient, and as strong as a circular one, and would conse- quently, in these circumstances, be everywhere adopted. Whenever and wherever men are thinking only of protecting their persons and property, they may be depended upon to use the best means available for that purpose without thinking whether other nations adopt other means, or whether people of their race are content with less complicated appliances while living under different circumstances ; and the Brochs were admirably adapted to the conditions of Viking life in the Orkneys, and for utilizing in the best manner the materials they found existing there. They were not however so necessary in their own more settled homes in Norway ; while, as the constructive necessities of the locality did not require the circular form being adopted, we need not feel surprised that no Brochs are found in that country. Had they, on the other hand, been of Celtic origin, they certainly would now be found in Pictland and in Ireland as frequently as in the north of Scotland. The conditions of Celtic life in these countries, must have been the same, at one time or another, and would consequently have been exemplified by the same class of buildings. It need hardly be added that no Brochs are found in any countries known to have been occupied exclusively by Celts, except the two in Forfarshire above alluded to. Tombs. Whatever others may think of the above chain of reasoning, to me it seems so clear and conclusive that I cannot but assume that the age and use of the Brochs are established by it, and all that follows is consequently based on that conclusion. If it is so, it simplifies very much the condition of the problem ; for it is not so much the question whether the tombs are those of the Celts or ORCADIAN ANTIQUITIES, 25 of the Norwegians, as whether the Broch builders erected also the various mounds or edifices that are found everywhere mixed up with them. I have no hesitation in answering this question in the affirmative, and would, I fancy, find no difficulty in proving the truth of this proposition to others were it not that our one infallible guide — common sense — here forsakes us. As before mentioned, in dealing with tombs or temples, sentiment and tradition play so important a part, that they govern our conclu- sions to a great extent. It is true that if we knew all we desire to know, they may, and frequently do, lead us to the most curious and interesting ethnographical and historical revelations ; but they also may lead astray where our information is defective, and they never are convincing to those who have not sufficient knowledge to enable them to follow or appreciate the reasoning on which the conclusions are based. Still the sepulchral remains in the Orkneys show a style of art so similar to that of the Brochs, and both represent a state of civilization so nearly identical, that it will be difficult to separate the one from the other, or to maintain that the tombs belong to the extirpated Celts while the Brochs Avere erected by their conquerors. Taking, as before, the topographical question first : if a circle is described with a radius of five miles round the stones of Stennis, it will include within it the principal group of Orcadian antiquities, the two stone circles so often compared to Stonehenge, and several earthen circles which seem of the same land, though carried out in difi"erent materials. It includes also Maeshow and several large conoid barrows, besides numerous ovoid and bowl barrows. There are also eighteen or nineteen Brochs within this area, a larger number than are to be found within so limited a space in any other part of these islands. One half of these, too, are not situated on the shore, as all the other Brochs in the Shetland and Orkney Islands are, but are grouped nearer the monuments, some way from the coast line.* It appears therefore, so far as this evidence goes, that this was the favourite spot, and the one most frequented in the islands by the Broch builders, though it is neither the most fertile nor the best suited for their purposes. If all this is so, it would indeed be strange that these Broch builders should settle here, among the tombs and temples of their despised predecessors, and more strange that they should have left them standing as we now find them at the present day. If this, too, Avas their favourite settlement it seems unaccountable that they have left no traces of their own sepulchral rites and arrangements anywhere in the neighbourhood, which would be the case if the Celtic theory were correct. * The facts are taken from Mr. Anderson's map, Arcliaoolog'a Scotica, vol. v. p 199, and Captain Thomas's map in yol. xxxiv. of the Archoeologia, pi. xii. 2G ORCADIAN ANTIQUITIES, Before going further, however, it will be convenient to discrimi- nate a little between the various classes of sepulchral monuments found in the islands. The first are the great straight-lined conoid barrows, of which Maeshow is the chief. Some of these contain chambers of more or less extent ; others only cists with urns, exhibiting evidences of burial by cremation. A second class, in the unscientific style of nomenclature too frequently employed by antiquaries, are called ^'Picts' liouseft/' though they never were originally intended as dwellings for the living, but always resting-places for the dead ; and to apply to them the distinctive appellation of Picts is simply to prejudge