[From ManxNoteBook vol ii,1886]

BALLAUGH OLD CHURCH

THE
MANX RUNES

HERE ARE NO REMAINS
OF ANTIQUITY IN THE ISLE
OF MANN EQUAL IN THEIR
INTEREST TO THE ANCIENT
CROSSES WHICH STAND IN
MANY OF THE CHURCH
YARDS. THESE CROSSES ARE
unimpeachable monuments at once of the ancient Celtic civilization of the Island, and of that subsequent Scandinavian occupancy and dominion which lasted for about four hundred years, from the close of the ninth century nearly to the end of the thirteenth.

These Manx monuments are sepulchral crosses, exhibiting two very peculiar types of ornamentation. The inscriptions are in a Norwegian dialed, and written in that singular Scandinavian alphabet which offers one of the most curious problems in the whole history of the art of writing.

The letters of this alphabet are called Runes, and the alphabet bears the name of the Futhork, from the first six Runes, which represent the sounds F, U, TH, O, R, K. In Sweden, Norway, and Denmark, about 2000 runic inscriptions have been found, the earliest being probably as old as the Christian era, while a few are as late as the middle of the 12th century, when the runic writing was finally superseded by the Roman alphabet.

The origin of the Runes has been the subject of prolonged controversy, the most diverse theories having been maintained. The opinion now generally accepted was first put forward in a book published in 1879, entitled "Greeks and Goths ; a Study on the Runes," in which the author of the present paper maintained that the runic writing was derived from a variety of the Hellenic alphabet used in the Greek colonies on the Black Sea about the 6th century B.C. It would seem that the Goths, who then occupied the region between the southern coast of the Baltic and the upper waters of the Dnieper, must have obtained a knowledge of the art of writing from the merchants of Olbia and other Greek colonies on the Euxine, who, according to Herodotus, voyaged forty days' journey to the North by the great trade route of the Dnieper. At some subsequent but very early period, certainly before the time of Ptolemy, a portion of the Gothic people migrated northward across the Baltic, where they have left their name on the Swedish province of Gothland. In the 2nd century A.D. other Gothic hordes marched southward and invaded the Eastern empire, reaching the Danube in the reign of Caracalla- The language of the Scandinavian and the Danubian Goths was the same. The inscription on the Rune stone in Norway, which is assigned to the 3rd century A.D., is pure Gothic, such as was used by Ulfilas in the 4th century for a translation of the Bible, made for the benefit of the Goths then settled on the Danube." The runic writing must also have been the common property of the undivided Gothic nation before these migrations began, since inscriptions in the ancient runes have been found not only in Scandinavia, but on the Dnieper and the Danube, and as far west as the Saone, along the track of the Gothic and Burgundian hosts who ravaged the Mediterranean lands, and finally founded kingdoms in Italy, Gaul, and Spain.

For the somewhat technical discussion by which these early Gothic runes are identified with corresponding letters of the ancient Greek alphabet, the reader must be referred to the work above cited, in which the arguments are set forth in detail.

Among the southern Goths, who came into contact with Greek and Latin civilization, the runic writing speedily died out, but in the North, where there was no competing alphabet, it survived for many centuries. It was employed by the Swedes, the Norwegians, the Danes, the Jutes, and the Angles, but seems to have been unknown to the Franks, the Saxons, the Lombards, and to all the purely German tribes. Hence no runic stones have been discovered in Germany, or France, or in those parts of England which were conquered by the Saxons; whereas runic inscriptions have been found in East Kent, which was conquered by the Jutes; in Lincolnshire, Yorkshire, and Durham, which were colonized by the Angles; and also in the Orkneys, the Hebrides, Cumberland, and the Isle of Mann, which were occupied, at a later period, by Norwegian settlers. The number, values, and forms of the Runes underwent considerable modifications as time went on, so that the runic inscriptions of different epochs can be clearly distinguished from one another. Thus the Kentish inscriptions are in Gothic Runes of the 5th or 6th century, those of Mercia and Northumbria in Anglian Runes dating from the 7th century onwards, while those of Cumberland, the Orkneys, and the Isle of Mann, are in Norwegian Runes of the 10th and following centuries. Thus the three series of runic inscriptions in our Island are to be referred to the three great epochs of Scandinavian conquest.

