ý'i 'L,

ý, "ý ýe. ý 's , ý.,, rC ,

! ý- ý r44

ý' ý

Ofprint from

ESSAYS IN THE

HISTORY OF ART

PRESENTED TO

RUDOLF \1'ITTKOWER

PHAIDON 1967 ERWIN PANOFSKY

Hercules Agricola :A Further Complication in the Problem of the Illustrated Hrabanus Manuscripts

`Ea memorare volo, explicare nequeo. ' I (W. M. Lindsay) Since the end of the nineteenth century, Hrabanus' Magnentius Hrabanus Maurus, born in 780, received De universo has aroused the interest not only of his education at Tours under Alcuin but spent the palaeographers and students of medieval literature but greater part of his adult life in the Monastery at Fulda also of art historians. Of the c. thirty manuscripts which he directed as Abbot from 822 to 842. After which have come down to us, ranging from the tenth five years (842-847) of withdrawal to the nearby to the fifteenth century, no less than five are illustrated, Monastery on the Petersberg-his own foundation- though only two provide us with a nearly complete he was elected Archbishop of Mayence, where lie died set of miniatures. And these illustrations bear witness in 856.1 to a constant representational tradition still rooted in In his retreat he composed-or, rather, compiled-a the art of classical antiquity. De big encyclopedia entitled naturis rerum (but mostly The two completely illustrated and completely pre- cited as De universo) which he dedicated to his old served manuscripts, separated by about four centuries, friend, Bishop Hemmo of Halberstadt, and a copy of are, first, the `Codex Casinensis 132' (hereafter referred which he offered to King Louis the Germanic. Con- to as `Cas.'), which has never left the Monastery of cerned with `omnibus rebus et quibusdam aliis', this Montecassino where it was both written and illumin- work is largely copied from the Etymologiae (occasion- ated in 1022-1023; and, second, a manuscript in the Origines) by Isidore Seville Vatican ally quoted as of (c. 560- Library, Cod. Pal. lat. 291 (hereafter referred it differs from its in 636); but model several respects. to as `Pal. '), which was written in South or Central is in different Germany The vast material presented a order in 1425 (the explicit even states the precise books instead (there are twenty-two of twenty, and the date: November 8) and is no less fully-as a matter of Hrabanus discussion starts, as thought proper, with fact, a little more fully-illustrated in a charmingly Trinity God, the and the angels instead of with the provincial version of what is now generally called the liberal arts); some of Isidore's entries are omitted; and `International Style of around 1400'. ' most of them are lengthily `moralized' by the applica- interpretatio Christiana foreign Many-unfortunately tion of an still to 3. not all-of its three hundred and sixty- illustrations Isidore. ' one are reproduced in P. A. Amelli, Miniature sacre e profane dell'anno 1023, Montecassino, 1896, but only in Cf. M. Manitius, Geschichte der lateinischen Literatur des 1. chromolithographs which, being hand-made, obscure the style Mittelalters, Munich, ff.; 1911-31, z, pp. 288 III, p. 1o62. of the miniatures and isolate them from the script. Photographic So far I know there is Hrabanus' De 2. as no critical edition of reproductions exist only of some thirty pictures or pages selected based comparative the according universo on a study of manuscripts which to the requirements of a given context. See, e. g., differ in individual but in E. A. Lowe, tend to not only readings also the Scriptura Beneventana, Oxford, 1929,1, P. 344, the be in numbering of chapters and, as will seen, the presence and II, pl. 59; R. Wittkower, 'The Marvels of the East', Journal sentences For Warburg or absence of single or even whole sections. the of the and Courtauld Institutes, v, 1942, pp. 159 if., being the text is from Migne's E. Panofsky, time normally quoted vol. cxi of p1.42b; It Significato nelle arti visioe, Turin, 1962, Latina (hereafter P. L. ), Patrologia referred to as cols. 9-614, p. 52, Fig. 13; and, above all, F. Saxl, 'Illustrated Medieval is reprint of G. Colvenerius (George Colvener Encyclopaedias', Lectures, London, which a straight 1957,1, pp. 228 if., and 11, Colveneer), Magnenti Hrabani Mauri ('p. or ed., opera omnia, pls. 155 a 107' should read 'p. Io6'), c, e ('p. 290' should Cologne, (not in P. L. ). This in 'p. b-d; 1626-7 1617, as stated turn read 291'); 156 157 a, c; 158 a, c, e; 159 a, c; 16o a, c largely the princeps: Rabanus, de ('P. 'P. agrees with editio opus 474' should read 472'), e; 161 a, c, e; 162 a, c; 163 a, c; de proprietate universo seu sermonum et mystica rerum signifi- 164 a, e; 165 a, c; 167 c; H. Giess, 'The Sculpture of the Strasbourg, Adolf Rusch, before Santa catione, probably shortly 1467 Cloisters of Sofia in Benevento', Art Bulletin, XLI, 1959, *13669); incunabulum (Hain-Copinger, photostats of this were pp- 249 if., Fig. 26. For the general problem of illustrated kindly disposal by Dr. Curt F. Bühler placed at my of the medieval encyclopedias, see Sax], loc. cit., and 'Die Bibliothek Morgan Library. In to the text I keep to the Warburg ihr Ziel', Vorträge referring order of and der Bibliothek Warburg, i, chapters adopted in P. L. Isidore's Etymologiae is accessible in 1921-2, pp. i ff.; further, A. Goldschmidt, 'Friihmittelalterliche Latina, Patrologia LXXXII, cols. 74-1054, and in the critical illustrierte Enzyklopädien', ibid., III, 1923-4, pp. 215 if. by W. M. Lindsay, Oxford, Cf. J. Engels, `La Pal, discovered edition 1911. 4. was and excellently described as well as 1'Etymologie Port6e de Isidorienne', Studi medievali, Spoleto, analyzed by P. Lehmann, 'Illustrierte Hrabanus Codices; Series III, i, 1962, pp. i if. Fuldaer Studien, II', Sitzungsberichte der philosophisch-his- [2 0] IV HERCULES AGRICOLA: THE PROBLEM OF THE ILLUSTRATED HRABANUS MANUSCRIPTS 21

