Composition is certainly a desirable art, and I think can scarcely be entered upon too soon. It should be one end after which I would endeavour, and the mode of effecting it will be farther illustrated in the sequel, to solicit a pupil to familiarity, and to induce him to disclose his thoughts upon such subjects as were competent to his capacity, in an honest and simple manner. After having thus warmed him by degrees, it might be proper to direct him to write down his thoughts, without any prescribed method, in the natural and spontaneous manner, in which they flowed from his mind. Thus the talk of throwing his reflections upon paper would be facilitated to him, and his style gradually formed, without teaching him any kind of restraint and affectation. To the reader who enters at all into my ideas upon the subject, it were needless to subjoin, that I should never think of putting a youth upon the composition of verse. From all I have said it will be sufficiently evident, that it would be a constant object with me to model my instructions to the capacity of my pupil. They are books, that beyond all things teach us to talk without thinking, and use words without meaning. To this evil there can be no complete remedy. But shall we abolish literature, because it is not unaccompanied with inconveniencies? Shall we return to a state of savage ignorance, because all the advantages of civilization have their attendant disadvantages? The only remedy that can be applied, is to accustom ourselves to clear and accurate investigation. To prefer, whereever we can have recourse to it, the book of nature to any human composition. To begin with the latter as late as may be consistent with the most important purposes of education. And when we do begin, so to arrange our studies, as that we may commence with the simplest and easiest sciences, and proportion our progress to the understanding of the pupil. With respect to grammar in particular, the declensions of nouns, and the inflexions of verbs, we may observe, that to learn words to which absolutely no ideas are affixed, is not to learn to think loosely, and to believe without being convinced. These certainly can never corrupt the mind. And I suppose no one will pretend, that to learn grammar, is to be led to entertain inaccurate notions of the subjects, about which it is particularly conversant. On the contrary, the ideas of grammar are exceedingly clear and accurate. It has, in my opinion, all those advantages, by which the study of geometry is usually recommended, without any of its disadvantages. It tends much to purge the understanding, to render it close in its investigations, and sure in its decisions. It introduces more easily and intelligibly than mathematical science, that most difficult of all the mental operations, abstraction. It imperceptibly enlarges our conceptions, and generalises our ideas.