The Isle of Mann has yielded no less than thirty-eight crosses, or fragments of crosses, which may be attributed to the Scandinavian invaders, and of these nineteen bear inscriptions in runic characters. The object of the present paper is to investigate their origin and their date, and to arrange them, as far as possible, in chronological sequence.

In attempting this task certain preliminary limiting considerations must be borne in mind. Since the stones are sepulchral crosses, they manifestly belong to the Christian period. The inscriptions are in a Scandinavian dialect and alphabet. The names are usually Norse, but occasionally Celtic. The elaborate ornament is mostly Irish in its character, but, in some cases, it has been modified by Scandinavian influence. Two runic futhorks, one earlier than the other, are employed. Hence the conclusions may be drawn that the monuments extend over a considerable period, during a portion of which time Celtic influences prevailed ; that the earliest stones are not only later than the immigration of the Scandinavians, but also subsequent to their conversion ; while, on the other hand, they must be earlier than the time when the Norse language died out and the Runes fell into disuse.

Let us now examine to what extent these considerations can be brought into connection with the facts of Manx history, so far as they are known.

The Manxmen belonged to the Goldelic branch of the Celtic race, and spoke a language closely akin to the ancient Irish. They had been converted probably in the 5th century, certainly not later than the 6th, by the colleagues or the successors of St. Patrick. The round tower, near Peel, belongs to the same type as the round towers of Ireland, and must have been erected under Irish influence. Taken in conjunction with the name of St. Patrick's Isle, on which it stands, it goes far to establish the Irish origin of the ancient Christianity and civilization of the Manxmen. About the close of the 9th century many Norwegian chiefs, still heathens, fled with their dependents from the iron rule of Harald Fairhair, and established themselves in Iceland, Ireland, the Hebrides, and the Isle of Mann. This emigration began after Harald's subjugation of Norway in 883 A.D., and probably continued for about forty years. The conversion of Norway, attempted during the latter half of the 10th century, was effected in the reign of Olaf the Saint, 1015-1030, A. D. The conversion of the Northmen in the Isle of Mann seems to have followed closely on that of the mother country. The early bishops of Mann bear Celtic names, and probably ministered only to the Celtic population. There seems to have been a break or dislocation in the episcopal succession shortly before the time of Roolwer (Hrafr), who is the first bishop who bears a Scandinavian name. The list of bishops in the Chronicle of Mann, which was compiled by the Monks of Rushen Abbey, begins with him, and says he was the first bishop before the time when Godred Crowan began to reign, that is, before 1080 A.D. The chronicler adds the significant statement that , there were many bishops from the time of the blessed St. Patrick," yet it is sufficient to have begun the account of the bishops from Roolwer", because we are entirely Ignorant who or what (qui vel quales) were the bishops before Roolwer's time; for we neither find any written documents on the subject, nor have we any certain accounts handed down by our elders."* It is plain that the monks of Rushen recognized no regular episcopate before Roolwer's time, and the phrase qui vel quales implies that the so called bishops of earlier date, mentioned in the Irish annals, who all bear Celtic names, were merely Culdee abbots of the Iona type, who exercised their functions only among those of their own race. Hence it seems not unreasonable to suppose that the conversion of the Northmen in Mann, and the organization among them of a regular episcopate of the Roman type, cannot have taken place before the middle of the 11 th century. To any earlier period than this it would, therefore, not be safe to to assign any of the crosses bearing runic inscriptions.

