read only imperfectly; and when he has found the passages I allude to, he will learn why one set of words, expressive of one set of ideas, must be duads, while those of another set must be triads; and how the same symbol, A, may suit two languages, whose radicals are respectively all vocal, as in Greek, or all consonants, as in Hebrew; provided the vocal symbols represent ideas not the same as, but similar to, the ideas represented by the symbols not vocal. For instance, if the vowels AIQ (I hear, or I perceive, in Greek; in Latin, I say,) represent the idea of sentient matter, the consonants M T R would represent the idea of matter not-sentient.

Hence, if we consider the vowels as radicals, the derivatives will be formed by the admixture of consonants; but, if the radicals be consonants, the derivatives will be formed by the admix

ture of vowels; and thus a similarity will be shewn to exist in the principles of both the Greek and Hebrew radicals, which those, who, like Valckenaer and Ruhnken, were partly initiated into the mysteries of the theory promulgated by Hemsterhuis and Schultens, saw at once; but for which <he uninitiated want an interpreter.

Now, that Hemsterhuis had some theory, and probably a correct one, of the origin of language is most certain; else he would hardly have stuck to it for forty years, as we are told he did; but that the theory was one he did not choose to promulgate openly, is also proved by the fact, that it never was made known, except partially to his favoured disciples; one of whom was Valckenaer; and even he did not know all, or, if he did, never communicated it to his pupils; and hence all our present ignorance of the real nature of the theory in question; which, doubtless, turned upon the solution of the problem, why the symbol of the God of eloquence amongst the Greeks was the same as the symbol of the God of gardens in Italy; and what connexion there is betwen συκή, a fig, and xn, the life or soul; and why the same letters in Greek, though differently disposed, mean NOOΣ mind, and ONO an ass; and lastly, why the Greek and Latin letters pater are the same as taper in English.

Connected with the theory stated above, respecting the formation of a

new set of words, by mixing conso-
nants with vocal radicals, or by adopt-
ing what has been called the cabalistic
change of letters, we may take AQ for
an example. Thus we have ΑΔΩ,
EAQ, IA2, oaq, and yao, with or
without the aspirate, forming a new
set of symbols expressive of a new set
of ideas connected with the idea ex-
pressed by the symbol ▲, but viewed
in a different light. Thus—
AAO. I please. Hence AAON-IE, thou
art pleasing, the name of Ve-
nus' lover.

EAQ. I eat. In Latin Edo.
IAQ. I um in a cold sweat.

Hence by

metathesis die in English. OAN. I give a scent. Hence the Latin Odor.

YAQ. I am in a warm sweat. Hence, the Latin Sud-or, similar to the Greek 'Yo-wp.

But the most curious fact developed by this theory is the clue it gives us for finding out the reason why the

characteristic of the future in Greek is the letter σ, and of the perfect, an asend in abo, ebo, and ibo, and the perpirate; and why, in Latin, the futures fects in avi, uvi, and ivi.

These, however, may form subjects of future enquiry; in the mean time, I beg to present J. T., with the following diagram.

Mr. URBAN,

HAVING thus disposed (in No. V.) of any fancied argument which might be drawn from the Hebrew tongue in support of the Origines Græcæ, I proceed at once to state or recount some direct objections to the Hemsterhusian doctrine; where, in its practical etymology, it appears to me to run counter to the sensus communis of mankind.

1. The derivation of nouns (except in external form alone) from the tenses of verbs (as κριτὴς from κέκριται, &c.) has been sufficiently shown to be unnatural and absurd in the 1st and 2nd of these letters.

2. For the assumption that the same syllabic sounds (as we now have them) were by some natural necessity or scientific convention originally attached to the signification of one set of objects or notions, and of no other but those (thus identifying in origin oppos and ópun, &c.), two very strong reasons were assigned in my 3rd letter to justify the rejection of that idea.

3. The boldness of the postulatum, that the Greek language is self-derived and complete in itself, a language perfect and insulated, cannot be more strikingly exposed than by Valckenaer's origination of warp, "altor," from raw," alo," (itself an assumed verb), through the medium of πéжатαι! when the word pater, fader, vater, &c. notoriously belongs to half a dozen dialects at least, which neither could derive it from the Greek, nor had any Táш and "éжаraι of their own to father it upon.

4. Even the plausible assumption, which in all cases takes the shorter form as the primitive one, appears to me to have been allowed much too readily to pass unchallenged. Of the two simple verbs, yaw, yévw (both likely enough to have existed), why should we allow the former, and reject the latter as a primitive? Is the letter (v) necessarily and in all verbs a servile letter, because it is found so in very many? In μévw, for instance, the letter (v) is unquestionably radical, and why not in σθένω and στένω

The objection to letting that assumption pass becomes still stronger where the shorter form is itself imagined, and the longer has a bona fide existence. Thus, when the Homeric words μήδος and Μήδομαι (with μήσο

pai, &c.) are derived from μáw, which, in the sense required, is a verb purely imaginary, what possible advancement in sound knowledge can be gained by such a conceit ?

