come up to the description of their ancestors in the Germania (Op. 62.405).

Of the poets Terence seems to have influenced Luther very much. This is quite natural, for Terence was one of the most popular authors of the Middle Ages. There are countless allusions to him in the Reformer's writings. But Vergil was perhaps his favorite. This author surpasses all others heroica gravitate, he says (Op. 62.343). He liked the Bucolics and the Georgics best; this is only what we should expect of the son of a peasant (Schmidt, 26). Luther quotes Vergil constantly: Quos ego .; forsan et haec olim meminisse iuvabit; fuimus Troes, etc.

The elegance and the charm of Ovid also gave keen pleasure to Luther (De W. 1.190). Oddly enough Luther rated Horace below Prudentius (Op. 62.104); still he borrowed many proverbial expressions from the Augustan poet. Juvenal was rather well known too. The Reformer bought a copy of Lucan when he was already middle-aged (Op. 62.343). The late Latin Disticha of Dionysius Cato were prime favorites. 'Next to the Bible, we have no better book than Catonis scripta and fabulas Aesopi' (Op. 62.459). Juvenal, Martial, Catullus, and the Priapea, he says, ought not to be read in the Schools (Op. 62.344).

Luther's knowledge of Greek is much less extensive than his knowledge of Latin. Where he acquired it is not definitely known. He studied under Melanchthon at Wittenberg after 1518 (De W.1.138), but knew some Greek before this time. He was able to translate the Greek New Testament in 1518. But, when one of his friends wrote several Greek letters to him, he became very angry, and said he wished he could reply in Turkish in order that his correspondent might also be forced to read a letter he could not understand (De W. 4.16). Another time he punished a similar offense by including in his reply a German nonsense verse written in Hebrew letters (Mathesius, Sermon 12). Still, he knew Homer rather well and made frequent allusions to the story and the mythology of his poems. Some of the other Greek poets are also quoted in Luther's writings, but there is no proof that he had any first-hand knowledge of them. Even the philosopher Plato he knew probably only at second-hand. Aristotle, however, the foundation of the late mediaeval philosophy and largely of the theology too, he had studied thoroughly, though mostly through Latin translations. He began his teaching at Wittenberg by lectures on Aristotle's Dialectics and Physics (C. R. 6.157). However, he soon came to the conviction that the Stagirite's influence in theology was harmful and therefore he often sharply attacked him (Op. 22.196; 31.344). Aristotle was to him the personification of his great enemy, Scholasticism. Later in life his criticisms of Aristotle were milder. The regard in which he held Aesop has already been mentioned. He made a translation of a number of the Fables (Op. 14.349). During his stay at the Castle of Coburg, while the Diet at Augsburg was in session, he wrote to Melanchthon that he intended to

build three tabernacles on this Zion, one for the Prophets, one for the Psalter, and the third for Aesop (De W. 4.2).

But what did the Classics do for Luther? In the first place, they gave him a vehicle of expression. His Latin is indeed far removed from the Ciceronian smoothness of Erasmus or Melanchthon and from the vigor of his own German style, but it is correct, clear, and expressive. He is a complete master of the idiom. Now and then he even wrote a few Latin verses, the most interesting of which are an adaptation of Martial 10.47, in which he paraphrases the thoughts of Psalm 128 (Ор. 17.265). His German poetry, too, he says was influenced by Vergil (Schmidt, 27).

Greater, but more impalpable perhaps, is the influence of the Classics in broadening his mind and opening for him new sources of knowledge. They made him familiar with the thoughts of a complex civilization; they gave him a wealth of illustration and allusion from history, philosophy, law, and natural science. They moulded his ideas of patriotism and nationalism, they opened his eyes to nature and to mankind, freed him from prejudice, and mellowed his opinions.