The great expedition of Magnus Barefoot, who set sail from Norway with one hundred and sixty ships, resulted in the conquest of the Island in the year 1098, and was followed by a fresh influx of Norwegian settlers. We are told that he was so much pleased with the beauty of the Island that he chose it for his abode, and that when he started for Ireland on the voyage of exploration in which he was killed, he took with him only sixteen ships. Many of his followers must have been left behind. To this increase of the Scandinavian element in the population we may probably attribute the peace and prosperity which prevailed during the time of Olaf, King of Mann. All the conditions point to his long reign, 1103-1153 A.D., as a period favourable to the erection of some of the more costly crosses on the Island. In 1154, after his death, the connection with Norway was drawn closer by the subjection of the Manx bishopric to the archepiscopal see of Drontheim. The invasion of the Scots in 1228, which reduced a large part of the Island almost to a wilderness,* the subsequent decline of the Norwegian power, and the final cession of the Island to Alexander III. of Scotland in 1266, after the battle of Largs, in which many of the Manx chiefs of Scandinavian lineage perished, are events which point to the first half of the '13th century as the probable limit of the period to which the runic monuments may reasonably be attributed.

On historical grounds, therefore, we may assign the runic inscriptions in the Isle of Mann to the two centuries of Scandinavian Christianity and Norwegian power, from about 1050 to 1250. It would also be reasonable to suppose that the crosses with Celtic names and pure Irish ornament belong to the earlier part of this period, during which the influence of the ancient Celtic art and of the Celtic population continued to be strong; while the crosses with Norse names and Scandinavian ornament may be referred to a later time. after the Scandinavian element had been strengthened by the influx of fresh Norwegian settlers about the year 1100, and by the ecclesiastical union with Norway which took place in 11 154.

We have now to examine how far these historical inferences are borne out by the internal evidence furnished by the monuments themselves, that is, by the artistic quality of the ornament, and by the contents, the dialect, the palaeography of the inscriptions.

We may deal first with the artistic style displayed in the ornamentation of the crosses. A few are of unmistakably Irish character, a few are as markedly Scandinavian, while the rest exhibit a gradual transformation from one style to the other. The essential difference between the two styles cannot be too clearly grasped : the one is essentially geometric, the other zoomorphic. The pure Irish art is exhibited in great perfection in the elaborate ornamentation of the splendid codices belonging to the Irish school of calligraphy, such ~h date from the end of the 7th century or the beginning of the 8th. It is also seen on many sculptured stones in Ireland and Scotland. The ornament is a representation, on parchment or stone, of complex interlacing cords, with elaborate knotwork of marvellous ingenuity and diversity. It is impossible to believe that such patterns could have been invented by scribes or stonemasons ; they must have grown out of something else, and can only be regarded as survivals of art work of an earlier period in some more suitable material, just as the Greek temple was an imitation in stone of a primitive wooden structure, the ends of the roof beams and the wooden pegs by which they were fastened together, leaving as survivals the triglyphs and the guttae, which became purely decorative details. The intricate scroll work of the great codices may be an imitation of the patterns of artistic rush mats which adorned the halls of Irish chiefs and abbots, while the stone crosses of early date, such as that at Kilklespeen in Ireland, shown in the plate on opposite page,"~. unmistakably suggest the supposition that they were merely reproductions in a more durable material of still older crosses constructed of interwoven wickerwork. If, during the Celtic period, the mortuary crosses were merely of basket work, this would account for the fact that they have all perished, leaving as survivals those elaborate imitations in stone which belong, as the runic inscriptions testify, to the Scandinavian period. This style of interlaced geometric ornamentation must certainly have been employed in the booksand very possibly in the stonework of the Celtic church at the time of the Scandinavian conquest, and would furnish the only available artistic models. Hence we must consider those Manx crosses which display pure Irish ornament as the earliest in date. A good example of this pure Irish art is the lovely Nial Lumgun cross at Kirk Michael, figured on the plate opposite page 97, which, as will presently be shewn, may claim to be the oldest of the inscribed crosses in the Island.

