that of nouns and verbs. But when we rigidly examine this definition, we find it to be defective; for, in the first place, the peculiarity here noted applies to some other parts of speech; and secondly, it is true, after all, of the article, that it does usually specificate the meaning of the words to which it is attached, or at least give them a meaning which may be called in some respects emphatic.

With Middleton we may say, however, that we despair of obtaining from the mighty master of logic and grammar, any just and adequate view of the nature of the article, by the definition which he has given. The celebrated grammarian, Apollonius Dyscolus, who flourished about A. D. 150, although he has said much of the article, has left us no express definition of it, by which we can learn his views exactly. He asserts, indeed, that articles and pronouns are different things, and yet, that if the article loses its noun, it then becomes a pronoun. Middleton avers, in respect to him, that "he has many facts, for the most part corroborating the theory which he [Middleton] suspects to be the true one." This may be so; but the inquiry of most importance is, whether the Greek language itself corroborates this theory.

Let us hear another celebrated Greek grammarian, who lived in the fifteenth century. I refer to Theodore Gaza, whose grammatical work was the souree and exemplar of almost all the earlier Greek grammars in western Europe. "The article," says he, "is a declinable part of speech prefixed to nouns. It is, indeed, divided into the prepositive and subjunc

tive; but properly the prepositive only is the article." He then adds, (what is the essence of his definition,) ποιεῖ δ ̓ ἀναπόλησιν προεγνωσμένου τοῦ ἐν τῇ συντάξει, i. e. "it serves to recal that which had previously been known (or mentioned) in the discourse."

Here, indeed, we have one important remark, viz. that the prepositive article is the only real and true article. Why the so called subjunctive article should ever have been named otherwise than pronoun, it seems difficult to imagine. But we are not brought much in advance upon our way, by the rest of Gaza's remarks. It is a very limited part of the article's office, to refer merely to what has been suggested or recognized in previous discourse. Even if the old rule of definition here-a potiori nomen fit-were applied, we should hardly be able to defend the definition of Gaza. Moreover, the relative and the demonstrative pronouns also serve to recal that which was mentioned or recognized in the preceding context; and how does the definition of Gaza serve to distinguish the article from them?

Mr. Harris, who in his Hermes has written so many curious and in several respects interesting things concerning the philosophy of language, speaks of the article as being nearly allied to the pronoun, and remarks that they may be best distinguished by the circumstance, that "the genuine pronoun always stands by itself, while the genuine article requires a noun for its support."

Lord Monbodo, who has speculated much and often to very good purpose on language, and who was

uncommonly well versed in the writings of the Greek philosophers and metaphysicians, remarks, that "the article is of as subtle speculation as perhaps any thing belonging to the language; particularly as it is used in Greek." In this he was beyond all doubt correct. He then goes on to show, that "its office is different from that of a pronoun of any kind, and that it deserves to be ranked by itself among the parts of speech." But after all, when he comes specifically to define it, he makes it "the prefix to a noun, denoting simply that the noun to which it is prefixed, is the same with that which was before mentioned, or is otherwise well known." But these uses of the article are far from being the only ones which it subserves. The definition, therefore, is incomplete. Middleton objects to this definition, however, that "it makes the article a distinct part of speech; and that if it be thus distinct, it is not conceiveable that it should become a pronoun when (as Apollonius affirms) its substantive is dropped; inasmuch as one distinct part of speech cannot go over into another.” But the correctness of this last remark will hardly be conceded. Does not an adjective often go over into a noun? Do not the primitive prepositions, when they are joined in composition with a verb, become adverbs? Do not forms of the infinitive mode very often become mere nomina actionis-simple nouns? What then is the difficulty in the case before us? Why cannot the article, in certain cases, go over into a kindred class of words (to say the least), i. e. into a pronoun, although it be of itself a distinct part of speech?

From this brief review of the former ancient and modern definitions of the article, let us come to those of some of our cotemporaries, who are, or have been great masters in criticism, lexicography, or grammar. Dr. Middleton, to whom I have already more than once referred, published, some twenty years since, a treatise on the Greek article, which he entitled: The Doctrine of the Greek Article, applied to the criticism and illustration of the New Testament. In this he says, (p. 4 of the New York edit.) "The Greek prepositive article is the pronoun relative ő, so employed that its relation is supposed to be more or less obscure; which relation, therefore, is explained in some adjunct annexed to the article by the participle of existence expressed or understood." His meaning is, for example, that ỏ piñócopos is in all cases equivalent to ỏ ὢν φιλόσοφος· in which case ὁ is the subject of an assumed proposition, v the copula, and pióropos the predicate. According to him, then, the article stands in all cases, in connection with its noun, in a proposition which differs from one that has a verb, only as an assumptive proposition differs from one that asserts, i. e. as ὁ ὢν φιλόσοφος differs from ὅ ἐστι φιλόσοφος.

To explain and defend this definition, he occupies twenty pages of his Essay. I have read this part of his work many times heretofore, and recently with all the attention that I could summon, and yet I feel compelled to say, that Aristotle's definition, which has been cited above, and with which Middleton finds much fault, is at least as intelligible to me as that of his corrector.

What is a part of speech
"so employed, that its

relation is supposed to be more or less obscure ?” Men employ language in order to clear away obscurity; and they always complain of an unskilful or ill use of it, when it is so employed as to be obscure. The author, however, to do him justice here, means to say, that the relative pronoun i is so employed as to be anticipative, (which is sometimes the case with pronouns really relative,) and that the noun anticipated, i. e. the noun to which the article has reference, is to be fully known only by the mention of it in the sequel. For example; i he who (for so, if Middleton be correct, we must translate it, when rigid exactness is applied to it,) stands with an uncertain meaning or reference, until gwos, piñóσopos, or some other noun is supplied. In this way a kind of relative meaning is made out for the article, and on this ground the author in question calls it a relative pronoun; contrary to the great mass of critics, ancient and modern, who, when they admit its pronominal quality, always make it, in the main, a demonstrative pronoun. Matthiae (Gramm. § 292)

does, indeed, admit that the article is used for the relative pronoun; but he limits this to the Doric and Ionic dialects, and to the tragic poets only among the Attics. It should be noted, however, that this relative use which he thus admits, constitutes but a small part of the instances in which the article is employed.

In fact, if we are to name the article a pronoun in all cases, we must evidently divide the generic denomination into several species, viz. into the demonstrative, the most frequent of all in ancient times; the relative, which is less common, and somewhat pe