Let us turn for a moment to see what were Luther's relations with the humanists, the official bearers of the torch of classical learning at the time. There was among the humanists a strong feeling against the abuses of the Church, against its financial exactions, against the ignorance of the clergy and the attendant ills, and against the laxity of its morals. On these points the humanists found themselves in sympathy with Luther. Luther, too, immediately adopted several of the salient principles of humanism-the authority and value of the original sources as against the later interpreters and commentators, and, further, the right of each individual to examine and interpret these sources. However, humanism concerned classical philology, the Reformation concerned the teaching of the Church; for the one the Classics were an end in themselves, for the other a means toward an end. Therefore, some of the humanists, like Melanchthon, became Protestant; others, like Erasmus, remained Catholic.

Before we discuss the part the Reformer wished the Classics to play in his scheme of Christian education, it seems best to say a few words about his great friend and assistant, the humanist Melanchthon, because the educational work of the two is so closely bound together.

Melanchthon, the grand-nephew of the well-known humanist Reuchlin, was more fortunate in his early training than Luther had been. The town of Bretten, where he received his elementary education, seems to have had a fairly good School, but, on the advice of his famous relative, he soon became a pupil of John Unger, a splendid teacher. 'He drove me to grammar and required me to construct sentences. He made me give the rules of construction by means of twenty or thirty verses from the Mantuan" (C. R. 25.448). When ten

Baptista Mantuanus, 1448-1516.

years old Melanchthon went to Pforzheim, where he read Greek and Latin poets (C. R. 2, Decl. 135). Reuchlin, whom he now met frequently, gave him a Greek grammar and a copy of his own Greek-Latin lexicon. Together with some friends Melanchthon produced one of Reuchlin's School comedies in honor of the humanist-at the age of twelve (Cam. 2).

The University of Heidelberg, which he then began to attend, was scholastic, yet he continued his own reading of the authors. He writes (C. R. 5.715):

'Inasmuch as I had learned to write verse, with a kind of boyish avidity I began to read the poets and also history and the drama. This habit gradually led me to the ancient Classics. From them I acquired a vocabulary and style. We read everything without discrimination, but especially did we prefer the modern works, like those of Politian'.

He also studied rather intently the structure of the orations of Cicero and Demosthenes (C. R. 4.715). Before he was fifteen he had written several Latin elegies (C. R. 10.669. For his poems see Index to C. R. 38). At the University of Tübingen, where there was more of a humanistic atmosphere, he continued his private reading of Cicero, took up Vergil, Galen, and other writers, and heard lectures on Aristotle (Hartfelder, 35).

In 1514, at the age of seventeen, he began to teach at the University of Tübingen; Vergil, Terence, Livy, and Cicero were his subjects (Cam. 7). Two years later he published an edition of Terence (C. R. 1.13), then of a part of Plutarch. He also began a translation of Aratus into Latin verse, but soon left this to begin an edition of Aristotle's Analytica Posteriora (C. R. 1.26. For his translations see C. R. 17-19).

The next vear Melanchthon became first Professor of Greek at Wittenberg. In his inaugural he says (C. R. 11.15):

'Only regard for the proper studies and the duties of my office could induce me to commend to you the study of classical literature. I have undertaken to plead their case against those who call back

the German vouth from the happy contest of letters by the more than Thracian remark that classical studies are more difficult than useful; that Greek is studied only by disordered intellects and that, too, for display'.

Interesting, also, in this address is the use made of a leading humanistic principle: 'When we go to the sources, then are we led to Christ'. Melanchthon began his labors at the new University by work on Homer and the Epistle to Titus.

During this time Melanchthon did not rest from his scholarly activity. In 1518 he published his lectures on the Epistle to Titus and wrote to friends that he had ready for publication a Greek dictionary, two treatises of Plutarch, a Greek hymn, Athenagoras, Plato's Symposium, and three books on rhetoric (C. R. 1.44, 50, 52). Two years later he published a hand book on Dialectics, edited the Clouds of Aristophanes, and sent

out a new edition of the Greek Grammar he had written some time before (C. R. 20.1), besides aiding Luther in the revision of his translation of the New Testament and writing a Latin Grammar for his own private School (C. R. 20.192).

This private School had been begun almost as soon as the young professor came to Wittenberg. Its purpose was to prepare boys for the University in Greek and Latin, rhetoric, logic, mathematics, and physics (Hartfelder, 491). It seems rather modern that he should have encouraged his pupils to render dialogues and comedies from Seneca, Plautus, and Terence, a practice which Luther approved (Op. 62.336). We may remark here that he also took an interest in ancient coins (Cam. 13).