a question which, to say the least of it, is still siih judlce. Like the Brochs, many of these so-called Picts' houses show evidences of secondary occupation ; but of a very different character. It is no longer by additions or improvements, but simply, that when they either had been opened like Maeshow from above, by robbers in search of treasure, or their earthen envelope had been denuded by rain so that access could be had from above, they were occupied as dwellings by the degenerated Celtic population. Any one who has travelled in the East and knows how generally the deserted tombs of the supeiior races are inhabited by the Aborigines, will understand exactly how this came to pass, and will recognize at once, as usual and familiar, all the phenomena presented by such tombs as those at Wideford Hill,* or Kettleburn,t and other sepulchres of the same character.J They are generally low and long mounds, and always contain one or more chambers con- nected with one another by passages, and communicate with the exterior by a very narrow entrance ; they are, in fact, the counterparts of the Gang-graben, or passage graves, of Northern antiquaries. A third class are the circular or bowl-shaped barrows : they are generally small but infinitely numerous, and scattered everywhere over the island. Generally it has been assumed that they are the graves of the Celts, while the straight-lined tumuli belong to the Scandinavians.§ It may be so, but it is hardly worth while arguing the point, here at least, as notliing hinges on it. My impression is that the Northmen may have copied this form from their prede- cessors, and it may consequently have been used by both races. The horned tumuli of Caithness, so-called, and others of fantastic shapes, form a fourth class, the counterparts of which are frequent in Scandinavia, but, as they are not found in the islands, hardly come within the limits of this paper. As mentioned above, * Wilson's Prehistoric Annal, p. S-i. t Arch£Bological Journal, vol. x. pp. 212 et seqq. \ Proceedings Scot. Ant., vol. iii. pp. 188 ot seqq. § Archeeologia, vol. xxxvi. p. 107. ORCADIAN ANTIQUITIES. 27 I ara fast coming to the conclusion that it will be necessary to form a fifth class out of the Broch-like tumuli. Such mounds as the Broch at Okstrow, and that excavated by Petrie at Hoxay, of which more hereafter, may become numerous when looked for.* In attempting to ascertain whether Maeshow and the other Orcadian tumuli owe their origin to the Picts or to the Nor- wegians, the first thing that strikes us is that all the a jyno7'i probabilities are in favour of the latter hypothesis. All the Northern men who flocked to these islands after the departure of the Romans seem to have buried their dead in tumuli — in Eng- land down to the tenth century certainly, and in the farther north to a still later period. On this point Sir John Lubbock may be accepted as an authority. The Codex Diplomaticus," he says, " contains references to more than sixty barrows or lows bearing the names of particular persons ; some of them — for instance, Wodnes Beorgh, or Woden's Barrow — are probably mythical ; but there is no reason whatever to doubt that some — for instance, Alfredes Beorh, ^thelwolde's Beorh, Cissan Beorh, Cwichelme's Hloew, Oswolde's Hloew, &c. — retain the name of the person who is really buried within. It appears that in England the habit of burying under tumuli ivas finally abandoned during the tenth century, The italics are mine. As, however, all the sixty names are Saxon,t and not one Celtic appellation is found affixed either to these or, so far as I recollect, to any barrows in England, their evidence, so far as it goes, is entirely in favour of a Northern origin. The most apposite examples, however, bearing on this subject are the often quoted tombs of old King Gorm and his wife the Enghsh Thyra at J ellinge, in Denmark. They belong undoubtedly to the end of the tenth century, and are, mutatis mutandis^ identical with Maeshow, the principal difference being that in Denmark, stone being scarce and wood abundant, the chambers of the tombs are lined with logs, while in the Orkneys, where the opposite conditions prevail, stone only is employed for these purposes. § The ultimate decision of the question, in so far as analogy is concerned, will probably be determined by the similarities or * In the first volume of the Scotch Archseologia, published in 1792, there is a paper by Colin Mackenzie, " Some Antiquities in the Island of Lewis." In a plate (p. 287), illustrative of the paper, two round towers are presented. The first, the Dun of Car- loway, is a true Broch, presenting all the known characteristics of these buildings. The second, however, the fort at Cromore, though badly dra^vn and engraved, is unmis- takably sepulchral, though at first sight so like a Broch as to deceive the unwary. Mr. Mackenzie adds, " These are universally agreed on to have been built by the Norwegians." Pennant merely calls them Danish towers, but does not pretend to know who built them. t Pi-ehistoric Times, 3rd ed. p. 117. X John M. Kemble, in Archaeological Journal, vol. xiv. pp. 119 ot seqq. : an article that ought to be considered final on this subject. § Rude Stone Monuments, pp. 244 et seqq. 28 ORCADIAX AXriQLiriES. differences tliat maybe detected between tlie Orcadian tombs and the eighty sepulchres above alluded to, which have recently been excavated