The native Scandinavian art, as exhibited on Swedish and Norwegian stones, is of a character wholly different. The ornament is zoomorphic, a style repugnant to the artistic conceptions of the Celtic races. Huge serpentine monsters, terrestrial or marine, engaged in fierce combat, with their writhing bodies Intertwined, are depicted on the Scandinavian stones*, just as the Kraken, the Dragon, the Snake, and the long Serpent?, were carved as figure-heads on the prows of the viking galleys.+ On some of the Manx crosses, such as the Thorlaf cross at Kirk Braddan (figured on plate, p. 97), we see an ornate form of this draconic decoration covering the shaft, while the head of the cross exhibits a survival of the Irish knotwork employed on the cross at Kirk Michael. It is plain that the stones which display the writhing serpents, so characteristic of Scandinavian design, must be referred to a period when the milder Celtic art, with its exquisite geometrical patterns, had been infected by the fiercer ideas congenial to the descendants of viking chiefs. These two characteristic crosses, the Thorlaf stone at Kirk Braddan and the Nial Lumgun stone at Kirk Michael, may be taken to represent the two extreme types-Irish and Scandinavian-to be found on the Island, and must evidently be separated by a considerable period of time. of intermediate date is a series of stones which exhibit a gradual transition from the Celtic to the Scandinavian type of ornament. The Kirk Malew cross' and the great jualfr cross§ at Kirk Michael may be taken as instances. In these, especially the latter, the main and central ornament is Irish, while the subsidiary and superadded ornament is Scandinavian. From a study of these monuments it becomes plain that the Irish cross, with pure geometrical ornament, was the original type, out of which the Scandinavian cross, with zoomorphic ornament, was gradually developed.

it is manifest, therefore, that the Manx stones, inscribed or uninscribed, may be roughly classified in accordance with their artistic character, the pure Irish crosses being the oldest, and the Scandinavian stones the latest in their date.

We next pass to the inscriptions on these stones, In search of any chronological indications which may be yielded either by the forms of the Runes, the dialect, or the contents.

Unfortunately, none of the Manx monuments can, with any certainty, be connected with personages known to history. The names Nial, Grimr, Dugald, and Utr, are, it is true, found on various crosses, and persons bearing these names figure in Manx history in the years 914, 996, and 1098, but these names were so common in the Isle of Mann that it would be unsafe to identify them, or to draw any inference from their appearance, with the possible exception of Utr or Otter, Earl of Mann, who was slain in 1095.

The foregoing conclusions as to the relative age of the crosses are supported by variations in the spelling of certain recurring words, attributable either to dialectic change, or to modifications in the phonetic values of certain Runes. The usual formula in the inscriptions is A.B. raised this cross to C.D.

The crosses with Celtic ornament usually have A.B. raisti krus thana aftir C.D.

On those with Scandinavian ornament we have A.B. risti krus thono aft C.D.

The crosses with mixed ornament vary between these spellings.* Such distinctions are perhaps of no very great value, but, so far as they go, they confirm the argument drawn from the artistic character of the monuments.

A far safer guide than the dialect is the palaeography. Fortunately for our purpose, several runic inscriptions exist in Cumberland and elsewhere, to which dates can be assigned, and hence, by a comparison of the forms of the Runes, approximate dates can be obtained for the Manx records. Especially valuable are the inscriptions from Cumberland,** as it was settled by Norwegian emigrants at the same time and under the same circumstances as the adjacent Isle of Mann. The fact that the Abbey of Furness nominated at one time to the See of Mann shews how close the connection remained in after times.

The monuments, mostly with inscriptions+ of precise date, which may be used for comparison with the Manx Runes, are as follows:

i. A cross at Collingham, in Yorkshire, erected in memory of King Oswin, who was killed 650 A.D.

