bullock; some in el, as, pike, pickrel; cock, cockrel; sack, satchel; some in ing, as, goose, gosling. These seem to be the only legitimate ones, as properly belonging to our language. The rest are derived from Latin, French, Italian, and have various terminations.

CHAPTER II.

OF THE ARTICLE.

LANGUAGE is chiefly composed of general terms, most substantives being the names of genera or species. When we find a number of substances resembling one another in their principal and most obvious qualities, we refer them to one species, to which we assign a name common to every individual of that species. In like manner, when we find several of these species, resembling one another in their chief properties, we refer them to a higher order, to which also we assign a common and more general name than that, which was affixed to the inferior class. Thus we assign the general name man to the human species, as possessing a common form, and distinguished by the common attributes of life, reason, and speech. If we consider man as possessed of life only, we perceive a resemblance in this respect between him and other beings. To this higher class or genus, the characteristic attribute of which is vitality, we affix the more generic name of animal *.

* It must be obvious, that the terms general and universal belong not to real existences, but are merely denominations, the result of intellect, generalising a number of individuals under one head.

Hence, when we use an appellative or common noun, it denotes the genus or class collectively, of which it is the name, as,

"The proper study of mankind is man," i. e. not one man, not many men, but all men.

"Metal is specifically heavier than water," i. e. not this or that metal, but all metals.

But, though our words are general, all our perceptions are individual, having single existences for their objects. It is often necessary, however, to express two, three, or more of these individual existences; and hence arises the use of that species of words which have been called numerals, that is, words denoting number. To signify unity or one of a class, our forefathers employed ae or ane, as, ae man, ane ox. When unity, or the number one, as opposed to two or more, was to be expressed, the emphasis would naturally be laid on the word significant of unity; and when unity was not so much the object as the species or kind, the term expressive of unity would naturally be unemphatical; and hence ae, by celerity of pronunciation, would become a, and ane be shortened into an. These words a and an are ́now termed indefinite articles; it is clear, however, that they are truly numerals, belonging to the same class with two, three, four, &c.; or, perhaps more properly, these numerals may be considered as abbreviations, for the repeated expression of the term one. By whatever name these terms a, an, may be designed, it seems evident that they were originally

synonymous with the name of unity, or rather themselves names of unity, emphasis only distinguishing, whether unity or the species were chiefly intended. Hence a and an cannot be joined with a plural noun.

Some grammarians, indeed, have asserted that in every example where a or an occurs, the term one may be substituted in its stead, without in the least degree injuring the sense. As far as the primary idea denoted by these words is concerned, this opinion is doubtless incontrovertible, for they each express unity; but with regard to the secondary or implied ideas, which these terms convey, the difference is obvious. An example will illustrate this: If I say, "Will one man be able to carry this burden so far?" I evidently oppose one to more; and the answer might be, "No; but two men will." Let us substitute the term a, and say, "Will a man be able to carry this burden?" Is the idea nowise changed by this alteration? I apprehend it is; for the answer might naturally be, "No; but a horse will." I have here substituted a, for one; the converse will equally show that the terms are by no means mutually convertible, or strictly synonymous. If, instead of saying "A horse, a horse, a kingdom for a horse," I should say "One horse, one horse, one kingdom for one horse," the sentiment, I conceive, would not be strictly the same. In both expressions the species is named, and in both one of that species is demanded; but with this difference, that in the former the name of the species is the emphatic word, and it opposes

that species to every other; in the latter, unity of object seems the leading idea, "one kingdom for one horse." In this respect, our language appears to me to have a decided superiority over those languages, where one word performs the office of what we term an article, and at the same time denotes the idea of unity. Donnez moi un litre means either "give me one book," i. e. not two or more books; or "give me a book," that is, "a book, not something else; a book, not a pen," for example.

I acknowledge that, in oral language, emphasis may serve to discriminate the sentiments, and prevent ambiguity. But emphasis is addressed to the ear only, not to the eye; it can, therefore, be of no service in written language. It is true also, that by attending to the context error may often be avoided; but let it be remembered, as Quintilian observes*,

* Non ut intelligere possit, sed ne omnino possit non intelligere curandum.—Inst. lib. viii. cap. 4.

I am inclined to think that our language possesses a superiority in this respect over the Greek itself, Εγένετο ανθρωπος απεσαλμενος παρα τε Θεs may signify either “ man in the species, or an individual, was sent from God." The author of the article Grammar, in the Encyc. Brit. observes, “that the word aveρwπos is here restricted to an individual by its concord with the verb and the participle." If he mean by this that the term must be significant of only one individual (and I can annex no other interpretation to his words), because a singular verb and participle singular are joined with it, he errs egregiously. Numberless examples might be produced to evince the contrary. Job, v. 7. ανθρωπος γεννᾶται κόπω, man (mankind) is born unto trouble :” where the subject is joined to a verb singular, Psal. xlix. 12. ανθρωπος εν τιμη ων & συνηκε, man being in honour abideth not."