But, according to the programme of the Reformers, Schools were not to be exclusively private enterprises. As early as 1520 Luther, in his appeal To the Christian Nobility (Op. 21.277), had advised that there should be public Schools in which even girls were to have at least enough instruction in German or Latin to be able to read the Gospels. Universal education was advocated also in 1524, in the pamphlet addressed to the Aldermen (Op. 22.168). Even those students who did not intend to enter a learned profession and for whose vocational training provision was to be made ought, he held, to have some schooling in Latin. This training was, of course, to be much more intensive for the future theologians, lawyers, etc. 'Although the Gospel has come and is daily coming only through the Holy Spirit, nevertheless it has come by means of the languages'.

Nuremberg, which was already fairly well supplied with Schools and with humanists, took up Luther's suggestions and asked Melanchthon's aid in realizing them, as head of a new Gymnasium. Although he declined, he went to Nuremberg and gave his aid and advice in founding the School (Cam. 31).

In 1528 appeared the Saxon Visitation Articles (C. R M.48) which give an outline of the School system as 26elanchthon wished it to be. In the first place, teachers should be careful to teach only Latin, not German, Greek, or Hebrew. The primary School was to consist of three classes, which, however, were not to be absolved in one year. In the first the pupil was to study the alphabet, the Lord's prayer, the Creed, Donatus, and Cato's Disticha; in the second class, he was to be occupied with Aesop (in Latin), extracts from the writings of the contemporaneous authors Mosellanus and Erasmus, and with Terence and Plautus. Besides, there was to be drill in grammar. The third division was to take up Vergil, Ovid, and the Letters or the De Officiis of Cicero, and to continue the study of grammar, dialectics, and rhetoric. In all classes the boys were obliged to talk Latin with the teacher and with one another. This elementary School was followed by the Gymnasium, in which the study of Latin was continued, Greek and Hebrew were begun, and Greek was continued through Isocrates, Xenophon,

Plutarch, Hesiod, Theognis, etc. (Heerwagen, Geschichte d. Nurnb. Gelehrtenschulen, Nuremberg, 1860). "We still have the correspondence between him <Melanchthon> and fifty-six cities asking counsel and assistance in founding and conducting Latin schools and gymnasia" (Richard, 134). Nearly all the Latin Schools of the sixteenth century in Germany were founded according to Melanchthon's directions. His textbooks were used in all of them. His Latin Grammar, for instance, which went through fifty-one editions from 1525 to 1737, was used in all Saxon Schools up to the latter date. His Greek Grammar, in forty-three editions, from 1518 to 1622, had a similar sphere of usefulness. Works on rhetoric and dialectics, psychology, physics, ethics, history, and religion were also popular; some of them were even used in Catholic Schools (Hartfelder, 211 ff.; Richard, 134-136).

Melanchthon's influence, however, was probably greatest through his reorganization of the Universities. The change, of course, came first at Wittenberg. Already in 1518 Luther wished to have one of the teachers at Wittenberg, who was reading on Thomistic logic, lecture instead on Ovid's Metamorphoses (De W. I. 190-222). In the same year Melanchthon dealt with Homer. Luther interceded for him, that he might be freed from the scholastic lectures on Aristotle's Physics (De W. 1.238). Other innovations in the curriculum are suggested in Luther's writing To the Nobility (Op. 21.277). Melanchthon was the soul of these changes, which revolutionized the Wittenberg curriculum. A summary of all the new departures is contained in the statutes for the government of the faculty of theology and liberal arts written in 1545 (C. R. 10.992).

The University of Marburg very soon adopted a plan almost identical with the order of studies at Wittenberg. Koenigsberg was founded in 1544, according to Melanchthon's directions, as was also Jena, in 1548. His plans were ultimately adopted by Tübingen, Leipzic, Heidelberg, Frankfurt a. O., Rostock, and Greifswald (Richard, 138). The outcome was a union of classical antiquity and of all sciences and philosophy with religious knowledge. The stand that Luther and Melanchthon took towards the Classics has determined the position of Latin and Greek in Germany to this day.