2. A cross at Bewcastle, in Cumberland, in memory of King Alcfrith, who died 670 A.D.

3. The cross at Ruthwell, near Dumfries, containing a part of

Caedmon's Dream of the Holy Rood, about 680 A.D.

4. Fragment of a cross at Leeds, in memory of King Onlaf, 872 A.D.

5. Inscription of King Harald, in memory of his father, Gorm the Old, who died in 935 A.D.

6. Inscriptions of King Sweyn, about 1000 A.D.

7. Inscription in Carlisle Cathedral by Tolfin, who was probably the Tolfin dispossessed by William Rufus, after his conquest of Carlisle, 1092 A.D.

8. Inscription at Bridekirk, in Cumberland, on a font probably carved by Bishop Pudsey's architect, 1,50-'170 A.D.

9. Inscription on a rock at Barnspike, in Cumberland, recording the death of Gilhes Bueth, 1160-1170 A.D.

10. Runes of King Waldemar, C. 1240 A.D,, preserved in a 1\1S., written about 1290.

11 Two inscriptions of ownership, in Manx Runes, on the broach of a warrior killed at the battle of Largs, 1263 A.D.

12. Inscription in Manx Runes, on the sword-belt of a warrior buried at Greenmount, near Dundalk, Co. Louth.

13. inscription in Manx Runes, by a hermit who lived in St. Molio's Cave, Holy Island, Arran.

The inscription on the cross at Beweastle, in Cumberland, is however in the earlier Anglian Runes, which are used also on the Yorkshire crosses and on the cross at Ruthwell.

14. Inscriptions of various dates, some in Manx Runes, at Maeshowe, Orkney.

15. A very early Christian inscription, relating to the payment of tithes, on a ring attached to the church door at Forsa, in Sweden, c. 1150-1200 A.D.

16. A very early Christian inscription on a bell at Holmen, in Norway, c. 1150-1250 A.D.

A superficial examination of these inscriptions will suffice to show that all the Manx records are decidedly later than the 7th century Runes at Collingham, Ruthwell, and Bewcastle, and also somewhat later than the 10th century Runes of Harald and Sweyn ; that they are nearly of the same age as the 11th and 12th century inscriptions of Cumberland, but older than the 13th century Runes of Waldemar. From a palaeographical point of view, the greater number can be assigned to the 12th century, though some may have been written at the end of the 11th, or the beginning of the 13th, a result which agrees with the historical limitations already obtained.

When we go into greater detail, it will be observed that on the Manx stones two runic alphabets, or futhorks, are used, one of which must be somewhat earlier than the other. It is important to determine the relative priority of these Futhorks, as one of them, which we will call Futhork A, is only used on the Nial Lumgun cross at Kirk Michael,* which has Celtic names and Irish ornament; and the other, which we will call Futhork B, is employed on the Thorlaf cross at Kirk Braddan, which has Norse names and Scandinavian ornament. Mr. Cumming, following Prof. Munch, calls Futhork A the new Runes, and Futhork B the old Runes. If this is correct, it would upset all our previous conclusions drawn from the artistic style of the ornament, and we should have to believe that the crosses of pure Scandinavian type, with Norse names, are earlier than those with Celtic names and Irish ornament.

The question of the relative priority of these Futhorks is therefore of some importance, as, incidentally, it settles the question of the chronological order in which the stones must be arranged.

The two Futhorks differ mainly in the form of the Runes for a, n, s, and t, and in the value of the Rune for o. We may confine ourselves, as a basis of comparison, to the consonantal Runes, as their phonetic values do not vary. We have in Futhork A, *~ = ill = s, and T = t, while in Futhork B, we have n, S, and t.

Now, If we compare these test forms with those in the dated inscriptions, it is clear that the earlier inscriptions exhibit the A forms, and the later inscriptions the B forms; and also, that the change took place in the course of the two centuries to which the Manx inscriptions must be assigned.