in Norway by the officers of the Christiania Museum. All their chambers are lined with stone slabs (dalles), and all are subsequent to the Roman times — for reasons stated above, I believe extending down to the ninth or tenth centmies. Un- fortunately no plans or sections of these tombs have yet been pubhslied, only verbal desciiptions, and these do not suffice for any satisfactory comparisons;* but the fact of so large a number of this class of grave being found in Norway should probably be considered as nearly decisive on the point. The examj^les on which antiquaiies seem most to have relied to sustain an opposite view of the origin of the Orcadian tombs are those situated on the banks of the Boyne near Drogheda. As the evidence now stands there seems no reason for doubting that these Iiish sepulchres were erected in the early centuries of the Christian era, before the conversion of the Irish to Christianity by the preaching of St. Patrick; t and if they were even nearly identical, the case would be a strong one. No one, however, can examine the very rude and clumsy construction of the chambers at New Grange — the finest of these Iiisli sepulchres — and compare it with the finished elegance of Maeshow, without perceiving that — supposing the conditions to be the same — a very long interval must have elapsed between them. "Whoever built it, the chamber in Maeshow is the most perfect and elegant known to exist in any sepulchre of its class on this side of the Alps, and consequently, according to the usual laws of architectural progression, may be assumed to be the latest, and erected by the people most capable of such magnificence. Whether this, however, is admitted or not, a considerable interval certainly elapsed between its design and that of the Iiisli tombs. The difficulty of ascertaining how long that may have been aiises principally from our ignorance of the circumstances of the two peoples at the time they were erected, and more because we do not know whence either the Irish or the Northmen first derived their models. Must we go back to the treasuries at Mycenae or to the tumuli on the Steppes for the original from which one or both derived their first idea'? Till in fact we know more than we do noAv we must l:e content with the certainty that Maeshow is more modern than New Grange ; how much so, must depend on what we can ascertain regarding the histoiy of the two monuments. Our information on * The information regarding these tombs is contained in four reports published in Norwegian, in the ^Memoirs of the Archaeological Society of Xorv.-ay, in l!^67, 18GS. 1860, and 1870, and resumed by M. Loranje, in a work entitled " Om Spor of Romersk Kultur i Norges a^ldre Jernalder." It need hardly be added that neither of these works is to be found in the British Museum Library. But they do exist in that of the Society of Antiquaries, where I have had an opportunity of consulting them. t Rude Stone Monuments, pp. I'J'J ot seqq. ORCADIAN ANTIQUITIES. 29 this head is necessarily somewhat vague, but as it stands there does not seem a shadow of reason for supposing that the Pictish inhabitants of these isles were ever either so organized or so far advanced in civiHzation as to attempt such a monument as Maeshow before their conversion to Christianity by St. Columba in the seventh century. They certainly were immensely behind their Irish contemporaries in this, as in every other respect, and accomplished nothing to be compared with it in Pictland, or any other part of Scotland where they were in larger communities and nearer the Roman and Irish civilization. After their conver- sion they did not want such a tomb as this. The Northmen on the contrary we know did bury in " hangs " down at least to their conversion at the end of the tenth century, and were sufficiently powerful and civilized to erect such a tomb without difficulty. Maeshow and Mousa are, in fact, the counterparts of each other, in so far as either constructive skill or artistic adaptation of means to an end is concerned, and cannot consequently, I believe, be far removed from each other in date. We are not, however, without direct written authority on this subject, though it may not be quite so full as might be wished for ; but the same Sigurd who built the Broch at Maerhsefui, as mentioned above, was buried in a How at Ekkialsbakki on the Dornoch Firth, where his grave was known in the thirteenth century and could how, no doubt, be identified if looked for.* Torfinn (Hausa Kliuf, A.D. 963) was buried in the How at Hoxa, and there his remains will probably be found when looked for.f Some years ago I tried to induce Mr. Petrie to excavate this How, but he was deterred by the magnitude of the undertaking, though I offered to pay the expense. Some twenty or thirty years ago a Broch had been excavated from this How, and may have been, sepul- chral, t but, as the existence of this form was not then suspected, we are left in doubt. Instead of the great How, Mr. Petrie excavated a smaller one about 110 yards from it. He found it to contain a building which he suspects to have been sepulchral, though it retains many of the features of a Broch. § In this, I beheve, he is correct, but whether it is the tomb of Thorfinn or not will only be known when the great How is excavated. Mr. Anderson points out several instances, especially in Caithness, where Brochs have been used as places of interment ;|| but