[From Manx Note Book, vol iii, 1887]

THE FOLLOWING CORRESPONDENCE ON DR. VIGFUSSONS READINGS OF THE MANX RUNIC INSCRIPTIONS IN THE LAST NUMBER OF "THE MANX NOTE BOOK," HAS APPEARED IN "THE ACADEMY": THE READERS OF "THE NOTE BOOK" WILL BE GLAD TO HAVE IT IN A COMPACT FORM FOR REFERENCE:—

Settrington, Feb. 8, 1887.

I have just received the new part of The Manx Note Book. It contains the long expected article, by Dr. Vigfusson, of Oxford, entitled "The Manx Runic Inscriptions Re-read," in which we were led to expect that the errors in Munch’s and Cumming’s transliterations and translations would be corrected. I must confess I am much disappointed with the results of these re-readings. Dr. Vigfusson spent some time in the Island last autumn, and personally examined the stones. He was also offered the use of the casts taken at Sir Henry Dryden’s expense in 1841, when the inscriptions were in a much more perfect state than they are now, after forty-five years’ additional exposure to the weather; but he declined to avail himself of the offer, on the ground that the casts "would only raise controversy." Sir Henry Dryden has most kindly placed at my disposal a set of squeezes from his casts, and I am bound to say that in numerous instances they establish the correctness of the old readings given by Munch and Cumming. In one or two cases entire words have now disappeared from the stones which are perfectly legible on the casts, and in numerous instances the casts plainly show letters which Dr. Vigfusson has either misread or found to be illegible. I am now collating the whole of the inscriptions, which is a work of time, as it can only be done in a good light; but I hope shortly to send the results to the Academy. They will, I think, show that Dr. Vigfusson would have acted wisely in availing himself of every possible means of verifying the true readings of these most interesting records.

ISAAC TAYLOR.

Oxford: Monday—Valentine, 1887.

Canon Isaac Taylor blames me for having read the Manx runic crosses from the stones, and not from casts in his hands—an extraordinary thing, to find fault with me for having read from originals and not from copies. The casts he speaks of I have never seen, and I know them only at secondhand and by chance, for Canon Taylor has never written a word to me about them; and now, here they are, a deus ex machina, not to solve a knot, but to raise dust and confusion.

With equal justice, and methinks with better grace, Canon Taylor should have found fault with himself for having, on the sole authority of casts, and without consulting the stones, still extant, legible, and within reach, made assertions and given readings and started theories as to relative age of inscriptions, which neither he himself nor anybody who understands the subject could have ever dreamt of, had he, if but for a few minutes, looked at the monuments themselves on the spot.

Casts are at best, even if genuine, but copies; and a copy can never take rank with an original. It is idle to plead that when the casts were taken the crosses were in better preservation than they are now. The early iconoclastic days of these poor crosses lie generations, nay centuries, behind all casts or rubbings. The crosses are some seven centuries old, and the casts only some thirty years—a short space of time indeed. As to the tread of feet that have worn the edges of the Kirk Bride cross, one sees that those feet that trod thereon—-generations of feet one fancies—had long since returned to dust ere any casts were known in this antiquarian world of ours. In one or two instances—the runes being written on slate —it is just possible that a bit may, since the casts were taken, have been knocked off. In that case the casts would come in for consideration, granting that they are shown to be tolerably accurate in those parts that we can still check, otherwise not.

Further, I know from my own experience that vellums (a much tenderer material than stone), which more than two centuries ago were ill-read or thought to be past reading altogether, and were left blank by scribes, are now found to be quite legible. Some such I have read myself. Keen attention, and taking pains, work wonders here as elsewhere, and I never used chemicals, except pure water. Mr. Savage and I tried the same on the stones, and we found it of good use in a few cases, where the runes were faded, like dim spots on vellum. We damped the weak spots gently with a wet sponge or rag, and this brought the letter or letters clearly and legibly forth.

But I have little time to spare, and no stomach for a runic war— litigiis tempus male teritur. I fear, were I to plead my eyes versus Canon Taylor’s casts, I should play the sorry figure of M. Orgon arguing with Mdme. Pernelle. I went to the Isle of Mann with no intent to tread on the toes of runologists, but in quest of pure air, and for the sight of the sea, which every day I miss here in this lacksun, lack-sea inland nook of mine. The Tinwall, too, has long been a day-dream of mine. It was my good fortune to fall in with Mr. Savage; and he, a practical patriotic man, put me on the runic task. We did the work in common, in all essentials accurately I trust, certainly, thinking ill of nobody, making no fuss about it whatever. In fact, it ~s a world of pity it has not been done long since.

