by a dot placed underneath the simple form of the letter. Thus, he conceived that the aspirate, or h sound, of ll, would be well expressed by a single l dotted below; and similarly th, or dd, the derived sounds of t, or d, by a simple dotted t or d. The Welsh w was to be represented by an underdotted u, or o; and the ordinary sound of ff by ph. Hereby a double advantage was secured: the necessity for the use of the type representing h was done away with, and space was economised by the reduction of the book to a smaller compass. It is proper, however, to mention that Dr. Roberts appears also to have been actuated by a further motive, less admissible, perhaps, than that of necessity. He appears to have been desirous of falling back upon the old lines, and substituting the general use of the orthography of older MSS. for that which had become familiar to his countrymen in his own day. In that it may safely be asserted he was in error. History does not retrace her steps, although, from another point of view, it has been rightly said that she "repeats herself". The orthography of every pure and unmixed language represents the pronunciation of that language in the stage of advancement in knowledge and refinement which it has actually reached at the period of its adoption, and the attempt to fall back upon it is as impracticable as to make the widened waters of the Thames or the Dee to flow back to their source from their estuaries below London or Chester, as to induce the English or Cymric peoples to return to the uncouth forms which were in use during the periods of the gradual progress of transition of their respective languages towards the perfection of their final development. If such were the case, it would be equally proper for the pronunciation to fall back in parallel lines with the orthography, and to pronounce words now written with th and dh, as though they were spelt with a t and a d; and with a v, as though they were written with an

f. Instances of signal failure of such attempts are to be found in that of Drs. Hare and Thirlwall, to revive old English spelling in their translation of Niebuhr's History of Rome, in which, among other solecisms, the final syllable of the past tense of verbs was spelt with t, instead of ed; and, again, in the well-known example of the orthography adopted by Dr. Owen Pughe, in his first edition of the Welsh Dictionary, in which we are puzzled to recognise syllables written with a z as those to which we had become familiarized from our childhood as spelt with a dd, pronounced by us naturally as dh; and again, in the reversion to the v of the MSS. of the fifteenth century, in Tegid's edition of the Works of Lewis Glyncothi, for the single ƒ of the sixteenth century, to which the national eye and ear had become irrevocably and irrecoverably accustomed by the nineteenth. These learned and indefatigable writers, to whom we of this generation must feel ourselves so deeply indebted for the enlargement of our knowledge in Celtic literature, would seem to have failed adequately to have imbued their minds with the conception of the fact that many of the words which they found in the ancient MSS. written with a single f, were originally pronounced with the hard sound of that letter, and that the necessity for the double ff was created by the gradual softening of some words so written to the pronunciation of v, in order to distinguish the latter from those in which the original hard sound was retained. If to this view it be objected that the orthography of the double ff for f is to be found in English books and writings of the sixteenth and seventeenth centuries, and that in writing English, the single ƒ has gradually, but at length totally (with, perhaps, the solitary exception of the word of), been substituted for the double ff, while in Welsh the double f has been retained to this day, in order to distinguish it from the single f; the use of which, in that language only, is con

fined to representing the sound of v; the answer seems to be that the unsightliness, approaching often to the grotesque, when the v form is employed at the end of a word, has rendered its use too unpalatable for general acceptance, however plausible, and even scientific, may be deemed the reasons for its adoption from a different point of view. A more striking example, however, of the sheer impracticability of maintaining such a system of orthography can scarcely be furnished than by the fact that the system adopted by Dr. Griffith Roberts himself was, perforce, partially abandoned by his immediate successor in the printing of Welsh works abroad, not to say his own devoted friend and admirer, Dr. Roger Smith, in the edition of the very next publication to the Athrawaeth, the Drych Cristionogawl, and in the apology which, as we have seen, he has offered for it in his Preface. There he tells us that he has substituted a dot over, for the dot under, the letters d and 1; because, having been at the pains to procure English type, he was unable to obtain a sufficient quantity of the latter. And, moreover, where his supply fell so far short that he was unable to carry out his own system of over-dotting the letters in its entirety, he was fain to introduce an h after d and ; and that not only for the reason already given, but also because to his own judgment, this method appeared preferable to that of doubling those consonants, which was then coming into use, being more agreeable to the practice of the ancient Welsh writers: "yr hen Gymreigwyr gynt." Despite his well-meant efforts, however, to counteract it, the system of doubling the consonants so "mightily grew and prevailed", that it quickly superseded every other, and spread so widely, that in our own day we find it adopted everywhere; and the other-save in the cognate dialect of Gaelic, and in the method of orthography introduced with equal failure of success into Edward Lhwyd's Archæologia Britannica-nowhere:

yet, doubtless, the soft sound of th, as in the English word the, is more naturally, as well as scientifically, represented by the use of the true symbol of the aspirate h after d, as dh. The whole story reminds us of the protest made by classical scholars against the introduction of the use of the word 'telegram' for a message by electric wire; whereas, the true classical usage would have required 'telegraphem', as, in fact, was abundantly proved by very learned letters, published in the Times and elsewhere. The principle of utilitarianism and expediency prevailed over that of grammatical correctness, to the triumph of 'telegram' over 'telegraphem', unless, indeed, we ought to call in Professor Rhys to assist us, who might possibly refer us, for the true explanation of the seeming incongruity, to the principle of ' phonetic decay'."

To revert, however, to Dr. Roger Smith's Preface, from which we gather information on another important point, namely the original scope of the "Drych". Of this he tells us that the MS. sent him by its author, Dr. G. Roberts, consisted of three parts, the two latter of which he purposed to bring out as soon as he could; a purpose, however, which, as far as we know, he never was able to effect, as nothing whatever, up to this time at least, appears to be known of their existence. The first part consists of about seventy pages, and is a treatise, as far as I have been able hitherto to ascertain from a cursory examination, on the Love of God. is still possible, but scarcely it is to be feared probable, that the other two should be brought to light at this distance of time, unless, indeed, copies may have providentially been preserved in MS. in the public or other library in Milan.

Of the author, Dr. Griffith Roberts, it is disappointing to find that so little information is forthcoming. Canon Williams, in his Dictionary of Eminent Welshmen, tells us that he was "a learned grammarian, of whom nothing further is

known than that he was educated at the university of Sienna in Italy, under the patronage of William Herbert, Earl of Pembroke," but gives no authority for the statement. Rowlands, in the Cambrian Bibliography, following apparently Moses Williams, calls him, "Griffith Roberts, Esq., M.D., i. e. Doctor of Medicine; but this is clearly apocryphal. Rowlands' editor, the Rev. D. Silvan Evans, annotating this, has succeeded in eliciting a ray of light to throw on this dark subject from a paper in the Harleian Miscellany (vii, 132), where it is said that he was Confessor to Cardinal (St. Charles) Borromeo at Milan; and Dr. R. Smith confirms this by calling him in his Preface, "yr Athro mawr o Dhinas Fulan yngwlad yr Idal", the great Master of the City of Milan in the land of Italy, proving the high estimation he was held in for his learning, and also, perhaps, for his piety. On reference to the Miscellany, the statement appears in a tract printed in London in 1590, with a very long title, headed "The English Romayn Life", etc., " Written by A. M., sometime the Pope's Scholler in the Seminary among them." A. M. was a person named Antony Munday, one of those disreputable spies, in which capacity he must have been a scholar, if at all, in the pay of Elizabeth's ministers, employed to ferret out information respecting the Catholics abroad in exile for their religion, with the view to found evidence against them upon it in the event of their return to England, and described, as stated in a note by an opponent of his, as having been "first a stage-player, after an apprentice, which time he wel served with deceaving of his master; then wandering towards Italy, by his own report became a cosener in his journey. Coming to Rome, in his short abode there, was charitably relieved, but never admitted in the seminary (as he pleaseth to lye in the title of his book)." His narrative contains (to say the least) monstrous and incredible exaggerations, of a character similar to those proved in the case of