as he did not suspect they could have been built for that purpose, he assumes that these were cases of secondary occupation. If, how- ever, the engraving on plate xvii. of the thirty-fourth volume of the Archaeologia is to be depended upon, the so-called Pict's house * Orkneyinga Saga, p. 107. % Arcliaeologia, xxxiv. p. 120. t Loc. cit. p. 207. § Proceedings Scot. Ant. vol. ix. p. 363. II Arcbseologia Scotica, vol. v. pp. 152 et seqq. 30 ORCADIAN ANTIQUiriES. at Burgher in Evie is a Broch of the usual dimensions and arrange- ments, but certainly erected originally as a resting-place for the dead, and never for the accommodation of the living, and mth the Brochs at Okstrow and Hoxa and many others may one day swell our fifth class to most unexpected dimensions. Be that as it may, all that is contended for here is that these Tombs and the Brochs and Picts' houses are all so inextricably mixed up together that it seems impossible to separate them, and to avoid the con- clusion that they are all of one age. When he first began his excavations, Mr. Petrie seems to have had an idea that the Picts' houses may really have been the dwellings of those savages ; but when his knowledge was greater, he wrote — In short, it apj^ears to me that Maeshow is identical with the so-called Picts' houses. If therefore the former was originally a chambered tomb — of which there seems to be no doubt — the latter may without hesitation be classified as the sepulchral buildings of the early colonists of Orkney."* If the word early" were omitted from this sentence, I believe almost all those who have studied the matter would agree with it. They are all so evidently of the same age, that if the date of one were ascertained the age of the rest becomes known within very narrow limits. When I wrote last on this subject,t I suggested that Havard was buried there, because he was slain in the battle close to this spot, in the year 970 A.D., and from his eminence and other circumstances, I thought he was woiihy of the most mag- nificent tomb in these islands.^ Mr. Anderson, on the contrary, suggests that he was buried on one of the larger tumuli close to the stones at Stennis. § To this I see no objection, if any one prefers it. All I contend for is, that these tumuli are the gi'aves of the Norsemen. And it may be that Maeshow is only the grave of some Viking who was more than usually successful in his career, and may have been exceptionally ambitious of sepulchral magnificence. It often happens in India that the tombs of subjects are more splendid than those of their masters ; and if this is the tomb of an early Viking, and not of a quasi-regal Jarl, it would remove some of the minor difficulties that prevent its general acceptance as a Norwegian tomb. All I care to contend for here is, that it is Norwegian, not Pictish. If that is conceded, the name of its original occupant maybe left for future determination. * Journal of Archaeological Institute, vol. sx. p. 37. t Rudo Stone Monuments, p. 250, to -wliicli the reader is referred. X The Quarterly Review, in a note at page l-i6, contends that the Pict's house at Papa Westray is not less remarkable than Maeshow, " and is even more impressive." No man, I believe, would make such an assertion who had seen either, or who was capable of forming a correct idea of a building from a plan. So far as I am capable of forming an opinion on such a subject, I have no hesitation in stating, that whether regarded from an artistic or a structural point of view, Maeshow is superior to the other in every respect, and more impressive in the ratio of at least 10 to 1. § Orkneyinga Saga, p. cviii. ORCADIAN ANTIQUITIES, 31 Circles. It only now remains, before concluding, to say a few words regarding the circles which are found at Stennis in the Orkneys, at Callernisli in the Hebrides, and occasionally on the mainland. They, however, are less communicative than even the Brochs or the tombs, and without some extraneous evidence it is difficult to say anything regarding them that is satisfactory. Here use and common sense wholly desert us. They were obviously erected because they gratified some feeling, or satisfied some aspiration which we do not quite comprehend, and till we do, we are groping very much in the dark regarding them. Still the setting up of a circle of stones is so obvious a way of marking off a sacred spot, either for worship, or for interment, or for assembly, or for any other purpose, that once done we need not be surprised it was repeated, and it may be by different races, at long intervals of time, and for many various purposes. With all these elements of imcertainty, it is pleasant to hear the voice of one so well qualified to speak as Sir Walter Scott pronouncing on the question without doubt or hesitation. " Tra- dition," he says, " as well as history, ascribes the stones of Stennis to the Scandinavians."