GUDBBAND VIGFUSSON.

Isle of Mann, Feb. 9, 1887.

Permit me, as a Manxman and a student of our many interesting antiquities, to express the thanks of myself and fellow-countrymen, which will be shared I feel sure by all antiquaries and scholars, for Dr. Vigfusson’s valuable contribution on the subject of our inscriptions which appeared in The Manx Note Book, No. 9, January, 1887.

Though I could not presume to set my opinion against that of so well-known a runic scholar in a question of the grammar or translation, I may fairly claim to do so in the reading of the runic characters of our inscriptions. I have lately devoted considerable time and attention to the subject, and, being a resident in the Island, I have many opportunities of examining the stones themselves. While heartily thanking Dr. Vigfusson, therefore, for his readings, I cannot allow them to pass unchallenged. The fact of his being so high an authority on the subject, and the certainty, therefore, that his readings, taken on the spot, will be generally accepted as correct unless the contrary be shown—the fact also that he emphasises his own certainty of their correctness as "the result of careful, and in some cases repeated examination," renders it obligatory on me to point out at once the most important of what, with all respect to him, I believe to be the mistakes he has inadvertently made.

It will be most convenient to consider the inscriptions in the order given by Dr. Vigfusson. (1) Respecting his first, which he calls the "Mael-Lomchon Cross, Kirk Michael," I have to say that the third word is distinctly RAISTI not nisri, and I read FUSTRA : SINE : TOTIR, not FOSTRA : SINA : TOTER, and KONA : AS : ATHISL : ATI, not KONO : ES : ATHISL : ATTI. These runes are all distinctly visible as I have given them, except the first of the word AS, which is broken across the middle in the line of what would have been the A-stroke. If it had been E, the central dot would still have been visible, as it is always deeper than the line of the stem, which in this case can be faintly traced past the centre, The last rune of the word SINE is undoubtedly our usual stung-rune for E; but just below its central dot, at its left side, is the lower dot of the N-twig, the stroke of which is obliterated, and, above it at its right side is the upper of the two dots which divide this word from the next; at first sight, therefore, it has quite the appearance of A, with the central stroke obliterated,

As regards the translation and the explanations given of the motto, I would ask why not render FUSTRA foster-daughter instead of "foster-mother"? We might then suppose that the "motto" was applied by Mael Lomchon to Athisl, the husband of his foster-daughter. Dugald, perhaps, had died leaving Mal Mora and a brother of hers, who proved a bad son. Mael Lomchon, who had no son of his own, adopted Mal Mora, and looked upon the husband to whom he married her as his son. When Mal Mora died, he raised this cross to her memory, and commemorated the fact that her husband had been a good man, who he was glad to think should succeed himself, for it was "better to leave a good foster-son than a bad son "—like that of Dugald; or, it might be that Mael Lomchon himself had had a son who proved unworthy and left him. This, 1 submit, while being an equally correct translation, is a more rational explanation of the motto. I should like further to ask why there "is little doubt that the B was not the peculiar Manx letter, but the ordinary Scandinavian B"? As to the legends being now known for the first time in their proper connexion, I would point out that they are so given in Gibson’s and in Gough’s edition of Camden.

(2) The Kirk Bride Cross has not before been published, except for the reading I hazarded about two and a half years ago, when I first began to pay attention on the subject. I took the first rune to be I, reading IRUIAN, but it may have been T, the top now obliterated I could distinguish the s of KRS, and AFCATHMIUL I took to be AFTETHM IRI. If Vigfusson’s reading be correct, we have here the word AFTIR contracted into AF, as it appears to be in the Ulfaac Cross, Andreas—a reading which, I believe, is not met with elsewhere in runes.

(3) The Malbricti-Gout Cross, Michael, is as here given, except that there is a very distinct and well-formed I in the first word; and, as to the spelling of GOUT, I have to say that the second rune is precisely the same as the second rune of the first word, and is neither the usual Manx O nor A, but is intermediate between our two forms of A, having the twig on the right side only of the stem. Instead of two dots there is a cross-mark after "Gaut," and there are no dots after AUC—these, however, are mere details. As to the rendering of BRUKUIN, Prof. Stephens seems to have fairly worked it out in his notice of this inscription (O. N Run. Mon. IL, 597), and his reading has the advantage that it does not require us to suppose that any of the letters "have become transposed." I would ask, Why suppose the cross to have been erected on the death of Malbrikti? The very reference to "Absalom’s pillar" would rather suggest that it might have been set up during his life-time.

