taught; and on the mode will depend, secondly, the amount of exact knowledge obtained. This latter we believe to be a secondary consideration; but I shall still endeavour to show that the amount of useful knowledge attainable in a good course of classical instruction may be increased much beyond what it ordinarily is. We suppose that our students in general will terminate their classical studies at an earlier age than they do in the two ancient Universities of Oxford and Cambridge, and that they will proceed from the study of Latin and Greek to some other branches of learning, or that they will begin to combine professional with general education somewhat earlier than is now usual. But we do not, therefore, suppose they will not learn Latin and Greek well we know that both languages can be learned well in much less time than they are often learned very imperfectly. The student may begin later, and he may end earlier; he will end with the instruction of the teacher, but he will be endowed with the power and the wish to continue the labour of selfinstruction.

The study of the ancient languages is generally commenced too soon. The age should depend partly on the capacity of the pupil, but principally on previous acquirements of the kind which have been described. The early age at which the Latin language is begun is the chief reason why the grammar system continues. In youth the memory is retentive, but the understanding is unformed; the accidents of words and the precepts of grammar can be committed to memory, and repeated long before they are understood. According to the grammar system both teacher and pupil do all that they can do; one repeats, and the other listens ;

the duties of both become mechanical, and consequently irksome.

An elementary grammar should be very short and simple; it should exhibit nothing more than what is technically called the accidence, which comprehends the declensions and conjugations. Nor should even this grammar be committed to memory as a preparation for the study of the language, but it should be used as a book of aid and reference in connexion with the teacher's instruction. To learn a language, the pupil must have a language before him. He must see specimens of the material on which he is called to exercise the faculties of observation, judgment, and memory. Some short sentences then, at least, are necessary for a beginner; and they are perhaps better than a continuous narrative, because the latter is very seldom found to be simple and clear enough in the structure of all its sentences. A short sentence, containing one proposition, is much easier than those found in consecutive description, which necessarily requires all the modes of speech. which constitute a language. As a continuous narrative, however, possesses the great advantage of attracting the pupil's attention by the nature of the subject, this is so far a good reason for preferring it; but it will be found to be a great aid to the student, and an almost necessary one, to break up each complex sentence into distinct propositions, or simple sentences, for the purpose of avoiding at the commencement the difficulties inseparable from long and involved periods. It woul be easy to form the first book of Cæsar into short sentences of the kind described, and to bring the student to the examination of the complex sentence after he has mastered all the varieties of the simple proposition.

2 B 3

But how is the pupil to prepare his lesson, to find out the meaning of this new language? Shall he try to do it with the aid of a general lexicon? Shall we ask him to perform the work with a bad or imperfect instrument, which he does not know how to use?

Some elementary books are furnished with a small lexicon, adapted to explain the words of the book in the sense in which they occur. This is a great improvement. Some teachers recommend what is called a literal translation, accompanied occasionally by one more free and elegant.

Perhaps the oral instruction of the teacher is better than all for young learners. It is the essence of good teaching to make things easy to be understood, and pleasant to learn; and this may be done by the teacher explaining to the pupil his short lessou, by giving him the meaning of each word, and by pointing out, as soon as he thinks it advisable, the modifications which its ending or termination undergoes from its connexion with other words. This is, in fact, teaching the elements of grammar; for which purpose it will soon be found useful to have a book that shall contain the general form or models under which are comprehended all the particular instances that the pupil can meet with in his lesson. It may be found convenient for the pupil even to commit to memory the general forms or examples of this grammar; but certainly not before he has frequently met with particular instances in his reading, and has been led to observe the advantage of comparing them, and arranging them for the purpose of aiding the memory. After the student has received instruction in his lesson, there remains the business of examination, which consists in proposing questions, in presenting the matter

in all the varieties of which it is susceptible, and thus leading the pupil to observe in new sentences, without the teacher's aid, the facts which he has already learned. The teacher may soon vary the lesson by the reverse operation of giving the pupil the English expression, and requiring him to repeat and also to write the corresponding Latin. In this way words are learned with their grammatical accidents, and, what is quite as important in a language, the proper order of words is acquired. A system that teaches by transposing the words of a language, ancient or modern, is not good.

It will be observed that this reverse operation is really almost the same thing as committing the lesson to memory; and when the lesson is well understood in all its parts, it is found, by experience, to be a good exercise to commit a small portion of it to memory, and one attended with very little trouble. A proper examination into every word will prevent its becoming a mechanical task. When I speak of committing part of the lesson to memory, I do not wish to be understood as recommending this practice to be applied to every lesson during a long course of instruction. The advantage of knowing by heart a portion of an author is briefly this. A student will make the part with which he is so familiar the test and the standard by which he will try and estimate the various words and phrases that occur in other parts of the book. It will be a grammar, because it has been carefully analyzed and thoroughly impressed on the memory by a proper study; and it will be more than a grammar, as it will furnish a great variety of modes of expression, applicable to the explanation of similar phrases that may occur. It is, in fact, the next thing to making the dead into a living language,

All the advantages to be derived from this practice are obtained by committing to memory the first ten or twenty pages of the book which is made the basis of the student's knowledge of a language. For we think that one book should be made the basis of all this knowledge, and that this book should be a prose author.

It is almost unnecessary to remark, that committing a lesson, or any thing else, to memory before it is understood, is just the reverse of what we are labouring to recommend.

In learning a new language we should begin to read an entire book as soon as we can; one that will give the teacher an opportunity of connecting with his grammatical instruction useful remarks on the subjectmatter generally. A book of prose is better than one of poetry, though some kinds of descriptive poetry may be occasionally used with profit; and a narrative or history is better for a young pupil than oratorical declamation or books on moral subjects.

When a student has made some progress, and when he has begun to read an author, he must be taught to ob serve more accurately the relationship of words to one an other, or the idiomatical usages of the language. These usages, when classified, are comprehended, at least part of them, under the head of syntax; and it is not unusual for the rules of syntax to be committed to memory long before the student either wants them or understands them. Sometimes, too, the rules themselves are written in Latin, and, as all may recollect who have been subjected to the experiment, they are for a very long time either misunderstood or not understood at all. If a complete system of rules is committed to memory, many of rare occurrence and some of doubtful character