is so fixed an idiom of our pronunciation, that to alter it, would be to alter the sound of the whole language. It must, however, be regretted, that it subjects our tongue to some of the most hissing, snapping, clashing, grinding sounds that ever grated the ears of a Vandal: thus rasped, scratched, wrenched, bridled, fangled, birchen, hardened, strengthened, quickened, &c. almost frighten us when written as they are actually pronounced, as raspt, scratcht, wrencht, bridl'd, fangl'd, birch'n, strength'n'd, quick'n'd, &c. they become still more formidable when used contractedly in the solemn style, which never ought to be the case; for here, instead of thou strength'n'st or strength'n'd'st, thou quick'n’st, or quick'n'd'st, we ought to pronounce, thou strength'nest or strength'nedst, thou quick'nest or quick nedst, which are sufficiently harsh of all conscience. (See No. 405.) But to compensate for these Gothick sounds, which, however, are not without their use, our language is full of the smoothest and most sonorous terminations of the Greeks and Romans.

360. By the foregoing rule of contraction, arising from the very nature of the letters, we see the absurdity of substituting the t for ed, when the verb ends in a sharp consonant; for, when the pronunciation cannot be mistaken it is folly to alter the orthography; thus the Distressed Mother, the title of a tragedy, needs not to be written Distrest Mother, as we generally find it, because, though we write it in the former manner, it must necessarily be pronounced in the latter.

361. By this rule, too, we may see the impropriety of writing blest for blessed, when a participle.

"Blest in thy genius, in thy love too blest."-Pope.

But when the word blessed is an adjective, it ought always to be pronounced, even in the most familiar conversation, in two syllables, as this is a blessed day, the blessed thistle, &c.

362. This word, with learned, cursed, and winged, are the only participial adjectives which are constantly pronounced in two syllables, where the participles are pronounced in one: thus a learned man, a cursed thing, a winged horse, preserve the ed in a distinct syllable; while the same words, when verbs, as he learned to write, he cursed the day, they winged their flight, are heard in one syllable, as if written learnd, curst, and wingd: the d in cursed changing to t, from its following the sharp consonant s. (358.)

563. Poetry, however, (which has been one great cause of improper orthography,) assumes the privilege of using these words, when adjectives, either as monosyllables or dissyllables; but correct prose rigidly exacts the pronunciation of ed in these words, when adjectives, as a distinct syllable. The ed in aged and winged, always make a distinct syllable, as an aged man; the winged courser : but when this word is compounded with another, the ed does not form a syllable, as a full-ag'd horse, a sheath-wing'd fowl.

364. It is, perhaps, worthy of notice, that when adjectives are changed into adverbs by the addition of the termination ly, we often find the participial termination ed preserved long and distinct, even in those very words where it was contracted when used adjectively: thus though we always hear confess'd, profess'd, design'd, &c. we as constantly hear con-fess-ed-ly, pro-fess-ed-ly, de-sign-ed-ly, &c. The same may be observed of the following list of words, which by the assistance of the Rhyming Dictionary, I am enabled to give as, perhaps, the only words in the language in which the ed is pronounced as a distinct syllable in the adverb, where it is contracted in the participial adjective: Forcedly, enforcedly, unveiledly, deformedly, feignedly, unfeignedly, discernedly, resignedly, refinedly, restrainedly, concernedly, unconcernedly, discernedly, undiscernedly, preparedly, assuredly, advisedly, dispersedly, diffusedly, confusedly, perceivedly, resolvedly, deservedly, undeservedly, reservedly, unreservedly, avowedly, perplexedly, fixedly, amacedly.

1

365. To this catalogue, may be added several abstract substantives formed from participles in ed: which ed makes a distinct syllable in the former, though not in the latter, thus: numbedness, blearedness, preparedness, assuredness, diseasedness, advisedness, reposedness, composedness, indisposedness, diffusedness, confusedness, distressedness, resolvedness, reservedness, perplexedness, fixedness, amazedness, häve ed pronounced distinctly.