In (4) the "Ufaac-Gaut Cross, Kirk Andreas," I much object to the word AFTIR. The first remaining word of the inscription, though broken, is distinctly THANA; this is followed by the usual division of two dots, and my only doubt about the next word, which is even more broken than the first, was whether it had been AF or AFT; the F-strokes are unmistakable. As to SAUNR, I expect one of the two dividing dots before this word was mistaken for S, and the S itself for A. With respect to the last two words, the reading suggested by Prof. G. Stephens, to whom I sent a photograph of the inscription, is KUB : KULS, "Seal-pate," a nickname. The first of these words is worn almost past recognition; but Mr. Cumming also read it KUB, and in his time — thirty to forty years ago —i t would certainly be more distinct than now. On the other hand, I must say that Kinnebrook, who could not read the runes, but was a painstaking artist, in his etching, published 1841, makes the strokes look not unlike this reading FRA, and my own note, taken two years ago, gives the first rune as F. The last rune of the next word might be I or S, it is now too broken to say positively. It does seem to me unreasonable and farfetched to connect the last word with "Cooley" in Kirk Michael—for supposing the reading FRA CULI "of Culi" to be right, there are places bearing such a name nearer to the cross, e.g., Cooildhoo, i.e., "the dark grove," in Jurby; Cooilbane, Lezayre; Ballacooiley, Ballaugh, and so on.

With respect to (5) "The Ufaac Cross at Kirk Braddan," [MA]LFIAAC may be right, but I read E for I. As to the name Ufaac, I have Prof. Stephens’s authority for saying it was a Norse name, "once not uncommon, now extinct"; and Prof. Munch says of it "Ufeigr is a very common Norwegian name," (Manx Society, Vol. xxii., 28.)

With regard to (6) the Ballaugh Cross, I desire to say that I think Prof. Vigfusson has at last settled the question of the two names which have been so variously read. There is a distance of an inch to 1³/₈ inches between the first o and the beginning of the sunken space in which the inscription is carved, but the next rune being only ⁵/₈ inches from the o. This led me to think Cumming had been right in reading TH as the first letter, though it is not now visible. I suppose it had worn away. The chief difficulty was with regard to the U, which certainly was difficult to believe was meant for an E. By adopting the present reading OULAIBR, Olaf—and, as I cannot see any part of TH, I have no hesitation in following it—this difficulty is surmounted. Of the next word I am not so sure. I took the first rune to be TH (þ), reading THIUTULB, because what Dr. Vigfusson evidently considered the right-hand twig of L is compared with the following rune so low down, while there undoubtedly is a complete bow as of þ. The lower part of the bow, however, may be accidental and simply caused by the stone flaking, for I find it is not so deep nor so distinct as the upper portion or L-stroke. The size of the runes frequently differs; so that in this word also I think the present reading may be correct. As for the blank space following the RA of the third word, I can plainly discern I and the dot of S, and the stems of TI K the US should be ES, the R being as distinct as that in RAISTI. The second rune in THANA is distinctly A, the left twig being visible, though it bears also a twig from the very top of the right side, as in L. The next word is AIFTIR, not AFTIR; and, as to the next, the L is most certainly not to be seen "on the flat circular band." A portion of the deep incision bordering the band, which is much worn, has evidently been mistaken for I, which, if anywhere, is immediately below the band. It is a mistake also to say UN of SUN "are written in one": they are distinct, and there is plenty of space. Lastly, the three dots only occur after the first word. These may appear trifling details; but they serve to show that though the learned professor has had the advantage of "careful, and in some cases repeated, examination of the inscriptions on the spot," he is liable to error as others who have gone before him.

(7) The famous Michael Cross has been curiously misread as to the first word. AULAFIR is most certainly wrong. Even if the first rune, I might be supposed to have had the A-stroke obliterated, there can be no doubt whatever that the third is A, and the fourth I, the twig being far more distinct than that of the TIN RISTI. Prof. Munch’s rendering of the word, Jóalfr, is doubtless correct. The eighth word is distinctly THONO not THANA, and the last two words MUTHUR SINO not MOTHUR SINA.

