consonant V. The only evidence produced to support this opinion, as applied to the Greek letter, seems to be the practice of the modern Greeks, who pronounce Sɛus, zefs; Bartheus, vasilefs; avtap, aftar, &c. But it is too well known to require a moment's illustration, that the pronunciation of the modern Greeks is in many particulars utterly abhorrent from that of their progenitors, and affords no grounds for any conclusion respecting the sounds employed in the ancient language. Numerous proofs might be accumulated to shew, that T, whether used singly or to compose the diphthongs av and ev, had the full force of a vowel; and that its sound (at least in the classical ages of Greece) was intermediate between that of I and that of OY, corresponding pretty exactly to the French u. Hence it was called inov, slender, to distinguish it, (not, as Dr. Marsh supposes, from a consonant answering to our English F,) but from the broader and stronger sound of the cognate OT, which, though represented by two letters, was in its own nature a simple vowel.

Equally destitute of foundation is the supposition, that the sound of the English consonant V belonged to the Latin V, when used as a consonant. V, when employed as a long vowel, answered to the Greek OT, and consequently expressed the sound which is denoted in French by ou, and in English by oo. Abbreviate this sound, so as to reduce it, in point of time or metre, to the power of a consonant, and it coincides entirely with that of the English W. Until, therefore, some good reasons are assigned to the contrary -some better reasons than the real or fancied " coarseness of the enunciation-we ought to regard this as the power of the Latin V when used as a consonant, rather than ascribe to it another sound, which cannot possibly be deduced from its known and admitted signification as a long vowel. This conclusion is further justified by the changes upon the power of this letter, practised by the Latin poets in order to suit the rhythm of their compositions. Thus Virgil, speaking of one of his pugilistic heroes, says, "Genua labant," converting the short vowel V into a consonant, so as to obtain a trochee out of a tribrachys. Who can doubt that in this, and in similar cases, the sound of the consonant V was that of the English w, which is only a rapid utterance of its vowel sound, and not that of the English v, which has no connection with the vowel sound? We might establish a similar argument upon the mode of forming various derivatives from verbs ending in VO, in which V is some

times a vowel, as in FLVO, TRIBVO, and sometimes a consonant, as in VOLVO, but which agree in conferring upon their derivatives the full force of the long vowel V, as in FLVMEN, VOLVMEN, &c. It would also be an amusing subject of research, to collect from those European and Asiatic languages which are of the same family with the Latin, the forms of words which correspond to Latin forms of the same words beginning with the consonant V. We should probably find that, in the vast majority of cases, these words begin with the sound of the English W. We present a few examples from the English language: vallum, wall; vasto, waste; ventus, wind; vicus, wick; vidua, widow; vinum, wine; volvo, wheel: and, if these instances be opposed by such words as virtue from virtus, revolve from volvo, &c. it is sufficient to reply, that such terms have been formed from the corrupted pronunciation of Latin in modern times; whereas those beginning with W were derived from the Latin of the classical ages, or are original words belonging, under different forms, to various languages which are allied to the Latin. Lastly, the opinion for which we contend, is supported by the examples of all Latin words transferred into the Greek language, such as Zεounрos, and Zεunpos, for Severus, in which the Latin consonant V is represented, imperfectly indeed, but as closely as the case would admit, by the Greek OT or T; and since the sounds of these two characters, when uttered with the rapidity of consonants, are found to coincide, all these examples must be regarded as conspiring to the same point, and therefore, when taken together, amply sufficient to overbalance the examples produced on the other side of the question, in which the Latin V is expressed yet more imperfectly by the Greek B.

The preceding observations tend to shew, that the digamma, when used as a consonant, had sometimes the sound of the English W. Such was the case in all those words where the insertion of the digamma restores the versification of the poems of Homer, and where also the corre sponding Latin forms of the same words begin with V. But it is probable that the ancient Greek F was often, though not, as Dr. Marsh contends, invariably equivalent to the Latin F and the more recently invented .

It will be allowed, that the source of the digamma, as well as of all those letters which originally belonged to the Greek alphabet, must be found in the Phenician or Hebrew. But in the ancient oriental alphabets, the digamma, there

called Vau, and occupying the sixth place, was probably employed both as a vowel and as a consonant, being analogous sometimes to the Greek OT, and sometimes to the Greek . This conjecture we might support, by producing instances of the manner in which the power of the sixth letter has been expressed by the ancient translators of the Hebrew Scriptures; but we cannot suppose that the oriental tongues wanted either of the two sounds which we have pointed out, for they are found with slight modifications in every known language. Now, if they existed in those tongues, we cannot doubt that they were both expressed by the sixth letter Vau, because the alphabets referred to afforded no other probable means of representing them. If this argument be valid, we must upon the same principle (as well as from the entire analogy of the ancient Greek alphabet to the Hebrew or Phoenician) ascribe the same double power to the digamma; for this letter was in established use among the Greeks before they had invented either T to denote the one sound, or to represent the other. The use of the two last letters was probably adopted to obviate the inconvenience and ambiguity arising from the double office of the digamma. This character then fell into disuse, except as a numeral.