The A forms of all three Runes are found in the following dated inscriptions : Bewcastle, 670 A.D., Ruthwell, 680 A. D., Gorm, 935 A. D., and Sweyn, 1000 A. D. On the other hand, the B forms of all three Runes agree with the forms in the early Christian inscriptions on the Forsa ring and the Holmen bell, 1150-1250 A.D.; with King Waldemar's Runes, 1240 A.D.; and with the Inscriptions in St. Mollo's caveand on the Louth sword belt, which are assigned to the beginning of the 13th century. Hence it is clear that the A forms are older than the B forms. To ascertain the precise time when the change came about, we notice that at Carlisle, 1092 A.D., we find the old s, the new n ' and both forms of t. Bridekirk, 1150-1170 A.D., has the new forms of n and s, and the old form of t. Barnspike, 1160-1170 A.D., has the old s and t, and the new n. Hence it appears that the new forms were coming into use before 1092, while the old forms were not altogether superseded in a remote district in Cumberland in 1170.

Hence we see that the results of the palaeographical investigation are in no way inconsistent with the historical probabilities. The historical evidence leads to the conclusion that none of the crosses are later than the battle of Largs in 1263, or earlier than the conversion of the Northmen in Mann, which took place a few years before 1080 A.D.; limits which are somewhat narrowed by the two great devastations of the Island, by the Scots in 1226, and by Godred Crowan in 1075. Moreover, we gather that the form of the Runes and the style of ornament may have been influenced by the fresh influx of Norwegian settlers after Magnus Barefoot's expedition in 1095, by the tranquility prevailing during Olaf s reign, 1103-1153, and by the closer union with Norway in 1154. On the other hand, a comparative study of the available monuments leads to the conclusion that the oldest Manx inscriptions must be considerably later than the year 1000 A.D., and earlier than 1240; while the change in the Ftithork, which was followed by changes in the dialect, and in the character of the ornaments, mav have taken place as early as 1092, but can hardly have been so late as 1170.

These dates, obtained by independent processes, agree * as closely as could be expected. Although the margin of uncertainty is not great, there are other considerations which point.to results even more definite.

Of all the inscribed crosses on the Island, the oldest is the Nial Lumgun cross at Kirk Michael, as is proved by the Celtic names, the form of the Runes, and the archaic style of the ornament. Referring to the plate on P. 97, it will be observed that the dress of the three figures reaches to the feet.'-'~ There is another cross at Kirk Michael to the memory of Rumun (Hramundr), which must be of somewhat later date, since the names are Scandinavian, the inscription in the later Runes, and the ornaments, though still Irish, decidedly more developed, exhibiting what has been called the pelleted-ribbon pattern, which frequently appears in later crosses. On this cross are two figures dressed in short kilts, not reaching to the knees. Now, as Mr. Cumming has pointed out, Magnus Barefoot, who conquered Mann in 1095, received his nickname from his adoption of the Highland kilt during the Scottish wars, and it seems not unreasonable to suppose that the cross with the earlier Runes, and the figures clad in the long garment, is older than Magnus Barefoot's conquest, and that Hramundr, the Northman, who is represented with a kilt on a cross with later Runes and later ornament, must have lived at a time when the dress worn by the Norwegian king had been adopted by his followers. One of these crosses, therefore, will be somewhat older, and the other somewhat later than the year 1095. They cannot, however, be separated by any very great interval, because, as we will proceed to show, both must have been executed by the same artificer. The proof is not difficult. Between these two we can interpolate two other crosses of intermediate date. These two crosses, both with the new Runes, were, as we learn from the inscriptions upon them, made by Gaut Bjornson. One is the Mall Brigdt cross at Kirk Michael,~',' the other, the Ufaig cross at Kirk Andreas.+ On the first, Gaut states that he "made this, and all in Mann." He must, therefore, have made the Nial Lumgun cross, with the older Runes, which stands in the same churchyard. On both the names are Celtic, and both have Irish ornament, and both, to judge by the execution, may have been the work of the same artist at different periods of his life. The Nial Lumgun cross, with the earlier Runes, is in the archaic Irish panelled style, with the ornament somewhat stiff, and exhibiting obvious survivals of a wicker-work prototype. On the Mall Brigdt cross, the ornament is less cramped, the treatment more free, and a very peculiar chain-work design, repeated on several later crosses, makes its appearance on the front of the cross, while the back has a plaited wicker-work ornament. The Ufaig cross at Kirk Andreas, also, according to its inscription made by Gaut, has the plaited ornament on the front, the later chainwork ornament not being introduced. Hence we infer that the Kirk Andreas cross was made before Gaut had invented the ingenious chainwork ornament found on the Mall Brigdi and other crosses. Themorearchaic ornament on the face of the one is relegated to the back of the other. Hence we conclude that the Ufaig cross at Kirk Andreas is the older of the two, and must be one of the crosses of which Gaut claims the authorship in the inscription on the other. We have, therefore, four crosses, all by the same artificer, which can be arranged in chronological order:

1. The Nial Lumgun cross, at Kirk Michael, with the old Runes and the long garments.

2. The Ufaig cross, at Kirk Andreas, with the new Runes; claimed by Gaut.

3. The Mail Bridgi cross, at Kirk Michael, with the new Runes, the chain work ornament, and the rope ornament ; claimed by Gaut.

4. The Rumun cross, at Kirk Michael, with the new Runes, the kilt, the chain work ornament, a rope ornament,* and the pelleted-ribbon ornament.

The first of these may be assigned to the end of the 11th century, and the others to the early part of the reign of Olaf 1103-1153. To the same reign we may also assign some other crosses, in Gaut's style, such as the Tynwaldt cross, a fragment at Kirk Michael, both early, with plain chain work, and the later Ballaugh cross,§ which has the chain work ornament, both plain and pelleted, and the rope ornament pelleted; as well as the Thurketel cross at Kirk Braddan,~-* and an uninscribed cross at Kirk Maughold+t, both early.

All these crosses are distinguished by their pure geometrical patterns, and the absence of the zoomorphic ornament which characterizes native Scandinavian art.

This new style appears in a rudimentary form on the Kirk Malew cross,'~ the two Conchan crosses, § § and, somewhat more developed, on an uninscribed cross at Kirk Michael.*~'-* On these Gaut's peculiar chain work has disappeared, and the interlaced rope work is less ingenious and more conventional.

The latest of the inscribed crosses, on which the Scandinavian style is most fully developed, are two at Kirk Braddan,"' one of which is figured on plate, p. 97. These crosses are manifestly separated by a considerable interval, perhaps as much as a century, from those of Gaut. Both of the Braddan crosses are plainly from the hand of the same artist, who, as we learn from the inscription on one of them, was named Thorblarn (ThorDjo5rn). The Irish ornament has now nearly disappeared, being replaced by zoomorphic designs of more artistic style than is found in Scandinavia itself. The execution is free, and instead of a somewhat clumsy cross being sculptured in relief on an oblong slab, the stone is itself shaped into a graceful cruciform outline. These crosses, which bear Scandinavian names, Thurlabr (Thorlafr), Eabr (jabr), Utr (Otter), and Froga, may perhaps be as late as the beginning of the 13th century. Somewhat earlier, but distinctly later than Gaut's time, is the magnificent jualfr cross+ at Kirk Michael, frequently called 11 The Dragon Cross," from the two dragons at the base. On this cross, the two styles of ornament, Celtic and Scandinavian, are associated, rather than blended as on the two Braddan crosses. The inscribed crosses may thus be distributed over rather more than a hundred years, from the end of the i i th to the beginning of the 13th century ; all but three belonging to the 12th, and no less than ten having been executed during the reign of Olaf 1103 -1153.

