most European languages, except before a double consonant, in which case it is pronounced as in ink, sing.

O, as the long English o, in mole, roll, dole; but short before a double consonant.

U, like the sound of the English word you, or ew in few, except before a double consonant, then it is short, as in luck, tuck, suck.

Y, like the English i, in sigh, die, white; or the German ei, in mein, sein; but somewhat more short and sudden, perhaps rather like the English y in ally, apply. The same sound, when long, is expressed by the following diphthong: AI, like the long English i, in dine, mine, whine. It is a true diphthong, generated from the coalition of a and i, as above defined; the first being heavy, the second light.

AO. This is a sound distinctly of two syllables, the a and the o being pronounced as defined in their respective places. AOO. This is a word of two syllables, the a and the oo being pronounced as defined in their places.

AU. This is a word of one syllable, pronounced like ow (which see below), but somewhat longer, and yet in such a way as not to run into two syllables like doo, (which see above). Thus >these four sounds, viz. ao, aoo, au, and ow, are very nearly similar, except to a good ear; and yet it is necessary they should be thus distinguished, for three of them are words of very different meanings, aoo, a cloud; au, the personal pronoun I, and ow, the possessive pronoun, thy. The same may be said of fao, a peg; faoo, load; and fow, a turban: also, tao, spear; taoo, to cook victuals under ground, and tow, war. AW, as in law, saw, paw. OW, as in how, allow, now.

When the w is preceded by a or o, it must be joined in sound with it, not with any vowel or aspirate that may follow; as, fawha is to be pronounced faw-ha, not faw-wha; for in point of fact, the w is not in such instances a separate letter, it only serves to give a peculiar sound, as above defined, to the a or the o. Ow is to be pronounced short, else it will run into the sound of au, which see above. OY, as in the English words, toy, coy, &c. OO, like the Italian or Spanish u, or like oo in tool, cool, &c., unless one of them be marked thus, ö, in which case they are pronounced distinctly. When oo is accented, the accent is uniformly upon the latter, thus oó. When three o's come to

gether, the two which are not marked thus, ö, constitute the diphthong; or if the first of the three is accented thus, 600, the two latter are the diphthong.

As to the consonants, the following only need be particularly mentioned.

Bis sounded between the b and p, but it has more the sound

of the b.

C, before a and o, is hard, and partakes in like manner of a little of the sound of the hard g. It never occurs before e and i. To express the sound of the soft c, the letter s is always used. D. The sound of this letter is scarcely known in the Tonga language. There is, indeed, a sound approaching to it, but it is only the careless conversational way of pronouncing the t. G. This letter is always hard, as in game, gill (of a fish), begin, &c. Wherever the sound of the English soft g occurs, j is always used to express it.

J. This letter has not, however, exactly the sound of the English, but between j and %, so that if the j in our English word jest were thus pronounced, it would sound between jest and zest, and not very much unlike chest.

CH, is always sounded as in the English words choose, change, chair, &c.

ER. The r is never pronounced strongly. When it follows e it is scarcely sounded, giving merely a power to the e similar to what it has in the French words, le, me, te.

NG, as in the words among, song, wrong. But this sound is not to be intimately joined with the following vowel. For instance, Tonga is not to be pronounced Tong-ga, as the English reader is apt to do, but thus, Tong-a.

Gn.

Here the g is not sounded strongly, but somewhat more so than in the word gnomon.

Upon the subject of accentuation, I must observe, as a general rule (to avoid the too frequent and unnecessary use of typographical accents), that in words of two syllables the emphasis is to be laid on the first; in words of three syllables, it is to be placed on the middle one; and in words of four syllables, on the first and third. In all exceptions to this rule, and in some of those instances where vowels themselves form syllables, typographical accents will be placed accordingly, It is to be noticed, that in words of three syllables, when the first only is ac cented, that the two others are light ones, and that the accent is always put upon the vowel of the emphatic syllable,

The Tonga language may be divided, like most others, into eight or nine parts of speech, and if nicer distinctions could aid in rendering the subject more clear, two or three might be added. But we apprehend that the usual number will be found quite as many as will suffice for our purpose: and as these are not always properly defined, the noun, adjective, verb, and par ticiple, being often one and the same word, distinguished only

by the general sense of the phrase, and sometimes scarcely by that, we might be disposed to lessen rather than to increase the quantity. In respect to those parts of speech which might be superadded, they consist of a peculiar particle used before the article, noun, adjective, and pronoun, according to certain rules, signs of the plural number, signs of the tenses of verbs, and two or three others, whose uses cannot be explained in a few words. Of these we shall treat under the different parts of speech to which they are generally attached, or to which they seem most referable.

The following, then, are nine divisions of speech, which we shall adopt in the present investigation, and of which we shall discourse in this order

Article, Noun, Adjective, Pronoun, Verb, Adverb, Conjunction, Preposition, Interjection.

THE ARTICLE.

We find in this language a very frequent use of these three particles, viz. he, co, and coe. On a strict investigation, how. ever, we find that only the first, he, can properly be called an Article, being chiefly used before nouns, and is sometimes used to distinguish them from verbs, though it is occasionally used before what in our language would be called the participle of the present tense, by which it is converted into a sort of noun. Whereas the particle co is used not only before nouns, but also proper names, to mark certain significations more distinctly. Besides which, it is frequently prefixed to pronouns. As to the

particle coe, it is only a coalition of the two former, he and co, the aspirate being dropped. Upon these considerations, I think we may venture to state, that there is but one ARTICLE in the Tonga language, viz. he. But as the particles co and coe seem to have a strong relation to it, we shall treat of them under the same head.