Space will not permit us to do more than to allude to the other leaders of Protestantism and their connection with antiquity. Zwingli, the Reformer of German Switzerland, was a thoroughgoing humanist before he became interested in Church reconstruction. Even afterward he taught the Classics at Zurich and had them taught in Schools under his supervision. Calvin, whose influence began in French Switzerland and spread to all Protestant Europe not under Luther's sway, began his career with an edition of Seneca's De Clementia. His School in Geneva was thoroughly steeped in Latin and Greek culture, and all the countless institutions which were founded after the model of this

'mother of Huguenot Seminaries' took from their parent a strong classical bias.

We may thus say that the leaders of Protestantism were all of them well trained in the Classics, that they had in fact a knowledge of the ancient literatures which is rare nowadays ever among professional classicists. They appreciated the classical authors and made them a part of their lives. That this heritage of antiquity might not be lost to their descendants they incorporated them in their Schools. Protestantism adopted humanism as its educational standard.

LAFAYETTE COLLEGE, THEODORE ARTHUR BUENGER.

Easton, Pa.

REVIEWS

Latin Sentence Connection. By C. W. Mendell. New Haven: The Yale University Press (1917). Pp. x+ 214. $1.50.

This volume aims to be both psychological and syntactical. Chapter I is devoted to a sketch of the general psychological background of speech as set forth some years ago by Wundt and Morris. According to this view, a thought in the stream of consciousness becomes a subject of observation while yet in a nebulous state, its indefiniteness is cleared up through a process of analysis, and this analysis and its results are mirrored in verbal expression.

To the reviewer it has always seemed that this treatment of the subject is inadequate. Indeed, there is a touch of the grotesque in the portrayal of an individual in whose mind there arises a nebulous thought which comes bit by bit under the lime light of the mind's attention, while the lips, pari passu, utter a running commentary. On that basis it would seem that the unfortunate individual did not really know what he wanted to say until he had finished the spoken sentence or period.

Any adequate treatment of the subject surely must reckon with three undoubted facts: (1) much thinking does not begin in a nebulous haze; (2) much speakingand more writing-does not proceed pari passu with the original marshaling of ideas in the mind of the speaker or the writer; (3) the constructive mental process by which words and constructions are chosen to suggest to the hearer or the reader a train of thought is not necessarily parallel to the genesis of the thought in the mind of the speaker or the writer.

We may illustrate very simply. If a pedestrian is about to step into the path of a swiftly moving car, the thought of the observer will be in no nebulous state as he apprehends that the pedestrian is in danger and that it is imperative that he halt. Yet, even in this crisis, the observer may not speak until the end of his original train of thought is reached; if so, he will probably reverse the order of his ideas as he suggests them to the pedestrian; e. g. 'Stop; you will be run down!' Not only is the order reversed, but the relation between the parts also is different. Probably the original nexus

was more complicated than appears on the surface; but surely there was an illative element involved, whereas in the spoken words the connection is causal.

Obviously it is not the process of inception of a thought that is of special interest to the student of syntax, but rather the mental process whereby forms of speech are chosen as a means of suggesting a train of thought to another person. Even in very informal speech, as shown above, the two processes may be any. thing but parallel, and in formal written expression the gap may be much greater. Take, for example, the poet who may be constrained by the limitations of his verse to say even what he had not planned; and if, as did Vergil at one period, he composes at the rate of a verse a day, a very labyrinth must intervene between 'original thought' and finished product. Nor does even this take into account the profoundly disturbing influence of speech habits.

Professor Mendell does not recognize clearly this very obvious distinction between original inception of a thought and the process by which speech material is chosen to suggest a train of thought to a hearer or a reader. Thus he declares (page 6) that, in language, "the object is to reproduce the original concept in another mind"; and (86) that "sentences, by virtue of the psychological processes which produced them, were of necessity related before ever they were put on paper or even analyzed into words". On the other hand, on pages 28 and 29 some apprehension of the real facts of the case seems momentarily to get the better of his theory.