Of the three other illustrated manuscripts two are great value were it not for the fact that blank spaces mere fragments: two isolated vellum leaves, probably are left for all the pictures planned but not executed- written and illuminated in West Germany about 1200, which enables us to compare its organization with that which belong to Book IV, Chapters 2-5 (De martyri- of Cas. and Pal. bus, De Ecclesia et Synagoga, De religione et fide, De clericis), in Paris, Bibliotheque Nationale, MS. lat. II 5 disiecta bra (more 17177; and the men: than one That these five illustrated manuscripts are interrelated hundred from pages or clippings pages) of a manu- cannot, I think, be doubted. ' There arise, however, Catalonian hand late script ascribed to a of the two questions. First, do all the manuscripts postdating fairly in fourteenth century, good quality with respect Cas. depend on it, or do some derive from a common illustrations but barbarously to script as well as cut archetype now lost? Second, if we do have to postulate Most, its fragments apart. though not all, of surviving such a parent manuscript, how do we have to imagine by Preussische Staats- were acquired and assembled the the genesis of this chef d'oeuvre inconnu? bibliothek in Berlin (Cod. fol. lat. 930) and are at The first of these questions has been conclusively Library Tubin- present on deposit in the University at answered by Paul Lehmann. He demonstrated that ` The fifth (Vatican Library, Cod. gen. manuscript Pal. and Reg. constitute a family related to but different Reg. lat. hereafter `Reg. ') dates 391, referred to as from Cas., and that, therefore, these two manuscripts from fifteenth the early century and was most prob- derive from a model antedating Cas. And, since Pal. in Italy (it in fact by ably produced was owned an was written and illuminated in Germany, this model Italian humanist The is as early as c. 1450). text was presumably produced at Fulda, perhaps within in complete, except that it abruptly ends the middle the orbit of Hrabanus Maurus himself. ' This Lehmann (XXII, but of the antepenultimate chapter 14); the proved by pointing out that no less than four pictures illustrator died, dismissed or in Pal. rather rustic either was present and planned for Reg. -witness the lost heart after having produced seven pen drawings lapis the beginning. ' Late, scantily and one sketch at 8. This to the Berlin fragments illustrated, indifferently in also applies which, according to written, and careless the Swarzenski, loc. cit., `seem to deviate from the other known rendering of the text, this manuscript would not be of illustrated copies of this work'. It is true that Paul Lehmann, 'Reste einer Bilderhandschrift', cannot see a `direct connection between the Berlin fragments and one of the other manuscripts'; torischen Klasse der Bayrischen Akademie der Wissenschaften but he admits that `not infrequently an affinity is evident' be- Cas. In zu München, 1927, pp. 13 if. The only trivial objection which tween them and my opinion this affinity is so close that direct descent Berlin fragments from may be raised is that the chapter references do not consistently even a of the Cas. might be The destroyed Decapitation refer to either the order adopted in P. L. or to the occasionally considered. miniature, showing a Scene formerly illustrating deviant numeration in Pal. itself. Cf. also E. Panofsky and and the chapter De theatro (xx, 36, F. Sax], 'Classical