The history of ancient and early mediaeval times requires to a far greater extent than more recent history the aid of various other sciences, not the least of which is the science of language. And, although the first object of these Studies was to demonstrate to specialists various unrecognized or imperfectly recognized linguistic facts, the importance of those facts in themselves is much less than that of their historical consequences.

The main historical result of this book is the settlement of c the Pictish question or rather of the two Pictish questions. The first of these is ‘What kind of language did the Picts speak?*. The second is ‘Were the Picts conquered by the Scots?

The first has been settled by linguistic and palaeographical methods only : it has been shown that Pictish was a language virtually identical with Irish, differing from that far less than the dialects of some English counties differ from each other. The second has been settled, with very little help from language, by historical and textual methods: it has been made abundantly clear, I think, to any person of impartial and critical mind that the supposed conquest of the Picts by the Scots is an absurd myth.

The Highlander, as we call him—the Albanach as he calls himself in his own Gaelic—is, indeed, in the vast majority of cases simply the modern Pict, and his language modern Pictish. To suppose that the great free people from which he is descended were ever conquered by a body of Irish colonists, and that the language he speaks is merely an Irish colonial dialect, are delusions which, I hope, no one will regret to see finally dispelled.

The next most important results of these Studies are the demonstration of the great prominence of the Belgic element in the population of the British isles, and the evidence that so many of the tribes known to us as inhabiting England and Wales in Roman times spoke not Old Welsh, as has hitherto been supposed, but Old Irish. Particularly notable for wide dispersion and maritime venture are the Menapians, and it is a pleasure to me to have traced to them the origin of the Manx nation and language.

As regards Continental history, the great Goidelic element is now shown to have extended with more or less continuity from the Danube to the mouth of the Loire, and from the Tagus and the Po to the mouth of the Rhine.

And here let me add a very necessary caution. Names which have not been purposely invented to describe race must never be taken as proof of race, but only as proof of community of language or community of political organization. We call a man who speaks English, lives in England, and bears an obviously English name (such as Freeman (or Newton) an Englishman. Yet from the statistics of ‘relative nigrescence’ there is good ground to believe that Lancashire, West Yorkshire, Staffordshire, Worcestershire, Warwickshire, Leicestershire, Rutland, Cambridgeshire, Wiltshire, Somerset, and part of Sussex, are as Keltic as Perthshire and North Munster ; that Cheshire, Shropshire, Herefordshire, Monmouthshire, Gloucestershire, Devon, Dorset, Northamptonshire,| Huntingdonshire, and Bedfordshire are more so—and equal to North Wales and Leinster ; while Buckinghamshire and Hertfordshire exceed even this degree and are on a level •with South Wales and Ulster. Cornwall, of course, is more Keltic than any other English county, and as much so as Argyll, Invernessshire, or Connaught.

What is sauce for the ‘Saxon’ is sauce for the ‘Kelt’. The Keltic-speaking peoples of antiquity may have incorporated other Aryan or non-Aryan tribes, and the Keltic language of any given region may have been introduced by quite a small minority of conquerors—like the English language in Ireland. Even as between the Irishman and the Welshman, the language-test is not a race-test: both in North

No one was more intensely ‘English’ in his sympathies than the great historian of that name, and probably no one would have more strenuously resisted the suggestion that he might be of Welsh descent: yet I have met his close physical counterpart in a Welsh farmer (named Evans) living within a few miles of Pwllheli and in South Wales many scores of thousands of the ‘Kymry' are probably descended from ancestors who spoke Irish ; and it is equally possible—though I know of no evidence for the supposition—that the Goidels of Ireland may have absorbed tribes, or portions of tribes, which originally spoke Kymric.

If, therefore, I have anywhere referred to any people, or the users of any given language or family of language, in terms which might be thought to imply that they were all of one primeval physical stock, I must disclaim that interpretation. In other words, such a term as ‘Goidels' is to be taken as meaning nothing more than an aggregate of people who speak Goidelic, or whose ancestors spoke it.

The chief linguistic result of the Studies (apart from the determination of the nature of Pictish and of the parentage of Highland Gaelic) is the fact that the loss of original, a loss supposed to be the distinguishing feature of the Keltic family of language, is of comparatively late date in the Goidelic branch—that, in fact, p was normally kept (see p. 205) for centuries after the Christian era, at Bordeaux till the 5th cent., in Pictish probably later still. I strongly advise those who read these Studies chiefly for linguistic purposes, or who would satisfy themselves of the soundness of the linguistic foundation, to pass to the Appendixes immediately after reading the first 8 pages. Two of those Appendixes have, indeed, been published before—‘Sequanian* as a pamphlet, 'Pictavian’ in the Zeitschrift fiir celtische Philologie; but the former has been largely revised and corrected, and the demonstration that the Rom tablet is in rimed metre is an important addition to the revised reprint of the latter.

My constant references to living scholars are themselves recognitions of indebtedness, yet I cannot help adding that, but for the Urkeltischer Sprachschatz of Dr. Whitley Stokes, ‘ the grand old man of Keltic philology, this book could never have been written.

