Pwo Latin grammars of the language followed, the Cambrobrytan Mere Lingum Itudiinenta' of Dr. David Rhys, in 1592, and the' Antiquin Linguao Britannicm mum communiter dictm Cambro-Britannicai Rudi menta ' of Dr. John Davies, in 1621, both of them much esteemed. Mere are now several grammars of the Welsh language in English, of which that by the Rev. Thomas Rowland, the second edition of which tbublilhed in 185/, may be rOcommended as the most satisfactory. Dictionaries are less numerous. Dr. John Daviee, the author of the ginunnuar, published in London in 102 ft 1rt1811 awl Latin dietionary, ' Antiqiuu Lingum BritsumiWn.flietiontartum Duplex,' which eontlnned In repute for A century and a half, till superseded by that of Dr. Owen '1 his won:, Gciriadur Cyniraeg a Sacsaneg, a Welsh and Englieln Dictionary, with 4 grammar prefixed, published In two closely. printed octavo volutnee In 1793, and again in 1829, is still the only Welch dictionary on a Sufficiently extensive scale, and it a work of great labour and Merit, but It is open to serious objections. The num ber of words it contains is nearly 100,000; but many of these are com pounds, formed on regular principles, which only servo to swell the bOok, and many of them are MAT words Which (night to exist in Welsh, aeccirtling to the leileographet s opinion, 6.tlier than words which actually exist in it. A third edition le now (1861) in course of publi cation, under the editorship of lt. J. who promisee that the work shall contain of thouganda" of words not in the previous editions. As comprises Welsh and English only, not Engliah and Welsh, the want of the latter Was fortnetly supplied by the excel lent English-Welsh dictionary of the Itev. John Walters, of which a new and improved edition wad published abntit 1825. The English and Welsh by Daniel Silvan Evans, in two thick volitnnes (Denbigh, 1552-1858), is however to that of Walters. There are a compendious grammar and dictionary by Mt. Spurrell of Carmarthen, both of which will be found useful. Seine information on the Welsh dialects is given in the ' Essay on the Present State of the Welsh Language,' by ND'. John fIuglies, published about 1820. There are differences in ptohunciation and idiom sufficiently marked to render it diffiehlt for persons from retnote districts to con verse with each other ; and 'orth Wales is more pure and correct In its language than South Waleg.
The System of spelling in Welsh coffee/ion:de with and represents the piononciation, and in that respect it has a marked superiority not only over Its kindred Celtic languages, the Irish and Gaelic, but, as we daily feel to out cost, Over the English. There are at the same time some peenliatities in Welsh spelling, the Motive Of which is not vcry plaid, and the effect of width is often ludicrous.. The sound which is generally repretehted by the letter v id In Welsh represented by f, even in proper names, so. that pronouncing Calvin and Virgil not unlike ourselves, a Welshintin wtltes " Caltin " and " Firgil." sound represented in the Etiglieln alphabet by f Is represented in Welsh by f ; and thus Fox and Franklin Meat bo 'written Ffot and Ffranklin, as, if spelt in the ordinary mariner, a Welsh reader might ptonounce them Vox and Wealth. Settle ellialtts have been made by Owen l'ughe and others to introduce the missing a' Into the Welsh alphabet ; it is Wend for English writers to folio* Ids system spelling Welsh names—to write 3letthyt-Tydvil, for instance, Itistezul of Merthyr-Tydfil. But the advantage of uniformity is so great, that though the present system scarcely dates Maher bade than the Pith century, and though there are several eccentricitiea like that with the e, teeistance to innovation like hitherto ttiuMphed.
It has been said, that in English the gramtnat is excellent and the dictionary is execrable. The grammar of English is, in fact, distin guished by its great comparative freedom front needlesa complexities, while its list of wade comprises thousande that a careful Writer will carefully avoid. in Welsh, almost the converge is the ease. For nee= less intricacy ita gramniar has a had pre-eteineneo. It is pervaded from first to lag by a certain law of "permutations," the natttre of which will best be understood by an Meta:nee. The word for "father"
in Welsh is fall ; the word for " Is fry : but to say Fg tad for "thy father" would be an unpardonable solecism. After fy, every word beginning with a t must change the to tilt, and the correct phrasd id therefore fa nhad. The Word for " thy " is dg but after thia a different change is requited: father" is dg dad. The word el in Welsh means " his " or " but according to the sense requires a different mutation to follow : father" le expressed by a dad, and "her father" by ei that!. some of the letters, as t, have three of these mutations to undergo, some only two, some only One, and some none at ail. There is a multitude of intricate rules to determine In what circumstances the different mutations are to be used, and at the game time the existence of letters which undergo no mutations whatever, and tho words beginning with which are just as elegant and forcible witty out them, proves to demonstration time utter uselessness of t110 IV1101e. in the other Celtic languages there is a similar system, hut hot carried quite so far as in the Welsh. The principle has been generally can metered to be peculiar to that family of letigtutgeg in Europe; but Mime Louis Linden Bonaparte, in his extensive researches among the obscurer tongues, has discovered Alight traces of mutation Id some of the dialects of Sardinia. In the complexity caused by an Intricate system of declenelons and conjugations there is generally some coin' pens.ation obtained in a superior precision and force; but there is nothing of the kind in mutation. Its origin was long a Subject of debate. In some compounds of words It was plain that attention to euphony had given rice to it. The opposite to " In English, is : both words are taken from the Latin, and the " In " privative, as it is is altered to " ire" in thin) and other eases before the letter p, to avoid unpleasantness of sound. The Welsh have also borrowed the word pceriU from the Latin, and for "impossible" they say anmhosibt The Welsh particle answering to onr " is en : the collision of n and p was also to be avoided ; but 'rigged Of altering the last letter of the nn, they altered the first letter of posiSf. It has lately been pointed out Ly Zeus§ and his school, that in other eases where p is altered to tnh, it was in the old forma of the language pre ceded by an n which no longer exists. Hence it may be inferred that the whole system of mutations was originally founded on the euphonic principle. While the effect remains, hoverer, the reason for it has disappeared in tho actual state of the languago. The rules for muta tions are at the present day mere arbitrary rules, the motive for which is as unknown to time who practige them as the Icemen for spelling " wright" with w and gh Is unknown to the schoolboy who Is being taught to spell. After all these sacrifices to harmony, the Welsh has never been considered harmonionts by those of whom it was not the ruother-tongue, though a write- In the %Cambrian ilegister' for 181R, affirms that " strangers to bath languages frequently mistake the Welsh for The principal beauty of Welsh we a Language consists in the facility ithieli it possesses of forming derivatives and compounds, and the com pleteness with Which that power has been exercised, making full and ingenious use of every root, and thus avoiding that useless borrowing of terms from other languages which has been carried to such an absurd extent in English. Many foreign Words have indeed been introduced into Welsh, but they have been so aselinilated to those of native origin as often not to be easily detected. The a hole language seems of a piece. But there is this advantage in a language in which the grammar is superior to the dictionary : that by a skilful choice of words, by limiting himself, like Metastasio, to a select vocabu lary, an author may exclude from his writings all but the beauties of the language ; while in the opposite ease, the defects—a9 in Welsh, the rules for inutation—are involved in every sentence and intrude at every turn. These rules alone make it much more difficult for an Englishman to learn to speak Welsh than for a Welshman to learn to speak English.