Among the Latins, who adopted the Greek or Pelasgic alphabet after it had received the addition of the vowel T, the digamma always retained the office of a letter, but was confined to the expression of its original power as a consonant. Hence those Greek words which were afterwards written with, such as AMA, TгEN, were expressed in Latin by F, as we see in FAMA, FVGIO. Nevertheless, in Etruscan inscriptions we find F still used, as it was by the Greeks, to denote the sound which was expressed in Latin by V having the time of a consonant.*

For the reasons now stated, we agree with Dr. Marsh in thinking, that the most ancient Greeks used the digamma in all cases in which was afterwards employed, and in which the digamma is still found in the Latin forms of the same words. Dr. Marsh also adduces satisfactory evidence to prove, that this letter was prefixed to those Greek words which now begin with an aspirated 'P, and which correspond to Latin words beginning with FR. Trypho, who

* Observe, that in the above account of the original double power of the digamma, we have considered the sound of the English W, as included ander its power as a vowel,

lived in the third century, asserts, that in the poems of Alcæus he had found the word puts written Fps. We now find in the Elean inscription Fρατρα for ῥητρα. Gre gorius informs us, in his treatise on the Dialects of the Greek Language, that the Eolians began such words with the analogous letter B; and as the sound of F, when used as a consonant, includes an aspiration, we are thus enabled to account, in a manner which confirms the hypothesis, for the existence of the aspirate still prefixed to such words. Upon this ground, we are enabled to judge how the aspirate ought to be expressed when it accompanies an initial rho. The sound ph, with which the words under review originally commenced, was an aspiration, modified by pressing the teeth upon the under lip. Omit this pressure of the teeth, and a simple aspiration remains, which however still precedes the initial P, and justifies the position of the spiritus before it, though sometimes represented as if it were sounded after. Thus the ancient FRAKTOΣ, corresponding to the Latin FRACTVS, became in the more recent and classical Greek ῬΑΚΤΟΣ, or ῥηκτος, pronounced Hractos.

When we consider the age, learning, and station of the celebrated author, upon whom we have been animadverting, and call to mind our obligations to him for some of the former productions of his pen, we should think ourselves chargeable with a high degree of ingratitude, as well as presumption, were we to blame him, even where censure is due, without great delicacy and respect. We therefore sincerely lament the deficiencies of the volume before us. His letters to Archdeacon Travis, and his Annotations on Michaëlis contain the clear and abundant proofs, not only of most extensive reading, but of independence and energy of mind, of sound judgment, of candour, accuracy, and ingenuity. The present treatise contains erroneous representations of the classics; assertions strong and positive in proportion as the evidence for them is slender and insignificant; inferences, dependant in trains upon one another, and all of dubious validity, yet ending in conclusions, which are produced as absolutely certain; and respectful appeals to the authority of writers, who, whenever they oppose his doctrines, are treated as undeserving of credit or regard. These circumstances indicate a deterioration of the understanding. Let the author, ere it is too late, beware how he devotes himself to the service of a party. Let the high and honourable reputation, obtained

by his former learned publications; let a superiority to the world, and a noble, ingenuous, and uncontaminated desire for the attainment and progress of Truth; let the memory of his venerable predecessor, and a strong sense of the awful responsibility attached to the episcopal dignity; in short, let upright principles and elevated affections influence him to employ his talents and opportunities in the promotion of those objects, which alone can appear truly important and valuable to the philosopher and the christian, and we may even yet have cause to number him among the greatest benefactors to his church, his country, and mankind.

ART. III.-An Account of the Natives of the Tonga Islands, in the South Pacific Ocean; with an original Grammar and Vocabulary of their Language. Compiled and arranged from the extensive Communications of Mr. Wм. MARINER, several years resident in those Islands; by JOHN MARTIN, M. D. 2 vols. royal 8vo. pp. 460–412. London, Murray, 1817.

To no single individual is the science of geography more indebted than to the late Captain Cook, who fell a sacrifice to the ignorance or ferocity of the barbarous regions he explored. In his first voyage the Society Isles were discovered by him; the insularity of New Zealand was ascertained, when the streights which separated the two component parts were distinguished by his name: and in the same voyage he explored the coast of New Holland through an extent of two thousand miles. In his second voyage he was enabled to negative the conjecture with regard to a southern continent within the reach of navigation; he added New Caledonia to our charts, the largest island in the South Pacific, New Zealand excepted; and also Georgia, in the latitude of Cape Horn, with an unknown coast that he called Sandwich Land, and which has been denominated the Ultima Thule of the southern hemisphere.

In his third voyage he revisited the Friendly Islands, discovered several smaller clusters on the tropic of Capricorn, and the Sandwich Islands to the north of the equinoctial line; he explored the western coast of America from 48 to 70 degrees of north latitude; he determined the proximity of Asia to that continent; and passing the streights between them, demonstrated the impracticability of a northern passage from the Atlantic to the Pacific.