Of (9) Dr. Vigfusson writes:

"As different readings have been given of the beginning of the existing part of the legend, and the runes are a little damaged, it is better to add that there is no doubt that RTI is the proper reading."

As some time ago I picked out the mortar which had filled the runes of the first word, it was more easy for him to read them than for Cumming and others; and yet he is further astray than they. To judge by the patterns on the faces of the cross, there would not be room for the words [GRIMR : INS : SUA]RTI as suggested; but, apart from this, the RTI of Dr. Vigfusson is beyond all doubt—notwithstanding the positiveness of his assertion—RIM, both twigs of the M being distinct, while what he mistook for the twig of a T is simply a break in the mortar above the top of the letter! Besides this a portion of the K can be traced, though not so distinctly as in the foregoing. I am supported in this reading by Prof. Stephens, who, on receipt of the photograph which I sent him, wrote to me on February 15, 1886, "I think with you the first word was KRIM and follow in the rest." THAN should be THNA. The runes on the smaller piece which I thought might be S]IN, Prof. Stephens suggested might stand for U] LNHULAN (true, faithful); between RUMUNT and this we may suppose "his friend" or some such words.

Of (10) I need only say that, as in the foregoing inscription, Dr. Vigfusson has misread THNA for THAN. The side -strokes or twigs of A and N are both faint but perfectly distinct.

With respect to (17) the Braddan Cross, I cannot see that the inscription "is perfect at the beginning." There is room for the two runes which would make UTR into GAUTR. The rest is now correctly given; but I must take exception to the last sentences— "The continuation given by Mr. Cumming is not now to be deciphered on the cross. In fact the strokes there seem to be almost aimless; and, if they ever meant anything, we can see from the existing marks that they certainly were not what he gives."

What are these marks? After the word FROKA, is the lower of the usual dividing dots, followed by seven strokes, an interval, two strokes, an interval, two more strokes, another interval, and three strokes. Cumining, about forty years ago, read "FATHUR SIN IN THURBIAURN SUNR." Before supposing that he had simply drawn on a vivid imagination for these unnecessary words, we should be quite certain that the existing marks could be no part of them. I have before me a rubbing which I took on November 4, 1885. On it the strokes are distinct, and I have no difficulty in showing by dotted lines that the upper portions of them would and must have been exactly as given by Cumming, excepting only that the stone is now entirely broken off at the third rune of THURBIAURN. The twig of the A and lowerpart of the bow of TH in FATHUR, the lower dots between SIN, IN, and the next word, and the lower part of the bow of TM in the last word, are quite legible on the stone.

As to the Conchan Cross, I need only say here, that, with Prof. Stephens’s notes before me, I cannot accept Dr. Vigfusson’ s reading of this difficult inscription. It is not bi-lingual in any seiise, and the names are good Norse. I think the sixth rune in MURKIBLU is not O but B. IESUCRIST and CRIST are perhaps misprints for ISUKRIST and KRUS, which words are quite distinct. RUNA has neither I nor R at the end of it, and it is followed by TM and the stem of another rune, the three omitted letters are not illegible, and the whole inscription can be read so as to make good sense.

In all the others, except about two, I observe what I take to be slight misreadings. But already this letter is too long, and I must reserve further remarks for some other occasion. I am sorry to have had to differ from Dr. Vigfusson in many particulars, and cheerfully acknowledge our great indebtedness to him for much that is valuable in his interesting article. The fact of such an expert having fallen into error only shows how difficult our inscriptions are to decipher. It confirms me also in the opinion that antiquaries and scholars, who may be unfortunately unable to see the actual crosses for themselves, will only be satisfied of the correctness of the readings when they see them reproduced without the possibility of error. This I am in hopes of being able before very long to do in a work illustrated by one of the new photographic processes, and having a separate large-sized plate for each face and inscription of every cross. The possibility of mistaken readings, or even of artist’s or engraver’s errors, will be thus avoided.

Meantime I am endeavouring to have what still remains to us of our beautiful crosses preserved from injury and weather by being placed under cover; and I trust that the interest taken in them by men of learning and culture may influence those of us who dwell among these noble relics of the past to exert ourselves in a matter wherein we shall have the hearty good wishes and support of scholars, antiquaries, and men of understanding all the world over.

P. M. C. KERMODE.

Settrington, Feb. 21, 1887.

If Dr. Vigfusson will read my letter more carefully, he will see that I did not blame him for reading the inscriptions from the stones, but for declining to verify his readings by means of the casts.

