530
of this part of speech, they are awkward and inconvenient. The syntax of the articles may be much better expressed in this manner: "Articles relate to the nouns which they limit," for, in English, the bearing of the articles upon other words is properly that of simple relation, or dependence, according to the sense, and not that of agreement, not a similarity of distinctive modifications.
OBS. 13.--Among all the works of earlier grammarians, I have never yet found a book which taught correctly the application of the two forms of the indefinite article an or a. Murray, contrary to Johnson and Webster, considers a to be the original word, and an the euphonic derivative. He says: "A becomes an before a vowel, and before a silent h. But if the h be sounded, the a only is to be used."--_Murray's Gram._, p. 31. To this he adds, in a marginal note, "A instead of an is now used before words beginning with u long. It is used before _one. An_ must be used before words WHERE the h is not silent, if the accent is on the second syllable; as, _an heroic action, an historical account_."--_Ib._ This explanation, clumsy as it is, in the whole conception; broken, prolix, deficient, and inaccurate as it is, both in style and doctrine; has been copied and copied from grammar to grammar, as if no one could possibly better it. Besides several other faults, it contains a palpable misuse of the article itself: "_the h_" which is specified in the second and fifth sentences, is the "_silent h_" of the first sentence; and this inaccurate specification gives us the two obvious solecisms of supposing, "_if the [silent] h be sounded_," and of _locating "words WHERE the [silent] h is not silent!_" In the word humour, and its derivatives, the h is silent, by all authority except Webster's; and yet these words require a and not an before them.