Mythology in Medieval Art', Metropolitan illustrated in Rotulus, no. 1583, and our Fig. 3), agreed with Cas., (Amelli, in it Museum Studies, IV, 2,1933, pp. 228 if., particularly p. 258, p. 489 pl. cxxx), that showed four spectators, beheaded henchman back and Figs. 37,41; E. Panofsky, Renaissance and Renascences, two victims already and the putting his into Stockholm, 1960, p. 207, Fig. 157. My grateful memory belongs sword the scabbard; whereas Pal., fol. 248 v (here xx, to the late Dom Guy Ferrari, O. S.B., who kindly provided me 35), shows only three spectators, one victim still alive and the henchman his for (our Figs. Cf. with a complete set of photographs. raising sword action 4 and 5). below, Pan (xv, 6) bears in 5. P. Lehmann, 'Mitteilungen aus Handschriften', Sitzungs- also note 17. a staff, as he should, berichte der philosophisch-historischen Abtlg. der Bayrischen the Berlin manuscript, Cod. fol. lat. 930, fol. 73, as well as in Cas., Akademie der Wissenschaften zu München, 1930, PP- 45 ff-; p. 389 (Amelli, pl. cxit), but a snake in Pal., fol. 191 v; the Theseus Labyrinth (xiv, cf. H. Swarzenski, 'Recent Literature, Chiefly Periodical, on and shown at the entrance of the Berlin, fol. lat. Mediaeval Minor Arts', Art Bulletin, XXav, 1942, pp. 287 if., 12; Cod. 930, fol. 65) is, if not a sword-bearing in Cas., (Amelli, least particularly p. 293. warrior as p. 348 PI. LXXXVII), at not a 6. P. Lehmann, 'Reste einer Bilderhandschrift des Hrabanus in monk as in Pal., fol. 170 v (where, incidentally, the chapter is Berlin', Zentralblatt JUr Bibliothekswesen, 1.v, 1938, PP- 173 ff-; numbered 13 rather than 12). If the personification of Terra (xii, lost but in Rotulus, cf. also Swarzenski, loc. cit. Some of the fragments, then thought i; apparently reproduced no. 1582) does in fol. but to belong to a manuscript of Isidore's Etymologiae, were offered makes an orant gesture as she Pal., 142, not in Cas., (Amelli, it in E. van Scherling's Rotulus, A Quarterly Bulletin for Manu- p. 294 p1. Lxxi), should be noted that this orant in Rubrum script Collectors, Leiden, in, 1933, PP. 3 and 8 if. In spite of gesture occurs the very similar personification of the Mare in Cas., (Amelli, Be it the erroneous attribution, this publication is still important in p. 276 pl. Lx11). that as may, on no Berlin fragments that it includes a number of miniatures which the Berlin Library account can the be said to represent a tradition different from is in four illus- was unable to secure and at least one of which (the Decapitation that which reflected the other Scene referred to in notes 8 and 17 and illustrated in our Fig. 3) trated manuscripts. is known to have been destroyed by fire (kind communication of g. Of the Paris manuscript Bibl. Nat., MS. lat. 17177, too little Professor Bernhard Bischoff). is preserved to establish a filiation; but the very fact that it was illuminated in West Germany 7. P. Lehmann, 'Fuldaer Studien, 11', pp. 41 if. The illustrations written and at a comparatively date it, like Pal., depends Cas. begin with I, I (De Deo) and end with in, 1 (De aliis quibusdam early suggests that not on but, directly indirectly, Fulda For Berlin viris sine feminis, quorum nomina in Veteri Testamento scripta or on the archetype. the leguntur). In this case, too, I have to thank the late Dom Guy fragments, apparently more closely akin to Cas, than to the Pal: Ferrari for photographs. Reg. family, see the preceding note.