366. The adjectives naked, wicked, picked, (pointed,) hooked, crocked, forked, tusked, tressed, and tretched, are not derived from verbs, and are therefore pronounced in two syllables. The same may be observed of scabbed, crabbed, chubbed, stubbed, shagged, snagged, ragged, cragged, scrubbed, dogged, rugged, scragged, hawked, jagged; to which we may add the solemn pronunciation of stiffnecked; and these, when formed into nouns by the addition of ness, preserve the ed in a distinct syllable, as wickedness, scabbedness, raggedness, &c.

367. Passed, in the sense of beyond, becomes a preposition, and may allowably be written past, as past twelve o'clock: but when an adjective, though it is pronounced in one syllable, it ought to be written with two, as passed pleasures are present pain: this I know is contrary to usage; but usage is, in this case, contrary to good sense, and the settled analogy of the language.

368, It needs scarcely be observed, that when the verb ends in t or d, the ed in the past time and participle has the d pronounced with its own sound, aud always forms an additional syllable, as landed, matted, &c. otherwise the final d could not be pronounced at all.

369. And here perhaps it may not be useless to take notice of the very imperfect and confused idea that is given in our best grammars,of what are called contracted verbs, such as snatch, checkt, snapt, mixt, dwelt and past, for snatched, checked, snapped, mixed, dwelled, and passed. To these are added, those that end in l, m, and ", or p, after a diphthong; which either shorten the diphthong, or change it into a single vowel; and instead of ed, take t only for the preterit, as dealt, dreamt, meant, felt, slept, crept; and these are said to be considered not as irregular, but contracted only. Now nothing can be clearer than that verbs of a very different kind, are here huddled together as of the same. Snatched, checked, snapped, mixed, and passed, are not irregular at all; if they are ever written snatcht, checkt, snapt, mixt, and past, it is from pure ignorance of analogy, and not considering that if they were written with ed, unless we were to pronounce it as a distinct syllable, contrary to the most settled usage of the language, the pronunciation, from the very nature of the let ters, must be the same. It is very different with dwelled; here, as a liquid, and not a sharp mute, ends the verb, d might be pronounced without going into t just as well as in fell'd, the participle of to fell (to cut down trees.) Here then we find custom has determined an irregularity, which cannot be altered, without violence to the language; dwell may be truly called an irregular verb, and dwet the preterit and participle.

370. The same may be observed of deal, dream, mean, feel, weep, sleep, and creep. It is certain we can pronounce d after the four first of these words, as well as in sealed, screamed, cleaned, and reeled, but custom has not only annexed t to the preterit of these verbs, but has changed the long diph. thongal sound into a short one; they are therefore doubly irregular. Weep, sleep, and creep, would

not have required to form their preterits, any more than perped, and steeped; but custom, which has shortened the diphthong in the former words, very naturally annexed t as the simplest method of conveying the sound.

$71. The only two words which occasion some doubt about classing them are, to learn and to spell. The vulgar (who are no contemptible guides on this occasion) pronounce them in the preterit art, and spelt; but as n and I will readily admit of d after them, it seems more correct to favour a tendency to regularity, both in writing and speaking, which the literary world has given into, by spelling them learned and spelled, and pronouncing them learn'd, and spell'd: thus earned, the preterit of to earn, has been recovered from the vulgar eurnt, and made a perfect rhyme to discerned. 372. To these observations may be added, that, in such irregular verbs as have the present, the preverit and participle the same, as cast, cost, cut, &c. the second person singular, of the preterit of these verbs take ed before the est, as I cast, or did eest; Thou castedst, or didst cast, &c. for if this were not the case, the second person of the preterit might be mistaken for the second person of the present tense.

373. I have been led insensibly to these observations by their connection with pronunciation; and if the reader should think them too remote from the subject, I must beg his pardon, and resume my remarks on the sound of the letter d.

374. The vulgar drop this letter in ordinary, and extraordinary and make them or'nary and extror'ury; but this is a gross abbreviation; the best pronunciation is sufficiently short, which is erd'nary and extr'ordinary: the first in three and the last in four syllables; but solemn speaking preserves the i, and makes the latter word consist of five syllables, as if written extr'ordinary.