* As far as tradition is concerned he is probably right, and could not express himself too strongly ; but if there had been any historical record of their erection in the Saga or anywhere else, it could hardly have failed to be quoted and insisted upon, and the controversy never would have arisen.f Even assuming that there may be some foundation for the conten- tion of the modern school of critics, which insists that nothing shall be received as authentic which was not recorded by com- petent contemporary witnesses, one may be allowed to protest against the further extension of this system, which has become fashionable in archaeological matters, and which excludes the testimony of any writer who, in narrating an event, adds any par- ticulars of an incredible nature, or can be convicted of an excess of credulity in any other part of his works. To do this is simply to shut out of court all Avriters of history who lived before the Reformation, and to afford a tabula rasa on which all sorts of prehistoric fancies may be written. It is a more difficult and laborious process to sift the evidence and gather up the grains of * Note to chapter xxxviii. of The Pirate. t What Sir Walter was probably thinking about when he used the word history, was an abstract he published from the Eyrbyggia Saga, in his "Illustrations of Xorthern Antiquities," p. 480, in which it is said — " The tirst object of the Scandinavian colony from Norway, when landing in Iceland, was to erect a temple to the god Thor, which is described in the Saga as a circular ring of upright stones." This, he adds, may confute those antiquaries who are disposed to refer such circles to the Celtic tribes and their Druids. All this is true enough, and an excellent corroborative example, but still not an historical notice of the stones at Stennis. 32 ORCADIAN ANTIQUITIES, truth that frequently lie buried under vast heaps of fable ; but it is, I believe, the only mode by which we shall ever arrive at any satisfactory conclusion regarding the age or use of these ancient monuments.* Neither in this, nor it is to be feared in any other instance, will history, in the sense in which that word is now understood, answer our inquiry regarding these rude stone monuments.f The people who erected them were utterly illiterate and wrote nothing ; and those who insist on rejecting every other evidence than con- temporary written records are asking for what they know does not exist, and, in some cases at least, asking it, not in the cause of truth, but because a challenge that cannot be accepted looks like the defeat of a rival, to those who are ignorant of the circum- stances of the case. If, however, we must in this, as in most similar cases, abandon the hope of direct written testimony, nothing can well be stronger than the traditional evidence of their Scandinavian origin. No one can well read the narrative of Principal Gordon, or the quota- tions from Dr. Henry in Dr. Hibbert's paper in the third volume of the " Scottish Archa3ologia," without being struck by the fact that the courts of law in the Orkneys, at the end of the eighteenth century, should gravely consider it an aggravation of an offence that the offender should have broken the Promise of Odin, made at the holed stone of the circle of Stennis. Looking at the tenacity with which the nations of Saxon origin cling to the Pagan practices of their forefathers wherever they exist, and the care with which they are handed down from generation to genera- * An amusing instance of how modern antiquaries treat the Utera scripta will be found in a quarto volume just published by Mr. W. Long, M.A., for the ^Yiltshire ArchjBo- logical and Natural History Society. It is a compilation of all that is known about Stonehenge, with quotations from all the authorities, from Henry of Huntingdon (1154), till the time when Dr. Stukely deceived his contemporaries by falsifying facts, and inventing fables, which have been too credulously accepted by some cf his successors. After printing fifty pages of these authorities, nine-tenths of whom favour the idea that the building was erected in the fifth century, Mr. Long quietly throws the whole of them aside, and expresses his belief that his friends Mr. Warner and Dr. Guest are right, and that it was erected by the Belgao before the Christian era. There is no pre- tence that any such monuments exist in Belgium, and not one of the authorities he quotes so painstakingly, alludes to the Bclga>, either directly or indirectly. It merely is that he and his friends having made up their minds that the building must be pre- historic, some prehistoric people must be found to whom it may be ascribed, and the less known about this people and their doings the better for the theory. t I cannot, however, help sometimes fancying that the Ring of Brogar is mentioned in the Saga. At least in Mr. Anderson's book, page 208, quoting from the Saga of Olaf, where it is said — " At this time Havard was at Steinsness, in Hrossey. There they met and there was hard fighting, and it was not long till the Earl fell. The place is now called Havard's Teigr." In a note it is explained that Teigr means an individual's share of the Tunland. In the " Archaeologia," vol. xxxiv. p. Ill, it is further defined thus : " Teigr, cultivated land of indefinite size, enclosed within a turf or stone dyke." Such a division of Tunland could have no possible application to a place where a king fell. But in a secondary sense it may have come to mean any distinct allotment or enclosure, and if this is go, would be singularly appropriate to the circle within which those slain in the fight lay buried. But this is a point which must be left for the deter- mination of scholars more familiar with Scandinavian languages than I can pretend to be. ORCADIAN ANTIQUITIES. 