I have now collated one half of the inscriptions. The casts show forty-five runes which Dr. Vigfusson has either omitted, inserted, or misread. They bear him out in eight of the new readings which he has proposed, and suggest seven other new readings which have escaped him. One of these is important, as it supplies the lost rune hagl, which has not hitherto been detected on the Manx stones.

Clearly, Dr. Vigfusson would have acted wisely if he had accepted the offer of the use of the casts.

ISAAC TAYLOR.

Oxford, Feb. 26, 1887.

Mr. Kermode’s letter on the Manx runes might with advantage have been a great deal shorter. Anyone that means to read inscriptions and comment thereon has first to learn the language they are written in. Mr. Kermode seems not to have served that apprenticeship. Read and learn before you write was the good old rule. Alas, how we have changed all that!

On the one side of Kirk Ballaugh (Mr. Kermode’s home), within three quarters of an hour’s walk is Kirk Michael. The chief cross and the legend thereon I found on September 30, immured "waist-deep" in the stones and mortar, the fag end of the one line just jutting out, as if stretching out its hands to the passers-by and crying, "Read me, relieve me, here I am, bound, hidden away, like a spirit in a bottle." The "Nial " -Lumcun anagram was hitherto the only answer until I came. On the other side, seaward of Kirk Ballaugh (I walked down in thirty-two minutes), is the old kirk of Ballaugh. Here, on October 5 I found the cross half sunk into the earth—the first letters below the turf, untouched, for long years past. A labourer’s spade (in a labourer’s hand, not in mine) soon dug it out, and we laid it flat on the ground. Here stretched on grass I lay down and read it. It took us just an hour—and here behold, O gentle reader, all three names were grievously misread, not one name read aright! Why, here was a pretty piece of work for a local Norse runic antiquary, if there was any; but there was none, and it was left to me—Kirk Michael, Ballaugh, and all. Mr. Kermode’s runic and rune studies are of a modern date, subsequent to, not preceding, my stay in the Isle of Mann.

The great and serious discrepancies between the legends on the crosses as I saw them, and the casts or copies whence the former received readings have been drawn, were a wonder and surprise to me. But, unless a spell of glamour was cast upon my sense, what could I do but trust my own eyes? There must be something wrong or rotten in these casts, thought I, for how else, in goodness’ name, could learned sensible men like Munch and Canon Taylor hit on their false readings; here false in the face of stone and slate, no mistake about it. Either their wits or their copies must needs be at fault; for that they had never seen the originals was self-evident. I have a shrewd suspicion that these casts were somehow meddled with by some one, who knew little Danish and less Old Norse. Such readings as cuinu for cunu of the stone are suggestive. Whence the spurious intercolated i—one of Canon Taylor’s missing letters, I should not wonder. But why speculate on dubious casts and copies when you have the originals still legible at hand?

These Manx runes are an odd queer sort of writing. It must be hard for an English reader to realise an alphabet where f can be read c, u as v, and a as b. Yet so it is. On one of the Kirk Andreas crosses some reader of thirty years ago, who clearly knew not Norse, has, with some little show of reason as far as "dustus literarum" is concerned, read cub, where I now, with full certainty and without fail, or doubt, or hesitation, read the well-known preposition fra. Cub-culi yields no meaning whatever; could not be construed into anything Norse; could not mean "seal-pate" any more than "silly-pate’ ‘—in fact, conveys no sense. Runes ought henceforth, in our enlightened age, to cease to be privileged above other alphabets. Read aright, a runic legend should always convey, grammatically, some substantial sense, or else should not be thought worth reading.

Respecting Mr. Kermode’s last sentence, I may add that there is already in the Isle of Mann a body of trustees, constituted by an act of Tynwald, for the preservation and protection of ancient monuments; and they are, I know, mindful of the crosses, and are working to get them preserved from wanton mischief and from weather. I should be glad to see the act extended against wanton readings and dilettanti runologists and antiquaries. Mr. Savage, the vicar of St. Thomas’, is the secretary; he and Mr. Moore could give fuller information on this head. My advice to my Manx friend and critic would be to join his countrymen; for among Manx people, as elsewhere, the old saying holds good, that vires unitae forties agent.

On my runic round in the Isle of Mann I was a sort of Man-in-the moon. My constant attributes were a pail of water and a clout; and now the Nemesis are upon me, for here I am, in consequence, doing weekly penance in scrubbing critics and losing my time.

GUDBRAND VIGFUSSON.