Within the limited space at my command, it has been impossible to deal with more than one of the many problems connected with the Manx crosses. I have selected for discussion the question of their date, because on this point I find myself in conflict with received opinions. The results at which I have arrived were to me wholly unexpected. When I took up the subject, I imagined that I should have merely to repeat and confirm the views of the eminent antiquaries who have preceded me in this inquiry. But an independent examination of the evidence has led me, perforce, to different and even opposite conclusions. The earlier crosses were assigned by Professor Munch to the 9th century, and by Mr. Cumming to the middle of the 10th, while Professor Worsaae considered they were all erected before 1077, with the exception of the Nial Lumgun stone, which he thought was later. I have, however, found myself compelled to relegate them all to the 12th and 13th centuries except the Nial Lumgun cross, which may, I think, be placed in the last years of the 11th. My predecessors consider the stones with Scandinavian names and ornament earlier than those with Celtic names and Celtic ornament ; I hold that the reverse must have been the case: and I differ from them as to the relative priority of the two Futhorks which are used. I cannot, however, expect such novel conclusions to be accepted without controversy, which however, if conducted in the right spirit, cannot fail to bring us nearer to the truth.

ISAAC TAYLOR.

* Numerous examples, some belonging to the heathen, others to the early Christian period, are figured in Stephens' " Runic Monuments," Vol. II., pp. 635-740.

' See " Corpus Poeticum Boreale," VOI. I., P. ~573.

"Chronicon Manniae," p. 14.
" Chronicon Manniae," p. 90.
From French, "Ancient Sculptured 'Stones," (Manchester, 1858.)
See " Corpus Poeticum Boreale," Vol. II., P. 217.
Cumming's " Runic Remains of the Isle of Mann,'" plate v, fig i,~. ~
Ibid. plate iv., fig 13.

* Three of the four Irish crosses read raisti, on the fourth the word does not Occur. The three Scandinavian crosses have aft; two of the four Irish crosses have aiftir or aftir. Three of the four Irisb crosses have the Rune 1 for the vowel in thana; the three Scandinavian crosses use the Rune which originally denoted but was afterwards softened into ~. of the crosses of intermediate style, those which incline most to the Irish ornamentincline also to the earlier spelling, andvice versa.

+ For accounts of these inscriptions see Stephens' great work on Runic Monuments.

Another inscription in these Runes, also from Kirk Michael, does not seem to have belonged to a mortuary cross. It reads

[1S]ITRA : 18 : I,AIFA : FU.S'I RA : GUTHAN. : SUN : ILAN

Better is it to have a good foster son than an ill son

The stone is believed to be lost, but there is a cast of it in the possession of Sir Henry Dryden.

* The older forms probably maintained themselves longer in Cumberland than in the Isle of Mann, which was in closer connection, political and ecclesiastical, with Norway, where the new forms seem to hive arisen.

Cumming, "Runic Remains," plate xi, fig. 28.
Ibid. plate ii, fig. 3.
Cumming, " Runic Remains," plate i, fig. i.
Ibid. plate iii, fig. io.

* This rope ornament, on the right hand side at the back of the cross, is absolutely, identical with the rope ornament on the right hand side of the front of the Niail Bridgi cross, and proves that Nos. 3 and 4 were made by the same person. The pelleted-ribbon pattern and the rope ornament are seen in the initial letter at the beginning of this paper, and the chain-work ornament on the cover of the number.

1 Cumming. " Runic Remains," plate ii, fig 5.
Ibid. plate ii, fig 4.
Ibid. plate i, fig 2.
Ibid. " Recently- discovered Crosses," No. i. ++
Ibid. No. 2.
Ibid. " Runic Remains," plate v, fig 15.
Ibid. plate vi, figs. 17 and 18. ***
Ibid. plate vi, fig. 19.

* Cumming, " Runic Remains," plate viii, figs. 22 and 23.
+ Cumming, " Runic Remains," plate iv, fig. 13.