375. Our ancestors, feeling the necessity of showing the quantity of a vowel followed by ge, when it was to be short, inserted d, as wedge, ridge, badge, &c. The same reason induced them to write colledge and alledge, with the d; but modern reformers, to the great injury of the language, have expelled the d, and left the vowel to shift for itself; because there is no d in the Latin words from which these are derived.

376. D like t, to which it is so nearly related, when it comes after the accent, and is followed by the diphthong ie, io, ia, or cou, slides into gzh, or the consonant j; thus soldier is universally and justly pronounced as if written sol-jer; grandeur, gran-jeur; and verdure (where it must be remembered that is a diphthong,) ver-jure: and for the same reason, education is elegantly pronounced ed-jucation. But duke and reduce, pronounced juke and re-juce, where the accent is after the d, cannot be too much reprobated.

F.

377. F has its pure sound in often, off, &c. but, in the preposition of, slides into its near relation v, as if written ov. But when this preposition is in composition at the end of a word, the ƒ becomes pure; thus, though we sound of singly ou, we pronounce it as if the ƒ were double in whereof.

578. There is a strong tendency to change the f into v in some words, which confounds the plu ral number and the genitive case: thus we often hear of a wive's jointure, a calve's head, and houze red, for wife's jointure, a calf's head, and house rent.

G.

379. G, like C, has two sounds, a hard and a soft one: it is hard before a, o, u, l, and r, as game, gone, gull, glory, grandeur. Gaol is the only exception; now more commonly written jail. (212.) 380. G before e and i is sometimes hard and sometimes soft: it is generally soft before words of Greek, Latin, or French original, and hard before words from the Saxon. These latter, forming by far the smaller number, may be considered as exceptions.

381. G is hard before e, in gear, geck, getse, geld, gelt, gelding, get, gew-gaw, shagged, snagged, ragged, cragged, scragged, dogged, rugged, dagger, swagger, stagger, trigger, dogger, pettifogger, tiger, auger, enger, meager, unger, finger, linger, conger, longer, stronger, younger, longest, strongest, youngest. The last six of these words are generally pronounced in Ireland, so as to let the g remain in its nasal sound, without articulating the succeeding vowel: thus longer (more long) is so pronounced as to sound exactly like the noun a long-er (one who longs or wishes for a thing;) the same may be observed of the rest. That the pronunciation of Ireland is analogical, appears from the same pronunciation of g in string-y, spring-y, full of strings and springs; and wronger and wrongest, for more and most wrong. But though resting the g in the nasal sound, without articulating the succeeding vowel, is absolutely ne cessary in verbal nouns derived from verbs ending in ing as singer, bringer, slinger, &c. pronounced sing-er, bring-er, sling-er, &c. and not sing-ger, bring-ger, sling-ger, &c. yet in longer, stronger and wanger; longest, strongest, and youngest, the gought always to articulate the e: thus younger ought always to rhyme with the termination monger, which has always the g hard, and articulating the vowel; and this pronunciation is approved by Mr Nares. Forget, target, and together, fall into this dass. See No. 409.

32. G is hard before i in gibbe, gibcat, gibber, gibberish, gibbous, giddy, gift, gig, giggle, giglet (properly gigglet,) gild, gill (of a fish) gimlet, gimp, gird, girdle, girl, girth, gizzard, begin, give, forgire, biggin, piggin, noggin: also derivatives from nouns or verbs ending in hard g, as druggist, wigguh, riggish, hoggish, doggish, sluggish, rigging, digging, &c.

393. G before y is generally soft, as in elegy, apology, &c. and almost in ail words from the learned languages; but hard in words from the Saxon, which are formed from nouns or verbs ending in ghard, as shaggy, jaggy, knaggy, snaggy, craggy, scraggy, quaggy, swaggy, dreggy, sprigru, twiggy, baggy, foggy, cloggy, buggy, muggy. "Gyve, from its Celtick original, ought to have the g hard, but has decidedly adopted the soft g.

GN in the same Syllable at the Beginning of a Word.

324. The g in this situation is always silent, as gnaw, gnash, gnat, gnarl, gnomon, gnomonicks : pronounced naw, nash, nat, narl, nomon, nomonicks.

GN in the same Syllable at the End of a Word.