Settrington, March 7, 1887.

Mr. Kermode’s excellent letter has forestalled a great part of my promised criticisms. Dr. Vigfusson has appealed to the stones as against the casts, and by the stones he has been judged; and I may add that, with hardly any exception of real importance, the evidence of the casts supports Mr. Kermode’s reading of the stones.

At the time when the casts were made many of the stones had only recently been exhumed, or extracted, from walls and foundations into which they had been built. If they had been exposed to the weather for six or seven centuries many of the inscriptions would have become ifiegible; as it is, the casts prove that much of the damage they have suffered has taken place during the forty-six years which have elapsed since the casts were made. As to reading from squeezes, all experienced epigraphists are aware that it is easier to read a corroded inscription from a squeeze than from the stone itself. On the stone the forms of the letters can only be made out from slight differences of light and shadow, often obscured by stains and lichens; whereas on a squeeze the deceptions arising from local colour are eliminated; while by reading in bright sunlight from the back of the squeeze the letters stand out in relief on a darker ground, instead of being merely duskier shadows on a dusky ground, as they appear upon the stone. I have been able to decipher inscriptions from squeezes which were illegible on the stones. Perhaps the most wonderful decipherment of recent times, Dr. Euting’s reading of Yehau-melek’s altar dedication at Gebâl, was made from a squeeze, the stone itself being hopelessly corroded. Hence, I maintain what to Dr. Vigfusson may seem a paradox, that for epigraphic purposes the casts of 1841 are more valuable than the present originals, and also that the squeezes from the casts are practically more useful than the casts.

Mr. Kermode’s letter having rendered it unnecessary to re-read the inscriptions seriatim, I will merely give a selection from my notes on half-a-dozen inscriptions which Mr. Kermode has either left unnoticed, or as to which his remarks may be supplemented, or, in one or two points, corrected from the casts.

The stones are referred to by the numbers used by Dr. Vigfusson.

No. 1. The Mal-Lumcun cross at Kirk Michael. On the front, Dr. Vigfusson reads:— MAL : LUMCUN : RISTI : CRUS : THENA : EPTHR : MAL-MURU : FOSTRA : SINA : TOTER : TUFCALS CONO : ES : ATISL : ATTI; and on the back, [BE]TR.: ES : LAIFA : FUSTRA : CUTHAN : THAN : SON : ILAN.

Mr. Kermode reads raisti, fustra, sine, totir, cona, and ati. The casts support him in all these cases except sine, which I think Dr. Vigfussofl rightly reads as sine.

In the first word of the inscription on the back of the cross the e of [Be]tra is visible on the cast, thereby confirming Prof. Stephens’s conjectural restoration. Dr. Vigfusson’s reading Mal instead of Cumming’s Nial is undoubtedly correct.

This cross is of unique interest, as it is inscribed with runes of an older type than those used elsewhere. The style of ornament is also more archaic than on the other crosses. On these grounds I contend that this cross is the oldest of all. Dr. Vigfusson, however, thinks that it is "by no means the oldest." This opinion may be tested by Dr. Vigfusson’s own admissions. He notices on this stone two words which belong to an archaic stage of Norse speech. He states that "in the whole range of Old-Norse literature" there is no other instance of the initial th being retained in the word than. It is, however, found in several ancient inscriptions from Scandinavia, which are older than any literary documents. Such, for instance, are the inscriptions at Husby and at Brösike (Stephens’s Runic Monuments, vol. i., pp. 704, 739, 929).

Dr. Vigfusson also points out that the form es is more archaic than er, which replaces it in the twelfth century. Now he assigns the oldest of the Manx crosses to the end of the twelfth century. This inscription, he thinks, is "by no means the oldest" and therefore, according to him, it cannot be earlier than the thirteenth. But, on his own showing, it has one form older than the twelfth, and another form older than anything in old Norse literature. It may, therefore, well belong to the close of the eleventh century—a date to which, on other grounds, I have assigned it.

Dr. Vigfusson accounts for these archaisms by the untenable hypothesis of English influence. The archaic form of the runes, the archaic style of the ornament, and these archaic words constitute a strong cumulative argument in favour of the antiquity which I have claimed for this cross.

On the other hand the two grammatical solecisms on which Dr. Vigfusson relies in proof of a late date seem to be mere blunders. The pure Celtic names and the pure Celtic ornament point to a Celtic influence which would account for an imperfect acquaintance with the niceties of Norse grammar.