33 tion, it seems impossible tliey could be mistaken in this instance. It is equally improbable that if these monuments had belonged to the despised Celts they should have appropriated it in this manner, and dedicated it to one of their gods, or looked upon it with any reverence eight or ten centuries after its erection. It seems equally improbable that the builders of the Brochs, which are so frequent in close proximity to these circles, should not have utilized them for their buildings, or for bridges or dams, which are much wanted in this neighbourhood, unless they were restrained by some superstitious feeling from so doing. Their preservation in fact is nearly as wonderful as their erection ; and had they not been regarded with reverence by the natives, they hardly would have reached our day in the perfect state in which we now find them.* In the absence of any direct testimony to that effect, one of the most obvious reasons for believing that the circles are of the same age as the tombs and the Brochs is tlie mode in which they are all mixed up together and apparently parts of one contempo- raneous group. It has been attempted to argue that the age of Stonehenge depends on that of the barrows which are found in its neighbourhood, though this opinion would hardly be maintained by any one who observes how evenly the barrows are dispersed over the whole fifty miles that extends from Dorchester to near Devizes. Here, on the contrary, any one who studies Captain Thomas's map in the thirty-fourth volume of the " Archa3ologia," must, I fancy, feel convinced that the monuments around the lake at Stennis must be connected with one another by some other ties than those merely of juxtaposition. The actual proof whether this is so or not, in so far as the circles at least are concerned, can only be arrived at from a careful examination of all similar monuments in England and Ireland. This of course cannot be attempted here, and without illustrations could hardly be rendered intelligible. It has, however, been attempted in my wo]k on " Kude Stone Monuments," and till some flaw in the argument is found — I at least believe — with success. Meanwhile there is an example so apposite that it may be quoted without carrying us beyond our limits. In or about the year 750, a great battle was fought on the Braavalla heath in Sweden, in which the king, Harold Hildetand was slain, and near which he was buried under a great mound, with all the pomp of heathendom. No one, I believe, doubts that the obelisks, the stone circles, the mounds, and other * In Wilson's " Prehistoric Annals," p. 102, will be found a curious account of tbe universal persecution that forced the unfortunate farmer to fly the country, who in 181-4 destroyed Odin's stone. Had he desecrated Kirkwall Cathedral, he might have baen fined or imprisoned for a few days and been little the worse, but to desecrate the Altar of Odin was a crime that no descendant of the Norsemen could ever either forget or forgive. 34 ORCADIAN ANTIQUITIES. nionmnents that still stand on that heath,* are the memoiials of that great struggle ; and the similarity between them and those that surround Ilavard's Teigr, near Steinness, is so striking that they can hardly have been placed there for different purposes. There are many other similar collections of circles and tumuli in Scandinavia, ^vhich, if drawn with care, would not only illustrate the history of their own countiy, but throw light on ours ; but the antiquaiies of that country, having made up their minds that they belong to the stone age, and are consequently prehistoric, dechne to waste their energies on the unknowable. My impression is, on the contrary, that there is not one of them that has not a history, and one that can easily be ascertained, so soon as the antiquaries will give up the absurd theory of the three ages, and be content to let every moniiment tell its own story without reference to any empirical system. It would be easy to extend these remarks to at least twice the extent, easier, in fact, than to write them so short as is here done, and it would require at least ivdcQ the amount of illustrative examples drawn from other countries to establish the propositions here announced with anything hke certainty. But such was not the aim or intention of this short essay. All that was proposed in wi'iting it was, while confining attention as closely as possible to the Orkneys, to show, in the first place, that there was no real difficulty in discriminating between what belonged to the Chris- tian Picts and Celts, and what to the Pagan Normans ; in the second place, to make out at least a strong 'prima facie case for the Brochs, the conoid and chambered tumuli, being the works of the Norwegians before they were converted to Christianity, or attempted such churches as that at Egilsey or the glorious cathedral of St. Magnus at Kirkwall. If the value of architecture for determining ethnographic or historical problems were at all understood or appreciated in this countiy, my impression is that these propositions would never have been mooted or thought doubtful for a single hour; but it is unfortunately equally my conviction, that till a more philosophical ^-icAv of architecture as a science shall prevail, the value of these Rude Stone Monuments as materials for history can never be properly appreciated, and controversies of this sort may be indefinitely prolonged without any satisfactory conclusion being arrived at. * Sijborg Samlingar, d'c, toI. i. pi. 12, fig. 40. DATE DUE UNIVERSITY PRODUCTS, INC. #859-5503