335. No combination of letters has more puzzled the criticks than this. Two actresses of distin guished merit in Portia in the Merchant of Venice, pronounced the word impugn differently, and each

found her advocate in the newspapers. One critick affirmed, that Miss Young, by preserving the sound of g, pronounced the word properly; and the other contended that Mrs. Yates was more judicious in leaving it out. The former was charged with harshness; the latter with mutilating the word, and weakening its sound; but if analogy may decide, it is clearly in favour of the latter; for there is no axiom in our pronunciation more indisputable than that which makes g silent before n in the same syllable. This is constantly the case in sign, and all its compounds, as resign, design, consign, assign; and in indign, condign, malign, benign; all pronounced as if written sine, rezine, &c. In which words we find the vowel i long and open, to compensate as it were, for the suppression of , as every other word ending in gn, when the accent is on the syllable, has a diphthong pronounced like a long open vowel, as arraign, campaign, feign, reign, deign; and consequently, unless the vowel u can produce some special privilege which the other vowels have not, we must, if we pronounce according to analogy, make the u in this situation long, and sound impugn as if written impune.

386. The same analogy will oblige us to pronounce impregn, oppign, expugn, propugn, as if written imprene, oppune, expune, propune, not only when these verbs are in the infinitive mood, but in the preterits, participles, and verbal nouns formed from them, as impugned, impugning, and impugner, must be pronounced impuned, impuning, and impuner. The same may be observed of the rest. Perhaps it will gratify a curious observer of pronunciation to see the diversity and uncertainty of our orthoepists in their notation of the words before us.

impune

Sheridan, Scott, Nares, Murray. Barclay says the g in this word and its derivatives is mute, but takes no notice of the quantity of the u.

Buchanan, Kenrick, Perry.

W. Johnston.

Sheridan, Scott, Nares, Murray.

Kenrick, Perry, Barclay.

W. Johnston.

Sheridan, Scott, Perry, Nares.

impun.

impung.

oppune.

oppin.

opping.

propine.

proping.

imprène.

imprén.

of the c.

expune.

expun.

impuner.

Barclay.

Nares, Murray.

Sheridan, Kenrick, Perry. Barclay says the g is mute, but says nothing of the quantity

Sheridan, Scott, Narcs.

Perry, Barclay.

Sheridan.

impuned. Murray.

impinner. Perry, Barclay.

oppugner. Sheridan.

propugner. Sheridan.

propuner. Scott.

propunner. Perry.

Nothing is clearer than that all these words ought to follow the same fortune, and should be pronounced alike. How then shall be reconciled Mr. Sheridan's pronouncing impugn, oppugn, expugn, and propugn, with the u long, and impregn with the e short? Kenrick, who has not the word propugn, is consistent in pronouncing the rest with the vowel short. The same may be observed of Scott, who adopts the long sound, but has not the word impregn. Mr. Perry gives the short sound to all but propugn, where he makes the u long, but absurdly makes the verbal noun propunner; and W. Johnston, who has only impign and oppugn, pronounces the vowel short, and spells them impung and oppung. Barclay, under the word impugn, says the g in this word and its derivatives is mute, without noticing the quantity of the vowels, but spells oppugn, oppun; and of impregn, only says the g is mute; but writes propugn, propung, in the manner that W. Johnston does impugn and oppugn: but Mr. Nares observes, that analogy seems to require a similar pronunciation in all these words, and that the vowel should be long. The same inconsistency is observable in Mr. Sheridan's pronunciation of the verbal nouns; for he expunges the g in impugner, and writes it impuner, but preserves it in oppugner and propugner. Mr. Scott has only the word propugner, which he very properly, as well as consistently, spells propuner. Mr. Perry has propunner and impunner, and Barclay impunner only. -The inconsistency here remarked arises from not attending to the analogy of pronunciation, which requires every verbal noun to be pronounced exactly like the verb, with the mere addition of the termination: thus singer is only adding er to the verb sing, without suffering the g to articulate the e as it does in finger and linger, &c. The same may be observed of a signer, one who signs: and as a corroboration of this doctrine, we may take notice that the additional er and est, in the comparatives and superlatives of adjectives, make no alteration in the sound of the radical word: this is obvious in the words benigner, benignest, &c. except younger, longer, and stronger. See No. 381.