The article he has no distinction either of gender or number. It may be used, we believe, before all nouns, though there are some occasions where it must be dispensed with, instances of which we shall give by and by. Its use may be exemplified thus he tangata, a man; he fafine, a woman; he togi, an axe. When the conjunction mo (and) precedes it, the aspirate is generally dropt, thus, he togi, móë coola, móë papalangi, móë jiawta; axes, and beads, and cloth, and looking-glasses. The occasions where it should be entirely left out, will be best explained when illustrating the use of the particle co; and some farther observations will be made upon this subject when treating of nouns. The particle co is very frequently used before nouns, prons, and proper names. The instance where it appears most

essentially to occur, is in answer to the question who or what? and will then generally bear to be translated by the verb, it is or it was, &c. Thus, who is there? a man, co he tangata: what is that? an axe, co he togi: who was with you? a woman, co he fafine: it is a man, it is an axe, it was a woman, &c.

Before proper names it is used in like manner, the article being left out, as in answer to these questions, who did you see there? co Finow: who else did you see there? co Toobo Nuha. But if the names of a number of persons are mentioned, the particle co is only put before the first, as, who came in afterwards? co Havili, mo Mooala, mo Talo, mo Latoo, &c. (mo being a repetition of the conjunction). In like manner, it is used before the proper names of brutes and of inanimate things, as dogs, hogs, canoes, clubs, axes, &c. ; for axes formerly had proper names, on account of their extreme scarcity and consequent value; and clubs also, which have become valuable on account of having been used in killing great chiefs, or from having done much execution.

Before the names of different varieties of the same species, this particle is also used, but the article he is omitted; as, co tooa, co coumele, co caho-caho, co gnoo, all which are different kinds of yam. But when speaking of the yam in general, they would put the article he after co: as co he oof, the yam; that is, in answer to a question, as before.

Sometimes (not particularly in answer to a question) the name of the species and variety are both mentioned, as the chief Ooloo Valoo, the man Boboto; and in such cases the particle co is always used (the article he being omitted) before the proper name of the person spoken of, and often before the word expressing chief, man, &c. ; but in this latter case (i. e. without the article) it seems to show that the party spoken of is supposed to be known to the person addressed; as co egi co Ooloo Valoo, the chief Ooloo Valoo; co tangata co Boboto, the man Boboto. If, on the contrary, the chief Ooloo Valoo, or the man Boboto, is supposed to be unknown to the party addressed, then the article he, as well as the particle co, would be used before egi, or tangata; as, co he egi co Ooloo Valoo, i. e. a certain chief called Ooloo Valoo; co he tangata co Boboto, a certain man named Boboto. *

We have intimated just now that the particle co is not always

* I do not mean, however, to lay down this rule as a fixed and certain one. To Mr Mariner's perceptions of the language, it appears in general correct; but he thinks they sometimes violate it through inadvertence, and, I may add, sometimes perhaps for the sake of euphony.

used before the words egi and tangata, and we shall now point out when it is not to be used. If the above phrase, the chief Ooloo Valoo, or the man Boboto, were to occur in the latter part of a sentence, the particle co would be left out before egi or tangata, but the article he would remain: as, nai how giate au he tangata co Paloo, there came to me the man Paloo: but if the arrangement of the sentence be altered, thus, the man Paloo came to me, then both the article and particle may be prefixed to tangala, as before: as, co he tangata co Paloo nai how giate au. These two examples, however, intimate that the man Paloo is unknown to the party addressed, but if the contrary were the case, the last form of the sentence only could be used, and the article he must be left out: as, co tangata co Paloo nai how giate au, the man Paloo came to me; but nai how giale au co tangata co Paloo would not be gramatically expressed for any

sense.

In consequence of the frequent use of co before he, the two, in the rapidity of speech, are coalesced into one, the aspirate being omitted; thus, coe instead of co he, as coe tangata co Boboto. We have hitherto expressed them separately for the sake of clearness, but shall henceforth write coe, according to the strictest pronunciation; for co he tangata would not sound very well in the ears of a Tonga chief who took pains to pronounce his language correctly.

THE NOUN,

The noun has, properly speaking, neither gender nor number: i. e. the gender is distinguished neither by any peculiarity in the word, nor by any sign; and the number is only distinguish ed sometimes by a sign, or by some other word of singular or plural signification; but the use of this prefixed sign or word will depend upon whether the noun be significant of an animate or inanimate nature; if of an animate nature, it will depend upon whether it be a rational or irrational nature.

The singular number of inanimate beings is usually expressed by the simple noun, with the article he before it: as, he togi, an axe; he falle, a house. When it is intended to lay a particular stress upon the circumstance of their being only one, the numeral is used with the word be (only), and the article is left out: as togi be taha, axe only one; falle be taha, house only one. When a certain and fixed number of inanimate objects are meant to be expressed, the numeral is used according to the following form: togi e ooa, axes two; falle e toloo, houses three; vaca e fa, canoes four; wherein it is seen that the particle e comes between the noun and the numeral, and which in all probability the article, with the aspirate omitted, and placed in this situa