study of modern languages is gradually established among us on a sound and thorough basis. That expectation may seem paradoxical, and yet I think it is reasonable. For, when people have fully realised that intellectual grasp of a modern language can be more surely tested by the power of composing in it than by the power of using it colloquially, then they will be less disposed to disparage the value of the same test in Greek and Latin. Then, too, perhaps we shall no longer hear that Nature cries aloud to us to teach the classical languages chiefly by an oral use of Dr Ollendorff's method. There is indeed one essential difference, as regards composition, between modern languages and the classics, which should be well noted here. It depends on the difference between idiomatic writing, and writing which is merely correct in point of grammar. In a modern language, when the learner knows only that he is grammatical, but doubts whether he is idiomatic, he can easily decide the point by an appeal to living authority. But in Greek and Latin, the only gauge of idiomatic truth is that which is furnished by the literature; and while it is comparatively easy to ascertain the rules of grammar, it requires very careful study—study which tasks not only intellect, but feeling and taste—to seize that subtle reflection of a living personality which in language appears as idiom, and which can still be apprehended, though sometimes but dimly, and with an inevitable element of uncertainty, in the literary records of the ancient world.