While it was passing through the press, Prof. Anwyl sent me a paper of his own which was likely to interest me. I sent him in return a sheet or two of my proofs and offered to send the rest. He has very kindly glanced through them all, and has sent me notes on them: two or three have been corrections of small slips, some have been cautious counsels to be occasionally less positive in statement, and others have been new facts and suggestions, always interesting, sometimes important, which I have made use of with due acknowledgement, and almost all of which will be found in Appendix VII. But, while congratulating myself on this kind service, I should be sorry if the reader considered Prof. Anwyl as endorsing any mistake which has escaped his attention. In a work of the extent and character of this, even a trained Keltic philologist would find it difficult to avoid absolutely all error of statement or of inference: in my case, I can only hope that such errors may be few and unimportant—I dare not dream that they have been escaped altogether.

The body of the book, that is to say pp. 9-111, was begun in Dec. 1900, and was meant to be quite a short paper on the Menapii, Parisii, and Belgae, in England—to be offered to the Zeitschrift fiir celtische Philologie as a sequel to my ‘ Language of the Continental Picts I was led on, however, from point to point till, at the end of September 1901, the 1 short paper ’ would have filled 90 pp. of the Zeitschrift; and, on my informing the editors, they very reasonably told me that they could not spare the space. In order to fit the material in some degree for another periodical, I then wrote the introductory matter on pp. 1-8, and was on the verge of completing it when a great breakdown of the heart prostrated me. On recovering sufficiently, I sent an analysis to the editor of a well-known Keltic society’s journal, asking if it would be of any use to forward the paper itself; but the favourable answer did not come for some seven months, and meanwhile I had thought best to print for myself. I mention these things for two reasons. Firstly, that the reviewer and the general reader may understand that the book was written simply to prove certain facts, in the smallest space, to a limited scientific circle. Secondly, that my friends, and the University to which I owe serious duties, may know that since the illness referred to I have not been guilty of writing more than about the last 34 pp.—besides the index, in which my daughter Myrtle has helped me. I have, indeed, not even read my own proofs more than could possibly be helped, but have been glad to avail myself of the aid of Mr. Strickland Gibson, M.A., of the Bodleian staff, who had already copied much of my manuscript for the printer. I have also to thank Mr. Gibson for calling my attention to Henri Monin’s Monuments des anciens idiomesgaulois, to which I owe my knowledge of the Am61ie-les-Bains tablets.

I must thank Mr. Hugh W. Young, F.S A. Scot., the owner of Burghead, for letting me reproduce photographs he had given me of the Burghead stone; Mr. James Milne of Arbroath for letting me reproduce three photographs of the St. Vigeans stone, taken by his late father ; and Prof. Camille Jullian for sending me two photographs of the Rom tablet for the same purpose. M. Ernest Leroux, the present publisher of the Revue Arch£ologiques has informed me that there is no longer any copyright in the plate of the Amalie- les-Bains tablets, but I have to thank him nevertheless for making no objection to my reproducing it.

The maps have had to be made against time. Using chiefly those of Haverfield and Longnon, I have asked Mr. F. C. Wellstood, of the Bodleian staff, to fix many of the positions and draw boundaries for me; have then written in the names on outline-maps of Messrs. George Philip and Son, with their permission; and have finally obtained from Messrs. Darbishire and Stanford, of Oxford, the services of a draughtsman, Mr. E. R. Bryant, to make a fair copy on duplicate outline-maps, which the Clarendon Press have reduced and collotyped.

It would be very ungrateful of me not to add the expression of my obligations to the Controller, Mr. Horace Hart, and the staff of the Clarendon Press, for the great pains they have taken to produce a most troublesome volume exactly to its author’s liking.

I have given that volume a shorter and wider heading, as well as a longer and narrower, partly for ease of comprehension and citation, but partly also because, if all goes well, I may in some future year issue another volume of‘Keltic Researches’, containing many separate studies on obscure points in British history and antiquities, chiefly between the Roman and Danish invasions.

I should have liked to add much on the vastness and richness of the harvest which awaits labourers in the fields of Keltic philology and Keltic antiquarian research. But, until I know a University which could—or a rich man who would —do something to provide the labour, I fear that I should only be wasting time.

But it may perhaps be of some little help to another cause which many more than myself have at heart—the preservation of the surviving Keltic languages—if I add here an extract from my own book ‘Golspie’. And what I have there said with regard to the languages of the British isles applies equally, nuitatis mutandis, to Breton.

‘No sensible man who wished the Highlander to live in intimacy and friendship with the other races which inhabit these isles, or who wished to see him cultivated and prosperous, would do otherwise than wish him to speak and read English well. But I hope the day will never come when Gaelic will become extinct in the Highlands, as unhappily Cornish was allowed to become extinct in the eighteenth century. In it are imbedded no small part of the Highlander’s history—the history of his settlements, the history of his descent, the history of his thought, the history of his culture. It is not only bad for a race to forget such things, but it is bad for science too: no study of a dead language can recover for us all of that knowledge which would have been transmitted by its preservation. Every Highlander, every Irish Gael, every Manksman, and every Welshman, should know and speak the speech of his fathers, and should see that his children also know and speak it. And every government should show for all such healthy developments of race-feeling that sympathy which is the best bond of union.’

You can download the book here