comparatives and superlatives of the words good, bad, great, small, much, little, &c. and these should be learnt among the first rules and variations, because they continually occur.

But as to other words which are less frequent, let but few of the anomalies or irregularities of the tongue be taught among the general rules to young beginners. These will better come in afterwards to be learnt by advanced scholars in a way of notes on the rules, as in the latin grammar called the Oxford grammar, or in Ruddiman's notes on his Rudiments, &c. Or they may be learnt by examples alone, when they do occur; or by a larger and more complete system of grammar, which descends to the more particular forms of speech: so the heteroclite nouns of the latin tongue, which are taught in the school-book called Quæ Genus, should not be touched in the first learning of the rudiments of the tongue.

II. As the grammar by which you learn any tongue should be very short at first, so it must be written in a tongue with which you are well acquainted, and which is very familiar to you. Therefore I must prefer even the common English accedence (as it is called) to any grammar whatsoever written in latin for this end. The English accedence has doubtless many faults: but those editions of it which were printed since the year 1728, under the correction of a learned professor, are the best ; or the English rudiments of the latin tongue by that learned North-Briton Mr. Ruddiman, which are perhaps the most useful books of this kind which I am acquainted with; especially because I would not depart too far from the ancient and common forms of teaching which

several good grammarians have done to the great detriment of such lads as have been removed to other schools.

The tiresome and unreasonable methods of learning the latin tongue by a grammar with latin rules would appear even to those masters who teach it so, in its proper colours of absurdity and ridicule, if those very masters would attempt to learn the Chinese or Arabic tongue, by a grammar written in the Arabic or the Chinese language. Mr. Clark, of Hull, has said enough in a few pages of the preface to his new grammar, 1723, to make that practice appear very irrational and improper; though he has said it in so warm and angry a manner, that it has kindled Mr. Ruddiman to write against him, and to say what can be said to vindicate a practice, which I think is utterly indefensible.

III. At the same time when you begin the rules, begin also the practice. As for instance, when you decline Musa, Musa, read and construe the same day some easy latin author, by the help of a tutor, or with some English translation: chuse such a book whose title is simple, and the subject of discourse is very plain, obvious, and not hard to be understood; many little books have been composed with this view, as Corderius's Colloquies, some of Erasmus's little writings, the saying of the wise men of Greece, Cato's moral Distiches, and the rest which are collected at the end of Mr. Rudiman's English grammar, or the latin Testament of Castellio's translation, which is accounted the purest latin, &c. these are very proper upon this occasion, together with Esop's and Phædrus's fables, and little stories, and the common and daily affairs of domestic life, written in the

latin tongue. But let the higher poets and orators and historians, and other writers whose language is more laboured, and whose sense is more remote from common life, be rather kept out of sight till there be some proficiency made in the language.

It is strange that masters should teach children so early Tully's Epistle, or orations, or the Poems of Ovid or Virgil, whose sense is oftentimes difficult to find because of the great transposition of the words; and when they have found the grammatical sense, they have very little use of it because they have scarce any notion of the ideas and design of the writer, it being so remote from the knowledge of a child: whereas little common stories and colloquies, and the rules of a child's behaviour, and such obvious subjects, will much better assist the memory of the words by their acquaintance with the things.

IV. Here it may be useful also to appoint the learner to get by heart the more common and useful words, both nouns and adjectives, pronouns and verbs, out of some well-formed and judicious vocabulary. This will furnish him with names for the most familiar ideas.

V. As soon as ever the learner is capable, let the tutor converse with him in the tongue which is to be learned, if it be a living language, or if it be latin, which is the living language of the learned world: thus he will acquaint himself a little with it by rote, as well as by rule, and by living practice as well as by reading the writings of the dead. For if a child of two years old by this method learns to speak his mother tongue, I am sure the

same method will greatly assist and facilitate the learning of any other language to those who are older.

VI. Let the chief lessons and the chief exercises of schools, v. c. where latin is learnt, (at least for the first year or more) be the nouns, verbs, and general rules of syntax, together with a mere translation out of some latin author into English; and let scholars be employed and examined by their teacher daily in reducing the words to their original or theme, to the first case of nouns or first tense of verbs, and giving an account of their formations and changes, their syntax and dependencies, which is called parsing. This is a most useful exercise to lead boys into a complete and thorough knowledge of what they are doing.

The English translations, which the learner has made, should be well corrected by the master, and then they should be translated back again for the next day's exercise by the child, into latin, while the latin author is withheld from him: but he should have the latin words given him in their first case and tense; and should never be left to seek them himself from a dictionary: and the nearer he translates it, to the words of the author whence he derives his English, the more should the child be commended. Thus he will gain skill in two languages at once. I think Mr. Clark has done good service to the publick by his translations of latin books for this end.

But let the foolish custom of employing every silly boy to make themes or declamations and verses upon moral subjects in a strange tongue, before he understands

N

common sense, even in his own language, be abandoned and cashiered for ever.

VII. As the learner improves, let him acquaint himself with the anomalous words, the irregular declensions of nouns and verbs, the more uncommon connections of words in syntax, and the exceptions to the general rules of grammar. But let them all be reduced, as far as possible, to those several original and general rules, which he has learned as the proper rank and place to which they belong.

VIII. While he is doing this, it may be proper for him to converse with authors which are a little more difficult, with historians, orators and poets, &c. but let his tutor inform him of the Roman or Greek customs which occur therein. Let the lad then translate some parts of them into his mother-tongue, or into some other well known language, and thence back again into the original language of the author. But let the verse be translated into prose, for poesy does not belong to grammar.

IX. By this time he will be able to acquaint himself with some of the special emphasis of speech, and the peculiar idioms of the tongue. He should be taught also the special beauties and ornaments of the language: and this may be done partly by the help of authors who have collected such idioms, and cast them into an easy method, and partly by the judicious remarks which his instructor may make upon the authors which he reads, wheresoever such peculiarities of speech or special elegancies

occur.