Syntax; but as this presumption is not always borne out by facts, it is necessary in the junior classes to commence the session with a series of easy exercises to enforce the ordinary rules of Syntax. But I have no faith in sets of exercises which are arranged so as each to illustrate some special rule. When a student finds the 'Dative,' 'qui with the Subjunctive' or 'Indirect Question,' at the head of a set of sentences, all he has to do is boldly to throw in the Dative or the Subjunctive wherever they can be inserted without absurdity, and in five cases out of six he will be right. In this way a teacher may find little to correct, and yet to his dismay discover at a later stage that his pupil has gained no real mastery over the constructions he has practised. It will be found accordingly that Parts I and II contain scarcely a sentence which illustrates only, or even mainly, one single rule. Headings have been prefixed throughout; but these only indicate that in the sentences which follow the teacher will find examples enough to illustrate the particular rule indicated, while alongside of these he will find other constructions from which it must be distinguished, and more especially those with which a careless student would be likely to confound it.

I have eschewed Simple sentences even in Part I. Simple sentences may be necessary for mere beginners; and nothing but the Ollendorf principle is suitable for children, whose minds are not capable of grasping the logical

relations of a compound sentence, even in their own language. But as the Dean of Westminster has well pointed out in the preface to his admirable edition of T. K. Arnold's Latin Prose Composition, it is impossible to make any real use of a language as an instrument of thought, for expressing even the most simple events of life, without introducing subordinate clauses: the attempt therefore to construct a series of exercises on a strictly progressive principle, so as never to introduce a construction in a sentence until it has been separately explained and illustrated, is not only very tedious in itself, but it postpones indefinitely the interest which a learner feels when he finds he can make real use for his own purposes of the language which he is studying. For this reason Compound sentences, especially such as contain simple Adjectival clauses, have been introduced from the very beginning. If the teacher finds they are beyond the strength of his pupil, it will be easy for him to begin by breaking up the Compound sentence, and to put before his pupil, for his first lesson, the simple sentences into which it may be resolved. Thus, while practising himself in the simpler rules of Syntax, the learner will, at the same time, and almost unconsciously, be acquiring some knowledge of more difficult constructions, and gaining by habit, as every child does when he learns his own language, some familiarity with the principles of composition.

Not less important is it that the teacher should insist, from the very first day, upon the necessity of observing the true Latin order of the words. From writing only simple sentences, which leave no room for variety, the learner acquires the fatal habit of following the English order of words in a sentence, and this habit it is most difficult to unlearn. It cannot be impressed upon a learner too soon that it is as gross a fault in writing Latin to use a non-Latin order of the words as it is to commit a positive error in construction.

Following here, too, the Dean of Westminster's example, I have made the English of the sentences as idiomatic, and as unlike the corresponding Latin, as I could. No process of thought is involved, no mastery over the construction of a language gained, when a pupil in translating is allowed to use the same words, the same constructions, which he would use in English: he cannot be taught too early that Latin and English are two different languages, and that he performs no act of translation if he merely takes the words of one language and translates them into those of another.

For the Exercises in Part I, a general reference has been given to the Public School Latin Primer, which the student should study carefully for each construction in succession. For the more difficult Exercises in Part II, he is referred throughout to the excellent book of the Dean of Westminster, where he will find not only

a full account of every important construction, and of almost all the niceties of Latin idiom, but also a vivid appreciation of the special points of contrast between Latin and English-all given with a force, point, and clearness which recall to us who had the privilege of being his pupils why we have always deemed him an almost unapproachable master in the art of teaching not only how to write Latin Prose, but how to think it also. In the chapters of the Dean of Westminster's book the student will find all he needs in the way of explanation and information; it will be for him to apply that instruction to his own sentences. All further hints, references, or explanations are omitted. It is good to indicate to a student where he will find the guidance that he needs to supply him with a finger-post at every turn is not to help him in his work, but to do his work for him.

Part III contains a number of easy, selected passages for translation into Continuous Prose, at about the level of our ordinary Pass Degree. They will be found graduated in point of difficulty, and consist mainly of simple historical narratives or anecdote, such as are useful for students who are making their first essays in acquiring a Latin Prose style. There are some excellent collections of passages for Latin Prose in existence, but I know none which contains a sufficient number of easy passages, to bridge over the gulf

between isolated sentences and passages difficult in thought as well as in style. For some of the passages in Part III, I have to thank my friend the Rev. C. Darnell of Cargilfield, whose remarkable power of teaching Latin Prose to boys is known to all who have examined his pupils.

Part IV consists of more difficult passages, all of them, it is thought, passages of literary excellence, and which have approved themselves as suitable for translation into Latin. They have been arranged in subdivisions, in accordance with the character of their contents. A few of these passages have appeared in other collec

tions.

It is not my intention to publish a Key to this collection: indeed, my main object in compiling it has been that there may be at least one Latin Prose book in existence which has no Key. My experience as a teacher is that nothing is so injurious to sound scholarship, nothing so much baffles the efforts of the teacher, and retards the progress of the learner, as the use of keys and translations,-especially by those who are not far enough advanced to know how to make a right use of them. To an advanced scholar, who can appreciate, if he cannot produce, what is good, nothing is more stimulating than to have put before him as a model a finished version by a good scholar; but for a student who has not yet reached this stage it is more