This confusion of thought as to the general psychollogical background of speech affects little the validity of Professor Mendell's work, for the simple reason that, though he talks much of psychology, his study of Latin sentence connection actually rests but very slightly upon distinctly psychological considerations. For example, in Chapter I much space is given to selecting a definition for the term 'sentence', but in the subsequent discussion this does not seem to lead to any quarrel with the traditional punctuation of the Latin sentences examined. Again, we are carried back to an original stream of consciousness in the mind of a speaker or a writer to learn the fact of sentence connection; whereas the simple truth of the matter, namely, that, by placing sentences in juxtaposition, a speaker or a writer normally implies some connection between them, is a fact so patent that we do not need the help of psychology to recognize it, especially when psychological considerations are so badly confused as they are by the author at this point.

So, too, in Chapter II, which deals with the fundamental categories of sentence connection, it is vigorously insisted at the start that there is imperative need of a background of psychological theory; but, as the chapter progresses, it soon becomes clear (15) that, as a

Here, as below, the italics are the reviewer's.

matter of fact, the investigation is based upon the ordinary principals of syntactical study:

An analysis of the means of sentence connection actually used by Latin writers makes possible a rough division into three groups according to the chief element which gives to each its power to express thought relations and so convey them to the reader.

Such a foundation as this could be laid without even mentioning the word psychology.

The main business of the treatise is to study means of sentence connection, primarily in cases where conjunctions are absent. Chapter II declares that, generally speaking, these means are Repetition, Change, and Incompleteness. In the subsequent chapters, in which this thesis is developed, another confusion of thoughtor at any rate of terminology-is everywhere encountered.

For example, the table of contents for Chapter III, which deals with the subject of Repetition, runs as follows:

Repetition. I. Repetition of Content: its fundamental characteristics; its various types; the relations expressed by it; the conjunctions used to supplement its force. 2. Repetition of Function: its characteristics, typical typical uses, and significance; conjunctions used to supplement its force.

Chapter III itself reinforces at great length this same conception, namely that repetition is a fundamental element in sentence connection, and that, by a study of its phases, there will be brought to light types of sentence interrelation ordinarily marked by conjunctions.

This approach to the subject is likely to arouse expectations that cannot be realized; for, as the chapter progresses, it is gradually revealed that repetition merely marks the fact of connection, with very little indication of the nature of the connection. Quite reversing his first position, Professor Mendell now does not hesitate to recognize the fact that mere juxtaposition of sentences may imply connection, and that the postposition of a sentence may suggest a notion of "sequence". If so, it follows that the rôle of repetition may be but trifling. Compare what is said on pages 28 and 56:

So, when language simply and accurately represents action, each sentence is subsequent to its predecessor both temporally and logically. The description of an action given in the order of its occurrence needs nothing to make clear the relation between sentences. To be sure, the unity given to an action by the constant presence of a given actor is often reproduced by the repetition of a word representing that actor, but this is not essential.

The principle of repetition of content is evidently made use of to indicate in a very general way that the sentence in which the repetition occurs is logically subsequent to the one from which the concept is repeated. Various forces determine into which of three types these instances fall. The second sentence may indicate merely an additional item; or it may indicate the result of the first sentence: or, third, it may indicate the explanation of the first sentence.

It seems at least infelicitous to have introduced into the classification the category of Repetition as a domi

nant factor through which types of sentence relation are to be determined, when, as a matter of fact, its function may amount to nothing more than a somewhat fortuitous emphasis upon a relation whose existence is indicated by the relative position of the sentences, and whose specitic nature is defined by "various forces" apart from the bare fact of repetition.

Within the limits of a brief review it is impossible to sift this matter to the bottom. But it is suggested that the inconsistent and ever-recurring tendency to ascribe to repetition the whole function of sentence connection shakes the reader's faith in the entire chapter, especially when it is found that "repetition" is stretched to cover sentences in which "repetition is implicit", i. e. in which there is no formal mark of repetition (48)2. The question is raised whether it would not be more confidence-inspiring if the problem were approached from the other point of view, that is, if sentences in juxtaposition were examined to determine the probable relation between them, this examination being followed by a study of all the elements that contribute to the reader's impression of the relation. Among these elements the bare fact of repetition would often be found to play a decidedly inferior part.