No. 3. The Mael-Brigti Cross at Kirk Michael. This is one of the beautiful crosses made and signed by the artist Gaut. On the Kirk Andreas cross Dr. Vigfusson admits that the name is Gaut. Here on theoretical grounds, he contends that it should be Gout. The casts ahow that Mr. Kermode is right in affirming that it is Gaut. Dr. Vigfusson reads the first word as Mal. The cast shows four runes, and the reading is either Mail, as Mr. Kermode says, or Mael, which would be a more correct form. Dr. Vigfusson reads thano. The cast has THONO, which is the way the word is spelt in three other places.

No. 4. The Ufaic cross at Kirk Andreas. Dr. Vigfusson reads this inscription . . . AFTIR: UFAAC : FAUTHUR: SIN : EN CAUTR: CIRTHI: SAUNR BIARNAR : FRA : CULl. This transcription at once illustrates Dr. Vigfusson’s power of suggesting briffiant emendations, and his curious inability to read runes which are perfectly plain to men of far inferior scholarship. Bracketing all defective letters, the cast reads: . . . TH[A]NA : [A]F : UFAIC : FAUTHUR : SIN : [E]N CAUTR : CIRTHI : SUNR : BIARNAR : FRO : CUL[I].

According to Cumming the fragment begins with the word thana, which Mr. Kermode still reads upon the stone. The next word is not aftir but af. The cast reads not Ufaac but Ufaic; a reading which is supported by Prof. Rhys’ s opinion that it represents the Gaelic name Ua-Feic. The cast has sunr, which is Cumming’s reading. Dr. Vigfusson, instead of mistrusting his own impossible reading saunr, courageously defends it on phonological grounds, although in six other cases in which the word "son" occurs he admits that the vowel is u and not au. Moreover, on this very cross we have fauthur, and it would be strange indeed for the vowels in "father" and "son" to be represented by identical symbols. The cast supports Mr. Kermode’s explanation of how Dr. Vigfusson fell into this error. He plainly mistook the sign of word division for a rune. There can be no doubt as to the correctness of Dr. Vigfusson’s explanation of the last two words, which have puzzled all preceding decipherers. The true reading, however, is fro : culi, and not fra : culi.

No. 6. The Thurlaf cross at Ballaugh. Dr. Vigfusson reads :— OULAIBR LIUTULBSUNR : RA .... US : THONA : AFTIR ULBSUNSIN. Dr. Vigfusson asserts that "all three names have hitherto been wrongly read." Cumming reads the first two names as Thorlaibr Thoriulbsunr. As to the first name it is quite certain from the cast that it begins with the rune thorn (th). The name, I think, should be Thurlaib[i]r. In the next name it is quite impossible that the first rune can be 1; and though it is somewhat defaced, I think there is little doubt that it was th, as Cumming has it. The next words are

R[A]STI : CRUs : TH[]N[A] : AFT : F.

No. 8. The first rime of Grim’s name, which Dr. Vigfusson omits, is visible on the cast.

No. 13. The Sondulf cross at Kirk Andreas. Dr. Vigfusson reads :—SONT : ULF : EIN : SUARTI : RAISTI : CRUS : THONO : AFTER : ARIN : BIAURC : CUNU : SINA.

Here I thinkDr. Vigfusson has wrongly read the five words—thono, after, Arm : biaure, cunu, and sina.

No. 17. The Conchan cross. I agree with Mr. Kermode in being unable to accept Dr. Vigfusson’s readings of this difficult inscription. He inserts, without any sign of hesitation, runes which were broken away in 1841 ; in other cases the casts shows that his readings are impossible; while the word he reads crist is plainly ares; and Cumming’s reading, aftir sun sin, agrees better with the cast than Dr. Vigfusson’s ift ...lusina, of which I can make no sense. The reading aucraither, is I think, impossible.

No. 20. The Distington cross. Dr. Vigfnsson reads :— N : ROSCIL : UILTI : I : TRIICU[M] : AITHSOARA : SIN.

This inscription is a fragment, the first portion being broken off. Dr. Vigfusson has omitted two letters which Cumming gives correctly — the reading being Roscitil, and not Roscil. All the words are divided by colons (:); but there is no colon, or any space for a colon, between the first two letters. Hence, the first letter, which Dr. Vigfusson reads as N must form a part of the name Roscitil. A portion of this letter is broken off; but the part which remains on the cast shows it could only have been h or n. Nroscitil would be impossible as a Norse name; but Hroscitil (=Hross-Ketill) is a regular form, and exactly what we should expect.

