evidence of the same use afforded by the Chertsey tiles, which present episodes from the Tristram saga1. It is quite natural to find the arms of the royal family of England introduced with intention in a work of English art of the end of the thirteenth century, which also illustrates episodes from the romantic life of Richard Coeur de Lion3. How well they were known by this time is shown by the constant reference made to them by Matthew of Paris', who died in the neighbourhood of 1259. On the other hand, there is a sense of novelty in the criticism made on them by Giraldus Cambrensis, in his De regimine principum, of which the final form was written 1216-17. In his hatred for the reigning Anglo-Normand family of England, the conquerors of Wales, he praised the French kings for adopting the fleur-de-lys as their arms, and the English loss of prestige and territory in France, under John, was occasion for the congratulatory note in the phrase which tells how the fleur-de-lys pardos vincere vidimus atque liones". However, the red field of the cognizance was not adopted at once, and for good, by those most concerned. In 1203 John presented his nephew Otto IV with a basin on which three leopards appeared on a red fields, and this was the cognizance adopted by Henry (1238-53), the ill-starred son of the brilliant emperor Frederick II, and his third wife, Isabella, the daughter of John. Finally, Konrad von Würzburg in his Turnier von Nantes, written before 12690, of which the chief interest is the description of the cognizances of the participants in a tourney, in which the king of England is the centre figure, attributes to the latter the coat of arms with three red leopards on gold", a confusion between the colours of the device and the field.

Unfortunately, Mr Loomis has not made a preliminary study of the use of armorial bearings on which to base a criterion for the value of his special plea. Yet excursions into heraldry can serve a very useful part in literary as well as in historical investigations, and have ready

1 R. S. Loomis, Illustrations of Mediaeval Romances on Tiles from Chertsey Abbey' (University of Illinois Studies in Language and Literature, 11, 2), pp. 50-1; Plates 15 and 16, pp. 49, 52.

2 Ib., 20-1.

3 R. S. Loomis, Richard Coeur de Lion in Mediaeval Art,' Publ. of the Mod. Lang. Ass., XXX (1915), pp. 514-6; Fig. 1, p. 520, Fig. 4.

4Ed. cit., VI, p. 472.

5 Opera, vIII, ed. Warner, 1891, pp. xv ff., 1—lii.

6 Ib., 320-1.

7 Ib., 321.

8 H. Grote, Gesch. der Welfischen Stammwappen, 1863, p. 323.

9 G. Seyler, Geschichte der Heraldik, 1885, p. 274b.

10 A. Galle, Wappenwesen und Heraldik bei Konrad von Würzburg. Zugleich ein Beitrag zur Chronologie seiner Werke,' Zeitschr. f. deutsches Altertum, LIII, pp. 244, 254. 11 Ed. Bartsch, 302 ff., cf. Galle, art. cit. p. 241.

resulted in dating with greater accuracy certain works of the Middle High German epic and courtly poetry1, if an attempt to apply the same methods in dating the two prologues of Chaucer's Legend of Good Women, has not been so successful. Further, a seductive thesis has been presented to show that the author of the Middle High German epic Virginal flattered Otto Visconti, archbishop of Milan, a patron of letters at the end of the thirteenth century, by connecting the youthful exploit of Dietrich von Bern, in destroying a dragon in the act of devouring a knight, with the story attached to the arms of the great Italian family, which represents a living child in a serpent's jaws3. GEORGE L. HAMILTON.

ITHACA, N.Y., U.S.A.

1 E. Schroeder, Zur datierung des Herbert von Fritzlar,' Zeitschr. f. deutsches Altertum, LII (1910), pp. 360-4; Galle, art. cit., pp. 209-259, especially pp. 254-8.

