words to their original state, as plorare and pluere, from llorar and llover; or stagnum and spica, from estanque and espiga. This is one of the most easy, and yet most important points to acquire in the study of languages, as after having detected the several disguises, the affinity of the different phraseology becomes such, that the memory is materially assisted, and that some appear to be but dialects of each other; as is observable in the Hebrew, the Chaldee, and the Syriac; the Latin and its coguate tongues; the English and the Saxon.

I

7. Some languages have the peculiarity of shortening their derivalives. Thus the French and Portuguese say seul, so, alone; voir, ver, to see; lire, ler, to read; mal, ma, bad; nu, nu, naked. It is also a peculiarity of this latter language, that it strikes off, or digammates consonants in the middle of words, as in coroa, a crown ; cea, a supper; fea, foul; nao, a ship; voar, to fly, and bateis, boats, The Portuguese also, in subservience to the genius of their language, turn the l into r, as in cravo, a nail; pranto, (planctus,) weeping, and praya, a coast. The Cornish, unable to express the initial v, compensated it by gu, as in guer, true; guenuyn, poison, and guest, a garment; or they articulated the English ch by Tsh, as in Tshappal, a chapel; or the v again by th as in seithyn, a week, and ethen, a bird, from septem and avis. They also assume the aspirate for the loss of the Latin s in Helik, a willow, and Huigeren, a father-in-law, from salix and socer, thus returning to the Greek undigammated words in eλ and exupós. The s, the English y, and the aspirate are nearly allied, and it is but according to the nature of the sound, that they should be often corrupted and substituted for each other. The Cornish seem also sometimes to pronounce the w like the Welsh, as in lew, a lion, and tarw, a bull. It is, in shape as in sound, nothing but the Greek w. The substitutions of the several other letters in Cornish are almost endless, and will be better understood from the extracts I have already given from the vocabulary, than from any more detailed account in this place.

8. There are sometimes sounds which, in the course of permutation into their own language, foreigners cannot pronounce; and which instead of corrupting, they entirely omit. Thus the Italiau c and the English ch have no correspondent in French, when we say cerass, cherry, cerise; nor have the Italians a sound like ch in charbon, The English combines the three sounds of c in chess, card, and censor.

9. The disguise of words is almost infinite, and cannot be reduced to any general rule. Some words are disguised to an extent, that could not have even been conjectured. Many of these instances, however, can admit of no doubt; and I think that many of those who use

This, together with the diphthong ou, and the soft e with a cedilla, seem to prove an affinity between the French and the Portuguese, which must be as ancient as Count Henry, a Burgundian prince, who with a body of military adventurers, founded the latter monarchy about the end of the eleventh century.

2 Thus vulgarly yute for gate.

either sherry or jalap do not know that they are indebted for both to the districts of Xeres and Xalapa. The mistake arises from the Spanish a, which is a guttural, and pronounced something like our sk; and we have expressed that a by a similarly sounding symbol in our language. It is impossible that the disguise of words should be any where more striking than in religious appellatives. Thus from,

we have in English, Jew; Juif, French; Giudeo, It., and Cornisli Jedhewon. It would be tedious to follow the same, if not greater, variations, in the words ἐπίσκοπος, ἐκκλησία, μοναχός, δε

10. The disguise of words is not confined to their spelling, but also extends to their meaning, though this latter is much less subject to variation than mere sounds. Most words impart their original signification to their derivatives; but still the exceptions are numerous, and therefore one cannot be too cautious in the interpretation of one language by another. Thus copy, evidence, reflect, repair, and supply, convey very different ideas in English from what they do in Latin. Some of the German Biblical commentators have fallen into this mistake, when they met with difficult words, or with such as occur but once in the sacred text, and had recourse to the synonyms in Arabic, and advanced interpretations from it, which are at variance with the most valuable translations.