387. But in every other compound where these letters occur, the articulates the latter syllable, and g is heard distinctly in the former, as sig-nify, malig-nity, assig-nation, &c. Some affected speakers, either ignorant of the rules for pronouncing English, or over-complaisant to the French, pronounce physiognomy, cognizance, and recognizance, without the g; but this is a gross violation of the first principles of spelling. The only words to keep these speakers in countenance are poignant, and champignon, not long ago imported from France, and pronounced poiniant, champinion. The first of these words will be probably hereafter written without the g; while the latter, confined to the kitchen, may be looked upon as technical, and allowed an exclusive privilege. See COGNIZANCE.

388. Bagnie, seignior, seraglio, intaglio, and oglio, pronounced ban-yo, seen-yur, seral-yo, intal-yo, and ole-yo, may be considered as foreign coxcombs, and treated with civility, by omitting the g, while they do not pervert the pronunciation of our native English words.

GM in the same Syllable.

389. What has been said of gn is applicable to gm. We have but one word in the language where these letters end a word with the accent on it, and that in phlegm; in this the g is always mute and the e according to analogy, ought to be pronounced long, as if the word were written fleme; but a short pronunciation of the e has generally obtained, and we commonly hear it flem; it is highly probable Pope pronounced it properly, where he says,

Perhaps it would not be difficult to reduce this word to analogy, as some speakers still pronounce the elong: but in the compounds of this word, as in those where gn occur, the vowel is shortened, and the pronounced as in phleg-mon, phleg-monous, phleg-matick, and phleg-magogues; though Mr. Sheridan, for no reason I can conceive, sinks the g in the last word. When these letters end a syllable not under the accent, the g is silent, but the preceding vowel is shortened: thus paradigm, parapegm diaphragm, apophthegm, are pronounced paradim, parapem, diaphram, apothem,

GH.

390. This combination, at the beginning of a word, drops the h, as in ghost, ghastly, aghast, gherkin, pronounced gost, rhyming with most, gastly, agast, guerkin; but when these letters come at the end of a word, they form some of the greatest anomalies in our language; gh, at the end of words, is generally silent, and consequently the preceding vowel or diphthong is long, as high, nigh, thigh, migh, weigh, inveigh, eugh (the obsolete way of spelling yew, a tree,) bough, dough, though, although, clough (a cliff,) plough, furlough, slough (a miry place,) through, throughout, thorough, borough, usquebaugh, pugh!

591. GH is frequently pronounced like ƒ, as laugh, laughter, cough, chough, clough (an allowance in weight,) slough (the cast skin of a snake or sore,) enough, rough, tough, trough.

392. Gh is sometimes changed into ck, as hough, shough, lough, pronounced hock, shock, lock; sometimes we hear only the g sounded, as in burgh, burgher, and burghership.

GHT.

393. Gh in this termination is always silent, as fight, night, bought, fought, &c. The only exception is draught; which, in poetry, is most frequently rhymed with caught, taught, &c.; but in prose, is so universally pronounced as if written draft, that the poetical sound of it grows uncouth, and is becoming obsolete. Draughts, the game, is also pronounced drafts. Drought, (dryness) is vulgarly pronounced drowth: it is even written so by Milton; but in this he is not to be imitated, having mistaken the analogy of this word, as well as that of height, which he spells heighth, and which is frequently so pronounced by the vulgar. See the words HEIGHT and DROUGHT.

H.

$94. This letter is no more than breathing forcibly before the succeeding vowel is pronounced. At the beginning of words, it is always sounded, except in heir, heiress, honest, honesty, honour, honourable, herb, herbage, hospital, hostler, hour, humble, humour, humorous, humorsome. Ben Johnson leaves out the h in host, and classes it in this respect with honest.

395. His always silent after r, as rhetorick, rhapsody, rheum, rheumatism, rhinoceros, rhomb, rhubarb, myrrh, catarrh, and their compounds.