H. Lange, Zur datierung des Gg-prologs zu Chaucers legende von den Guten Frauen. Eine heraldische studie,' Anglia, XXXIX (1916), pp. 347-55; Über die farben könig Richards II von England in beziehung zur Chaucerdichtung. Eine heraldische studie, zugleich ein weiterer beitrag zur legendenprologfrage'; ib. XLII (1918-19), 142–4, 352-6; Beiblatt zur Anglia, XXIX (1918), p. 358;Die legendenprologfrage. Zur steuer der wahrheit,' Anglia, XLIV (1920), pp. 72-7; V. Langhans, Untersuchungen zu Chaucer, 1918, p. 218; Zu Chaucers Legendenprolog,' Anglia, XLIII (1919), pp. 69-90. 3 J. Lunzer, Arena,' Zeitschr. f. deutsches Altertum, LIII, pp. 30-47, 50-4.

M. L. R. XV.

28

REVIEWS.

Ordo Rachelis, by KARL YOUNG. (University of Wisconsin Studies in Language and Literature, No. 4.) Madison, 1919. 8vo. 65 pp. 50 cents.

Of the four types of liturgical play belonging to the Christmas cycle elaborate studies have already appeared of the Stella by H. Anz, of the Prophetae by M. Sepet, and of the Pastores by Professor Young himself, who here presents a careful edition and discussion of all the extant texts of the Ordo Rachelis or Interfectio Puerorum.

This last is, of course, intimately connected in subject with the Officium Stellae, several versions of which end with Herod's order for the massacre, while one from Freising actually introduces the pueri themselves with a song. Thus it is evident how natural an extension of the Stella the Interfectio would be, although, of course, its liturgical occasion is different.

Of the Ordo Rachelis four texts are known: two independent, a very simple version from Limoges (MS. eleventh century) and a very elaborate one from Fleury (MS. twelfth century), one from Laon (MS. thirteenth century) appended to a Stella, and one from Freising (MS. twelfth century) appended to a Pastores. The relation of these texts raises all manner of interesting and intricate questions, a perfectly satisfactory or at least certain answer to which is perhaps beyond the range of the available evidence. Two views have been advanced by previous writers. According to W. Meyer (Fragmenta Burana, 1901) the original was an elaborate and comprehensive version composed somewhere in southern Germany, from which were derived the clearly in many ways unoriginal Freising text and a French recension made under French liturgical influences. From this last sprang the three highly divergent texts of Limoges, Laon, and Fleury, the last being also influenced by Laon. On this view the process is in the main one of disintegration. Anz's view (Die lateinischen Magierspiele, 1905) is the exact opposite of this, since it regards the process as one of development and elaboration. According to him the original was a French version, either Limoges, or more probably one closely similar. The source of Freising arose either independently or as an elaboration of the French original. From that original was also derived, by one or more steps, the Laon text. Fleury represents a combination of the source of Freising with Laon, or perhaps some text intermediate between Laon and the original.

The present editor indicates his own view rather by way of criticism on his predecessors than in a formal or dogmatic fashion. Although he

by no means accepts all Anz's conclusions, much less his reconstructions, he evidently inclines towards his general position, and sums up by saying that we are sure of a French tradition that includes Limoges and Laon and of a German tradition that includes Freising; and in some manner the two traditions seem to be united in Fleury.'

Professor Young is inclined to believe that the Ordo Rachelis originated rather as an extension of the Stella than as an independent officium. It is a difficult question and one not to be decided without great care in definition. But unless we deny to Limoges the status of a play (as in his last paragraph the editor appears to do) it is difficult, in view of the independence of our simplest and earliest text, together with the difference of liturgical occasion, to endorse the view that the Innocents are a bud upon the parent stem of the Epiphany play, rather than a later graft. If Limoges be ruled out, then, no doubt, it may be fairly argued whether the independent or the annexive development be the earlier, but this would appear to be a question of very subordinate interest, especially as the two were probably synchronous.