11. Though it is one of the essential properties of languages, that their pronunciation should be different, yet sometimes they have features of resemblance in that respect, and where it might have been the least expected. Particular sounds in one language thus become common to another, as has been already observed in the Italian c and the English ch, and it is not improbable, that the r so frequent in Portuguese, is the Hebrew derived to it from the Moors. The most striking feature of the kind, kowever, is the aversion of the Spanish language from the letter fƒ, a circumstance well known to every scholar, as being common to the Greek in the disappearance of its digamma. Possibly the Spaniards experience the same difficulty in pronouncing that letter, which the Greeks did; and if so, it must have been from their long familiarity to the same sounds, and incapability of uttering any other. Be it as it may, the coincidence is most complete. Thus we have hacer, to do; hijo, a fig; hoja, a leaf; honda, a sling; horca, a fork; hosco, dark; huir, to fly; humo, smoke; hierro, iron, and several others which are derived from the Latin f. Those words which begin with two consonants retain the ƒ; as in ficgrar, to sparkle; frangir, to break; and this is also observable in the Greek as in ppioow, to tremble, and ppovéw, to be wise. From

By a most strange perversion we have bottega from anothen, which the Italians borrowed from the Greeks; called also by the French boutique; and as men in similar circumstances have recourse to the same means, the Americans call a shop, a store; and the Spaniards give the name of lavaderos to the places, where the gold dust is collected, like the Cornish miners, who call the corresponding places for tin, stream works.

filum we have hilo and filo; and viùs, after having been disfigured in filius and all its derivatives, is at length found again in Spanish in almost its primitive form in hij'o.

12. Words are still more highly disfigured, when adopted in the dialects of a barbarous age. Under some of the foregoing heads, I have considered words as reported by voyagers, who accommodated them as much as they could to sounds in their own language. Little dependence can be placed on their accuracy. But how much more inaccurate must be the derivatives, which are found in the modern languages! The influx of rude and barbarous nations into the Roman empire corrupted the Latin; or, to speak more accurately, it began to be pronounced according to the particular accent of the invaders. This change of pronunciation necessarily created a disguise which from its combination with continual solecisms produced a new dialect. This new production for a long time was despised, and being uneither committed to writing, nor having any other fixed standard, became subject to still greater vicissitudes. Words, which were at first but slightly altered, at length became so disguised, as to lose every type of their original resemblance. While they borrowed from foreign languages, the vulgar did it in the most ignorant way, so as to be even ridiculous, as in the Spaniards mistaking the Arabic article for a part of the word itself, in nearly all they took from the Moors, as in algodon, cotton; alcalde, a magistrate; alchymia, alchymy, &c. Thus they wrote a solecism and a disguise in the very same word. After these alterations of the common people, another important change still remained, when the language began to be cultivated. With a view to be polished, and reduced to a grammatical system, the words still underwent a much greater aberration from their roots. All these processes are so many different steps, which account for a more considerable corruption than when words are reported according to the ear of a traveller, or when common use transplants them from a living tongue, retaining the orthography, if not the pronunciation. The modern languages, French and Italian, were in their infancy much less disguised in their Latin derivatives, than they are at present. Petrarch spelled much nearer Latin than the modern Italians; and the French have since dropped many unpronounced letters, though some are still retained in the plural of their verbs. The pronunciation may also be supposed to have participated in the deterioration of orthography; and what was still articulated with a Roman accent in the fifth century, gradually departed from it, so as to leave us no doubt, that the former is not less corrupted than the latter.

13. The reverse of the above is true with respect to words, which are but of late introduction into modern languages. They are indeed the words which are the least disguised, as they labor under the disadvantage neither of having descended to us through distant ages, nor of having been imported from unwritten dialects by the deceitful ear of travellers. Of this number may be reckoned the Greek names, which have been adopted by the moderns to designate particular arts and sciences. Thus polity, philosophy, physic, &c. have been altered

in nothing but the termination; and the same rule also holds good with respect to the Latin words which have been lately naturalised among us, and which eminent authors have recommended for the sake of elegance and energy, as in to concede, to interpolate and to lubricate; and in conciliation, detrusion, obliquity, and recrimination. From this we may infer that it is by foreign conquests, and in barbarous ages, that languages become corrupted; and that on the other hand, whatever learned or fixed languages borrow from each other, is but comparatively little altered in the transition.