36. H final, preceded by a vowel, is always silent, as ah! hah! oh! foh! sirrah, hallelujah, Messiah. 397. This letter is often sunk after w, particularly in the capital, where we do not find the least distinction of sound between while and wile, whet and wet, where and wear. Trifling as this difference may appear at first sight, it tends greatly to weaken and impoverish the pronunciation, as well as sometimes to confound words of a very different meaning. The Saxons, as Dr. Lowth observes, placed the A before the w, as heat; and this is certainly its true place; for in the pronunciation of all words beginning with wh, we ought to breathe forcibly before we pronounce the w, as if the words were written hoo-at, hoo-ile, &c. and then we shall avoid that feeble, cockney pronunciation, which is so disagreeable to a correct ear.

J.

398. Jis pronounced exactly like soft g, and is perfectly uniform in its sound, except in the word hallelujah, where it is pronounced like y.

K.

399. K has exactly the sound of hard c: it is always silent before n in the same syllable, as knee, invel, knack, knight, know, knuckle, knab, knag, knap, knare, knave, knit, knock, knot, knoll.

400. It has been a custom within these twenty years to omit the k at the end of words when preceded by c. This has introduced a novelty into the language, which is that of ending a word with an unusual letter, and is not only a blemish in the face of it, but may possibly produce some irregularity in future formatives; for mimicking must be written with the k, though to mimic is without it. If we use colic as a verb, which is not uncommon, we must write colicking and colicked; and though physicking and physicked are not the most elegant words, they are not quite out of the line of formation. This omission of k, is however, too general to be counteracted, even by the authority of Johnson: but it is to be hoped it will be confined to words from the learned languages: and indeed, as there is not the same vanity of appearing learned in the Saxon as in the Latin and Greek, there is no great fear that thick and stick will lose their k, though they never had it in the original.

L.

401. Ben Johnson says L melteth in the sounding, and is therefore called a liquid. This, however, cannot be the reason that r is called a liquid: for no two letters can, in this respect, be more opposite. See No. 21.

Lis mute in almond, calf, half, calve, halve, chaldron, falcon, folk, yolk (better written yelk with the Bounded,) fulfil, halser, malmsey, salmon, salve, talbot (a species of dog.) See SALVE. 402. Lis mute also between a and k in the same syllable, as balk, chalk, talk, stalk, walk.

403. L is silent likewise between a and m in the same syllable, as alms, balm, calm, palm, psalm,

qualm, shalm; but when the m is detached from the I by commencing another syllable, the becomes audible. Thus, though the is mute in psalm, it is always heard in psal-mist, psal-mody, and pal-mistry; but in balmy and palmy, where they is an adjective termination of our own, no alteration is made i In the sound of the substantive which sinks the (386.) Calmer and calmest ought to have the mute, as they are only degrees of comparison; and palmer, and palmerworm (except in the language of Scripture, where the 7 in palmerworm ought to be heard,) are only a sort of verbal nouns, which never alter the sound of the original word, and therefore ought to have the mute. But though is sometimes mute in the noun salve, and in the verb to salve, it is always heard in salver (a kind of ti plate.) See SALVE.

404. Lought always to be suppressed in the auxiliary verbs would, could, should: it is sometimes suppressed in fauit; but this suppression is become vulgar, (see the word.) In soldier, likewise, the is sometimes suppressed, and the word pronounced so-jer; but this is far from being the most correct pronunciation: ought alv ays to be heard in this word, and its compounds soldierly, soldiership, &c.

405. L, preceded by a mute, and followed by e, in a final syllable, has an imperfect sound, which does not do much honour to our language. The 1, in this situation, is neither sounded like el nor le, but the e final is suppressed, and the preceding mute articulates the 7, without either a preceding or a succeeding vowel; so that this sound may be called a monster in Grammar-a syllable without a vowel! This will easily be perceived in the words able, table, circle, &c. which are pronounced as if written abl, tabl, circi, &c. and in those still more Gothick and uncouth abbreviated participial terminations, peopled, bridled, saddled, trifles, gaffies, &c. pronounced pee-pi'd, bri-dl'd, sad-dľ'd, triflz, gaf-flz, &c. (359) 472.)