Dr. Vigfusson elsewhere remarks that there are only two words in the Manx inscriptions in which the letter h could have occurred, this being one of them; but, not being able to find it, he adds that the letter had evidently dropped out of the Manx language of that day, and says that the very form of the Manx letter is therefore unknown. The cast, however, shows a portion of this rune just where we should expect to find it, and the portion that remains indicates that the rune had the later Scandinavian form.

In the last word of this inscription there are certainly four runes, not three, and the reading is either sine or sino.

There are many minor corrections of Dr. Vigfusson’ s readings which I have noted; but this letter has extended to such an inordinate length that I must omit them. Enough, however, has been said to show that Dr. Vigfusson’s new recension cannot be considered final. In fact, while I find in ten inscriptions forty-five runes as to which I cannot accept Dr. Vigfusson’s readings, I have only discovered fifteen runes in Mr. Cumming’ s readings of the same inscriptions to which reasonable exception can be taken.

At the same time, I freely acknowledge the superiority of the translations which Dr. Vigfusson offers, and I think it probable that it is owing to the very excellence of his Norse scholarship that he has fallen into so many errors. He has read the inscriptions as they ought to have been written; whereas they were incised by illiterate masons, whose native speech was probably Gaelic, and whose knowledge of Norse must have been imperfect.

ISAAC TAYLOR.

Oxford, March 14, 1887.

Canon Isaac Taylor goes on arguing from the shadow (a hollow, distorted shadow), instead of from the thing that cast the shadow. I have been on the spot, seen cunu on the stone, yet he goes on arguing and drawing inferences as if there stood cuinu. I have seen the Ballaugh Cross, and read thereon the names Olave, Liotuib, and Ulb. This is manifest even to the dinunest and dullest of eyes, for the runes are almost as distinct as they were when first carved. This is admitted (privately) by your Manx runologist, yet the Canon goes on arguing from his casts that it must be a Thurlaib, Theoduif, Ub, and so on. Now, the difference between a copy and an original is this, that when ever they differ the copy goes to the wall, just as when your watch is behind time you send the watch, not the sun, for repair.

A cabman who argues from what he has seen is better than either clerk on canon arguing from the shadow of a thing he has not seen. I have been on the spot, dug out the crosses, seen and read what is thereon; and Canon Isaac Taylor has not. I have taken some trouble to make sure of my facts, and he has taken none. He further appeals from me to a gentleman who neither knows Norse nor runes—crede experto used to be said, crede inexperto is the new device.

From this otherwise barren controversy I gather this fact, that the casts of 1841 were taken after the crosses had been immured or covered up with earth. The workman, one sees, did not bother about what was underground, and what was above ground he executed ill, By the way, "aucraithcr," on the Conchan cross, is a mere misprint, long since noticed by me; transpose i and th and you are right. Further, the Distington cross in Cumberland I never saw; a copy was sent to me at Oxford, and it was inserted in our paper (which I finished on October 6, ere leaving the Isle of Mann). I forgot to mention this. Nineteen crosses I saw and read on the spot, the twentieth I saw not. Very sad; better, though, than to have seen none.

Canon Taylor never wrote to me, and never offered to me the use of his casts (those absurd, good-for-nothing casts, would they were clean out of court!) He wrote earnestly, I am told, to the editor of The Manx Note Book, who lives in his native Isle. Informed of this, I at first dissuaded my Manx friend, and begged him not to make a stirabout, as it were, of a bad copy and a fair original. But at last, pressed hard, I proposed that Canon Taylor should send his casts or rubbings to me here in Oxford, or, if he preferred, to a friend here whom I named, and I promised to have a look at them and to report on them; but he never sent any.

Canon Taylor, trusting in his casts, and not looking out for better information, last summer went into print about these vexed Manx runes; since then the originals, partly hidden underground when he wrote, have risen out of the earth to testify against him. Woe to me who have raised that ghost! for authors in the wrong or worsted are the most unreasonable creatures on this earth of ours. We are all of us so, and no better; rather than yield up our quills to the victor, as did the knights of old their swords, we fill graveyards with columns of dead print.

GUDBRAND VIGFUSSON.

P.S—In my last letter (ACADEMY, March 5, p. 168, col. 1), line 28 from bottom, for "v," read "r"; and line 26 from bottom, for "dustus," read "ductus."