14. Languages either add or take away letters from words for the sake of softening the pronunciation, or to be adapted to the national idiom. The theory of the Greek digamma, about which so much has been written, and which, it must be allowed, was a fortunate discovery, amounts to no more than this definition. It is in fact not peculiar to the Greek; but traces of it may be discovered in all languages; and though it may chiefly affect the ƒ and v, yet it is also sometimes applied to other letters. I take the digamma to be nothing more than a suppressed consonant, whatever it may be. A very short discussion will render this evident. The Greek, like other smooth languages, dropped harsh or sibilant letters, for the sake of a concourse of vowels, as in wor and cos, while the Latin retained the primitive forms in ovum and sol. The Greek words, as they now are, have been, if the expression can be allowed, truncated and smoothed down. The reason is plainly this, that finding it difficult and unmusical to articulate particular letters, the Greeks either removed them, or sometimes compensated them by an aspirate, as in olvos, and èorépa, Hesper and Vesper. Even the k may be a substitute of the Digamma, as the Hebrew ; Greek, képas; Latin, cornu; English, horn. From the Hebrew, the Greeks have digammated their λos, and the English their whole; and from exaoros, it is thus that we have each. The Cornish also has brohal' and brochal, a sleeve; and carhar and carchar, a prison. Next to the ƒ and v, the s seems to have been oftenest struck out by the Greeks, as in ciui, sum, I am; èkupòs, socer; р, sеrpo; aλs, sal; e, six; éπrà, seven, &c. Again, idw, takes the v in the Latin video, and the s, as in the English, to see; or as inu, to send. The Greeks do not always reject the y, since they have retained it in yata, and not in aia. The t and the d are also signs of the digamma, since the Hebrews say, the Greeks Epa, and the English, earth; instead of what the Latin and its cognates have expressed by terra, and the Cornish by dôr and tyr. The

2

Thus the Hebrew is expressed in Roman characters by hand ch, as NN is in the Septuagint 'Axàß, and in our English translation, Akab,

2 Umbra resumes its digamma in the Spanish sombra. The following word has taken at different times the three cognate digammatic letters, s, and h, oúxor, ficus, and higo. Andalusia has lost the digamma from Van dalusia. The Indus is now the Sinde.

Hebrew in it, the grave, disappears in its English derivate hell, The Romans unable to pronounce the harsh and guttural in

he made, approximated to it in facio, which from the genius of the Spanish language, is restored to something like its original form in hacer. I have already observed that vios returns at length to hij'o; but in Latin the digamma is expressed in that very word by f and l. The Italians and the Portuguese, in imitation of the Greeks, have dropped the I for the sake of a more pleasing concourse of vowels, as the poetic plurals of the former have augeil, birds; cavai, horses; strai, arrows; for augelli, cavalli, and strali; while the latter have in the plural, baris, barrels; and bateis, boats, instead of bariles, bateles. Naus after having retained its digamma in Latin, resumes its Grecian softness in the Spanish nao; and I apprehend also that alua is a digammated form of the Hebrew DŢ, blood.

K

I have just observed that the y is sometimes the letter restored in Greek for the digamma, as in yata; and the parallel also holds out through the modern languages, as yéveaus gives us in Cornish henath; August in French is Août;' Germanus makes in Spanish hermano, and in Portuguese irmam, a brother. The transitions of ei are remarkable; in Latin eo, in Italian, pres. ro, 3d of the perfect poetic, gio, inf. gire and ire in English go; in Spanish ir, and in Portuguese hir; all of which, varied as they are, leave no room to doubt of their common origin. The Greeks generally omitted the y, for which they had no equivalent, as in 'Hoau, and 'ny; except in a few proper names, in which they expressed it by y, as in láza, feßáλ; but it is not so commonly known, that they had before employed the x for the same purpose, as from y, they made képaž, from which have sprung the several derivatives of corvus, corvo, corbeau, and crow. Some Portuguese words want the n to be restored to their digammated form, as in pessoa, a person; some the d, as nua and erua; some the final vowel, as acçoens and varoens, heroes; and some even substitute nothing for a vowel, but an aspirate, as folho, a leaf, and molher, a woman; or after losing a syllable and its consonant they contract it into a circumflex monosyllable, as iér, a color, dór, grief, and mór, greater. I should therefore be induced to conclude from these observations, that the digamma is nothing more than a particular disguise of words, and that there is perhaps no language from which instances of it might not be selected. It is also evident that the modern languages have had largely recourse to the expedient of truncating, or digammating their words like the Greeks, and that in that

1 Voltaire retains the Latin form, and writes it Auguste.

2 It has frequently been observed that stammerers find a particular diffienlty in pronouncing the s and k. Such persons therefore are exactly in the situation of the Greeks and Hebrews, who digammated words beginning with those letters owing to an imperfect pronunciation. The e in the French reeif, is lost in our English reef.