then proceeds to the reading and wrlting of easy sentences, perhaps in such a book as “Gradatim”; and after say a year or more he will be reading easy selections from a Latin author. All the while he recapitulates his grammar and extends his grammatical horizon. This is, in any case, an immense improvement on the older plan of learning the whole of the old Eton Latin Grammar in its Latin dress without understanding a word of what is meant by its “as in praesenti” and other mysteries. If wisely administered, this method may also avoid the error of “Henry‘s First Latin Book,” which taught an intolerable deal of Accidence and Syntax to a half-pennyworth of text; though, on the other hand, Henry’s First Latin Book was an attempt to accompany the learning of grammar with the reading of easy sentences from the very beginning, and in so far was better than the method we are considering. For I must maintain, with all deference to the opinion of others whose experience is wider than my own, that we are as yet far from having drawn the full conclusions of the process of reasoning on which we have entered. There should be no preliminary study of grammar apart from the reading of a text. The declensions and conjugations, learncd by rote apart from their applications, cannot be properly assimilated or understood, and often prove a source of error rather than enlightenment in subsequent study. They have to be learned over and over again—always in doses which are too large for digestion, and the pupil has meanwhile been encouraged to form a bad habit of mind. Half knowledge in this case too often leads to the unedifying spectacle of the Sixth Form boy or the University undergraduate who is still so shaky in his accidence that he cannot pass his “Smalls” without a special effort, though in some respects he may be a good scholar. But still more serious is the effect of the false conceptions which are inevitably implanted in the mind by this method of grammar without understanding. The pupil learns mensā, “by or with a table,”