406. This letter has not only, like f and s, the privilege of doubling itself at the end of a word, but it has an exclusive privilege of being double where they remain single; though by what right cannot well be conceived. Thus cording to the general rule, when a verb ends in a single consonant, preceded by a sing el, and the accent is on the last syllable, the consonant is doubled when a participial termination is added, as abet, abetting, beg, begging, be gin, beginning, &c. but when the accent is not on the last syllable of the verb, the consonant remains single, as suffered, suffering, benefiting, &c. but the 7 is doubled, whether the accent be on the last syllable or not, as duelling, levelling, victualling, travelling, traveller, &c. This gross irregularity, however, would not have been taken notice of in this place, if it had not suggested an absurdity in pronuncia tion, occasioned by the omission of . Though the latter is useless in traveller, victualler, &c. it is not so in controller: for as ll is a mark of the deep broad sound of a in ball, tall, all, &c. (84;) so the same letters are the sign of the long open sound of o in boll (a round stalk of a plant,) to joll, noll (the head,) knoll (a little hill,) poll, clodpoll, roll, scroll, droll, troll, stroll, toll: for which reason, leaving out one l in bethral, catcal, miscal, overful, forestal, reinstal, downful, withal, control, and unrol, as we find them in Johnson's Dictionary, is an omission of the utmost importance to the sound of the words; for as the pronunciation sometimes alters the spelling, so the spelling sometimes alters the pronunciation.* Accordingly we find some speakers, chiefly the natives of Ireland, inclined to give the a its middle sound, to words commencing with al, followed by another consonant, because they do not see the ll in the all with which these words are compounded: thus we sometimes hear Almighty, albeit, so pronounced as to make their first syllable rhyme with the first of al-ley, val-ley; and extol is pronounced by the Scotch so as to rhyme with coal; and with just as much reason as we pronounce control in the same manner. For though compounds may, in some cases, be allowed to drop such letters of their simples, as either are not necessary to the sound, as in Christmas; or might possibly lead to a wrong one, as in Reconcileable (which see;) yet where, by omitting a letter, the sound may be altered, the omission is pernicious and absurd. (4.) The same observations might be extended to the numerous termination full, where, in compounds, one / is omitted, though nothing can be more certain, than that ful, with a single l, has not the same sound as when this letter is doubled; for who could suppose, without being used to the absurdity, that fulfil should stand for fulfill: but this abbreviation is too inveterate and extensive to afford any hope, that the great arbiters of orthography, the printers, will ever submit to the additional trouble of putting another l.

M.

407. M preserves its sound in every word, except comptroller; compt and accompt are now universally written as they are pronounced count and account; and though m and p are preserved to the eve in the officer called a "comptroller the word is pronounced exactly like the noun controller, one who controls.

N.

403. N has two sounas; the one simp e and pure, as in man, net, &c.; the other compounded and mixed, as in hang, thunk, &c. The latter sound is heard when it is followed by the sharp or flat guttural mutes g hard, or k; or its representatives e hard, qu or x: but it may be observed, that so prone is our language to the flat mutes, that when n is followed by k, or its representatives, the flat inute g scems interposed between them: thus thank, banquet, anxious, are pronounced as if written, not than-k, ban-quet, an-xious, but thangk, bangquet, angkshus. But this coalition of the sound of n and g, or hard c, is only when the accent is on them; for when the g or hard c articulates the accented syllable, then becomes pure: thus, though congress and congregate are pronounced as if written conggress, and cong-gregate, yet the first syllable of congratulate, and congressive, ought to be pronounced without the ringing sound of n, and exactly like the same syllable in contrary. The same difference may be observed in the words conscourse and concur; the first word, which has the accent on the first syllable, is pronounced as if written cong-course: and the last, which has the accent on the second syllable with a pure. It must, however, be carefully observed, that the secondary accent has the same power of melting the n into the succeeding hard g or c as the primary (522;) thus congregation and concremation have the first syllable pronounced as if written cong.

409. It may, perhaps, be worthy of notice, that when n is followed by k, the k has a finished or complete sound, as in link, think, &c.; but when n is followed by hard g, the g has an unfinished or

* This omission of the letter L, I see, has been rectified in the last quarto edition of Johnson's Dictionary; and it would have been well if the Editors had acknowledged their obligations and extended their emendations to the work Bodle, and several others.