These criticisms, of course, do not mean that the treatise under examination lacks merit. The psychological background does seem somewhat irrelevant and not clearly thought out; and the emphasis placed upon such factors as repetition is thoroughly misleading; the real merit of the work lies in the care expended upon the large mass of material examined, and in the thoughtful observations upon the (variously defined) types of sentence relation found. Careful reading of the book cannot fail to show its suggestive value. That it is definitive in its method is much to be doubted.

Chapters IV, V, VI and VII deal with Retrospective Incompleteness, Change, Anticipatory Incompleteness and Parenthetic Incompleteness. The method followed is the same as in Chapter III, and it requires no further comment here. Chapter VIII sums up the whole situation, as the author sees it, and forecasts important discoveries through the further application of the method used.

.

The book is not, of course, of a kind suited for reading on a warm summer afternoon. For a work of its character, the style is fairly clear. But there are some blemishes that a subsequent edition should remove, e. g. "It may be possible in conclusion to win to some generalizations ." (14); "Experience has brought it about that the very fact of juxtaposition indicates a relation ." (16); "Also they will make the instances to follow more clear" (38); "a concept behind a noun" (48); "But that subject does not pick an item from the preceding sentence ." (50). Seldom is there real obscurity, as at the end of page 20:

This order of consideration will necessarily divide the discussion of the principle of Incompleteness, but with *Chapter IV returns to the charge in the following words (86): "Repetition is the means employed to define to the reader the particular nalure of the relation".

the elements of the problem so interwoven as they are in the present question, some violence cannot but be done to logical order and distinct advantages will be found in making the differentiation between retrospective and anticipatory means, the factor to determine the order of investigation.

In conclusion, one criticism of the general method of syntactical study now in vogue may not be out of place; and this criticism is made perhaps with a better grace, inasmuch as the reviewer is of the guild, and himself doubtless guilty in times past. Let it be confessed, then, that the attempt to draw psychology into syntactical discussion has resulted rather unfortunately in some respects. Most students of syntax are not psychologists, and they understand none too well the methods of that discipline; yet, as illustrated in the volume here under discussion, the writer of a syntactical treatise feels it incumbent upon him to provide a "psychological background", whether it is really essential or not, or even whether it is carefully thought out or not. Moreover, the effect upon the terminology of syntactical discussion has been very unfortunate. Often a full page or two of stilted exposition, embellished with psychological verbiage, serves to obscure an idea which, for all practical purposes of the discussion, might have been expressed by a terse clear sentence of a dozen words, with no reference to psychology. There is real danger that the general philological reading public will come to pass upon us the verdict which, in a different connection, Cicero passed upon the Epicureans (Tusc. Disp. 2.7): Quam ob rem, quoniam quem ad modum dicant ipsi non laborant, cur legendi sint (nisi ipsi inter se, qui idem sentiunt) non intellego. UNIVERSITY OF CALIFORNIA.

H. C. NUTTING.

The Antique Greek Dance, after Sculptured and Painted Figures. By Maurice Emmanuel. Translated by Harriet Jean Beauley. New York: John

Lane Company (1916). Pp. xxviii + 304. $3.00. Maurice Emmanuel's work La danse grecque antique was published in Paris in 1896. It was the first attempt to study the evolutions of the dance of the ancient Greeks in a logical, rational way. The author collected references in Greek literature to the dance, then studied in Greek sculpture and painting representations of dance movements, which he sought to interpret by comparison with the steps and the figures of the modern French ballet. Emmanuel suffered from lack of archaeological training, and there are consequently many gaps in his accumulation of illustrative material, but the book is unique in its field and its value has been recognized and appreciated for twenty-one years.

The chief qualifications for a translator of a book are, first of all, knowledge of the language into which the translation is to be made, and ability to write that language lucidly and comprehensibly, the next a thorough knowledge of the language in which the original is written, and, finally, some familiarity with the subject treated. The present translator exhibits none of these qualifications. The English used in the