5. This very strange fondness for extreme simplification in Greek etymology, may be traced at least as far back as the Etymologicon Magnum. The wrong, or, if it be so, the right of that doctrine, therefore, must be referred, in its elementary notions, to an early beginning. But in its most extravagant shape, perhaps, it may be discovered in the Familie Etymologica of DAMM. The Lexicon Homericum itself, first published (1765) at Berlin, is now, by the Glasgow reprint of it, at any one's command; and by a most judicious arrangement of its (generally speaking) valuable contents, all the Homeric words are now to be found classed in alphabetical order, with the Familie Etymologice placed at the end.

I have repeatedly had occasion to shock the rationality (as to me it seemed) of my reader, by proposing for his entertainment, if not for his belief, propositions of the most eccentric description. Let me be forgiven once more, if I say that I hardly expect to be credited in the following statement.

Dammius, then, on a rough calculation, makes the whole materia of Homeric Greek to consist of 160 radical words; of which 120 are verbs, with 40 words alone of any other part of speech. In that list of 40, three animals only occur as having radical names, βοῦς, μῆλον, διs, and varying strangely in the number of their derivative attendants ; βοῦς and μῆλον, each with a tail, the first of 47, the second of 15 followers; poor is with not a tail to follow him but his own.

This surely is to tax the faculties of belief with a vengeance! And yet such is the naked truth, without disguise and without exaggeration !

Every effect must have had its adequate cause; and therefore some account may be expected of the birth and parentage of the Familia Etymologica aforesaid. I confess myself at present not much inclined, or it may be not much qualified, to meet that demand.

The grossest absurdities, however,

of that etymology may well deserve to be noticed.

And, first, the analysis of Greek words, when it proceeds till it reduces them very often into such very gaseous forms, as a, a, ó, &c. or even aw, ow, &c. little short of non-entity, is best consigned to the vis medicatrix of common sense, which, at the first approach, rejects what is so offensively offered.

But, secondly, the gratuitous assumption operating in the predominance of those 120 verbs amidst 160 radical words, may be discerned in a moment. The principle so boldly assumed can be neither more nor less than this; that the verb naturally is the parent of the noun. And this, be

yond a doubt, seems to have been long regarded as the alte terminus hærens in the foundation of languages; most probably indeed taken on credit, from the common mode of representing the Hebrew in particular.

Yet what says our Professor Lee in his Hebrew Grammar? He strongly inclines to believe (2nd edit. pp. 74,5.) that the NOUN rather ought to be considered as the root, than the VERB; and particularly on this account, that a verb in the state of conjugation, either is or must be considered as compounded with a pronoun, and therefore in a state unfit to be taken for a primitive word.

Luckily for our instruction, besides, the Anglo-Saxon yet in daily use amongst us present many decads of words, if not some hundreds, which, radically the same in idea, are with us called nouns or verbs, just according to their use in construction with other words. Fear, hopé, dislike, hate, love, with dust, water, fire, sand, oil, &c. are obvious examples both for substances and notions; which may suffice for the present.

Much more luckily, however, the Latin language, without ever exhibiting the word in its crude state, as ours frequently does, shows it so distinctly in grammatical combination, that we at once perceive curo to be the VERB I think, and cura to be the noun thought, grammatically personified. Need I refer to servus, slave-нE, and serva, slave-SHE, persons in real distinction of sex, for the purpose of illustrating names personified with gender, such as pugna and ludus?

The plain truth lies in a small compass. Many words having the form

of nouns, especially such as denote state, action, thought, feeling, &c. by their composite or derivative character naturally carry the mind back to certain verbs in their elements more simple; so that no doubt or difficulty can exist as to the commodious referring of those nouns to those verbs for the purpose of grammatical arrangement, v. g. ȧpornp, curatio, &c.

On the other hand, many names of plants, animals, substances, &c. appear before us with such strong certificates in their physiognomy for an original appellation of their own; that, though perhaps some few may be fairly traced to a verbal root, the rest may be justly left in possession of their own title, as being the oldest of the family. Such words, apparently, are pódov, yépavos, Xilos, &c.

6. But when the Hemsterhusian doctrine comes to be applied, as Lennep on a large scale has avowedly done, to illustrate the etymology of the Latin language by constant derivation from the Greek, then it is that principles hastily assumed most strikingly show, in erroneous results, the unsoundness of their foundation.

The first assumed principle was this, that the Latin is a dialect of the Greek, a descendant from it, which has degenerated from a pure original : whereas the Greek and Latin languages have for some time now been considered by all competent scholars as two distinct dialects similarly but separately formed; neither of them pure, but each variously mixed, from the common tongue of the conquerors having been blended (in different modes and degrees) with that of the original population of the countries conquered by them.

On quoting thus from the Quarterly Review (vol. xlvi. pp. 339, 40,) to express a conviction which has long been my own, I have great pleasure in appealing to such high authority as that of Adelung, adopted by Dr. Alexander Murray, of course with a view to give to my own opinion confirmation and strength.

The second great source of aberration in the Latin etymologies of Lennep and Scheid, as in the Greek abundantly, arises from a different quarter, not so much in the assumption of a wrong principle, as in the frequent neglect of a right one; in the inversion of the view from that of simple and sensible to metaphorical and ab