There is one matter of some importance in respect to which it may be questioned whether the critics of the liturgical drama have taken quite a sound line. In a footnote Professor Young writes: 'It is scarcely necessary to remark that, in general, the dates of the extant manuscripts have small bearing upon the question of the interrelation of the versions. An early version may be preserved only in a late manuscript.' In so far as this is a caution against undue reliance upon manuscript evidence it is very right and proper: in so far as it is advanced as an excuse for simply neglecting that evidence it is less innocent. I cannot speak on the point from any expert knowledge, but I would nevertheless suggest, for the consideration of those who are so fortunate as to possess it, that where an early liturgical dramatic form is found in a late manuscript, it is more often that this belongs to a backward or conservative use than that it is a copy of an early original. In the present case we have four manuscripts, each presenting a totally different version-not four texts of one play, but four quite distinct plays. There is no direct evidence, apparently, that any other copy of any one of these plays ever existed. The manuscripts in which they occur are not of a kind that would be widely multiplied; the fact that the Freising play is written in a twelfth century hand on an erased page of an eleventh century manuscript suggests that opportunities of transcription were infrequent. Is it not likely that transcription would as a rule only take place when extensive revision of the liturgical accretions of the local use were in contemplation, and that the chances are that the new compilation would contain fresh recensions rather than mere transcripts of the liturgical dramas found in the earlier service book? If that is so, each text will, as a rule, be an individual belonging to the surroundings in which it occurs, not simply one copy of a common type. Doubtless there are many exceptions; but if it is true that we must be on our guard against assuming that a text in an earlier manuscript is necessarily more original than one of later preservation, it would appear much more dangerous to argue that we

are at liberty to assign any early date we please to a text irrespective of that of the manuscript in which it is found.

A protest should, I think, be entered against the expansion 'Xpistuc.' W. W. GREG.

LONDON.

Douglas's Eneid. By LAUCHLAN MACLEAN WATT. Cambridge: University Press. 1920. 8vo. xii +252 pp. 14s.

Mr Watt brings to the task of elucidating Douglas's translation unflagging industry and glowing enthusiasm. The comprehensive references to the literature of the subject in the first section of the book (The Man and his Fame) and the scope of the entire study are an attempt to clear Douglas from the unfairness of the casual treatment to which as a minor writer he has been particularly subject, more especially at the hands of the older historians. But most readers of the book will probably feel that Mr Watt in seeking to redress the balance has leant too far the other way. This is not so evident, perhaps, in the comparative estimates made from time to time of the translator's position and worth. Mr Watt admits that his author 'stands far short of the peak of Parnassus' (p. 122) and on several occasions he allows for the mixed tentative character of the work. The want of proportion is rather to be seen in the general tone and treatment, in the comparisons suggested, the quotations from other authors used as illustrations, and, above all, in the metaphors employed. Yet the abundant citations from the text, which allow even the reader who does not know or possess the original to form some judgment of his own, show that there is in the body of Douglas's work an undoubted vigour, at times a certain sense of strong and picturesque phrasing, a certain capacity for first-hand description, which have been perhaps overlooked except in the better known Prologues. These Prologues naturally stimulate Mr Watt's appreciative powers to their uttermost. Metre and meaning of each are analysed in full.

A similar want of proportion or discrimination lies at the root of the criticism which can be brought against the whole of the first and more lyrical portion of the book. The historical importance of Gawain Douglas lies in his transitional character, mediaevalism behind, humanism before. An intimate understanding of what is implied in both these terms is indispensable to the study of such an author. But here the critic and historian find themselves in country none the less difficult because so often traversed. Once more, in a few estimating and qualifying sentences, Mr Watt makes some of the necessary allowances. Of the Renaissance, for example, he says, characteristically: [it] was a process rather than an explosive event. Men could not set their watches by it, but they could float their spiritual emprise upon it' (p. 26). But constantly elsewhere, the lyrical warmth of his description of this comprehensive re-birth provokes reservations in the judiciously minded reader (cf.