LE!PZ: G / IV 22 ERWIN PANOFSKY

With spaces left for them in the text-are without parallels regard to the second question-the genesis of be `Hrabanus Cycle'-it is in Cas.: Pal., fol. 117 v (IX, 24 [not `X, 24']), A Heap what may called the now illustrations of Ashes; Pal., fol. 126 v (X, io), The Occupations of generally conceded that the of the-pre- firmly in the Months; Pal., fol. 241 (XX, 4), Four Men with sumably Carolingian-archetype are rooted Horns and Trumpets; Pal., fol. 272 (XXII, 12), the classical tradition and that the models they reflect in Ernst Kitzinger felicitously Sleighs and Their Runners in a Wintry Landscape. were produced what `sub-antique' And it may be added that Pal., in spite of its late date, called the period of the sixth or seventh The brought occasionally avoids iconographical inaccuracies com- century. wealth of comparative material by Fritz Sax113leaves little doubt mitted by the illuminator of Cas., whereas the opposite together about these is is true in other cases. points. What is in doubt, however, whether the In the left-hand column of fol. Igo (XV, 6), for `Hrabanus Cycle' was assembled de novo or-like the example, the illuminator of Pal. correctly equips Mer- illustrations of the Aratea manuscripts, the Psycho- cury with footwings (our Fig. 2) whereas the illumina- machia, the Notitia dignitatum, the Comedies of tor of Cas. (p. 386, Amelli, Pl. CIX, our Fig. i) mis- Terence, etc.-is based upon a cycle already estab- understood these footwings as a complete bird flying lished several centuries before; in other words, whether between the legs of the god. 10In Pal., fol. 187 (XV, 4), the original illustrator of Hrabanus' De universo col- the subject, taken from Exodus 7: 20 (Moses Changing lected his more than three hundred and sixty pictures the Waters into Blood), is much more clearly indicated from a great number of separate sources or merely than in Cas., p. 379 (Amelli, Pl. CIV), even though the appropriated, revised and rearranged a basically illuminator of Pal. forgot to provide Moses with a identical set of compositions originally intended to halo. The Saturn in Pal., fol. Igo, left-hand column, illustrate the parent work of Hrabanus' encyclopedia, has retained the classical sickle; whereas in Cas., p. 386 the Etymologiae of Isidore of Seville. (Amelli, Pl. CVIII), he carries the more modern The first of these alternatives would be at variance Ii scythe. And in the initial miniature of the chapter with the known habits of Carolingian and Ottonian De this gentium (XV, 6) the demoniacal nature of the book illuminators, who normally did not invent, or winged figure on the right is correctly emphasized by assemble from many heterogeneous sources, a whole horns in Pal., fol. 188 v; whereas he looks like an individual figures, Cas. angel in Cas., p. 383 (Amelli, Pl. CVI). and Pal. differ as follows. Cas. shows fifteen species of monsters arranged in four (the Pygmies Conversely, the counterpart of this figure (Silvanus? ) rows are represented by two specimens). Pal. shows, in the left-hand is omitted in Pal., and there are other miniatures column of fol. 75 v, twelve species in three rows of four figures in which, from an iconographical point of view, are less each, and the right-hand column of the same page-a miniature not mentioned by Lehmann Wittkower-two in Pal. than in Cas. Where the illustra- either or nearly correct further species in one row: a Hippocentauress and the Chimaera. tion of XVIII, 5 in Cas., p. 448 (Amelli, Pl. CXXIV), Ten species are common to both Cas. and Pal. Cas. lacks four in Pal.: blessed lower lip correctly shows the nimbed Apollo instructing his son, species present the people with a large enough to cover the entire face as to protect it from the Aesculapius (also in the the so nimbed), art of medicine, sun ('Iabeo subteriore adeo prominente, ut in solis ardoribus corresponding picture in Pal., fol. 222 v, transforms totam ex eo faciem contegant dormientes'); the people whose Aesculapius into And mouths are so small that they can `suck up nourishment only by a patient walking on crutches. ('aliis fora- that the Pan in Cas., (Amelli, Pl. CXII), means of straws' concreta ora esse modico tantum staff of p. 389 mine, calamis avenarum haurientes pastus'); the Hydra; and has become a snake in Pal., fol. 191 v, has already the Chimaera, the two last-named portenta mentioned in P. L., B. Cos., hand, has five present been mentioned. 11 In certain cases it may thus be col. 198 on the other species not in Pal.: the Androgyni (or Hermaphrodites); the Antipodes; the to `reconstruct' by possible the archetype synthesizing, Hippopodes; the Minocentaurs; and the Onocentaurs. We may as it were, the renderings in Cas. with those in Pal. 12 thus assume, with all due reservations, that the archetype exhibited nineteen species (to plus 4 plus 5), distributed, I think, over two miniatures as in Pal. One of these would have shown so. This discrepancy was already pointed out by Panofsky and twelve species in four rows (again as in Pal. ), beginning with the Saxl, `Classical Mythology in the Middle Ages'; for the cyno- Androgyni and ending with the Antipodes. The other, consider- cephalous aspect of Mercury, see below, note 37. In the same ably smaller, would have shown seven species in one row of four miniature, however, Bacchus wears his rightful wreath of vine (showing the Hippopodes, the Pygmies, the Hydra, and the or ivy leaves in Cas.; whereas he appears bareheaded in Pal. Chimaera) and one row of three (containing the more space- Centaur family, i i. See note 8. consuming representatives of the viz., the Minocentaurs, Hippocentaurs Onocentaurs). 12. This may apply not only to the miniatures showing Hercules, the and the Lectures, in Cf. E. Panofsky Mars and Apollo, which form the main subject of this article 13. cited note 3. also and F. Saxl, Dürers Kupferstich `Melencolia. I' (Studien der (see below, pp. 23 ff. ), but also to the illustration of the `monsters' Bibliothek Warburg, Leipzig Berlin, if., described in the chapter De portentis (vii, 7): Cas., p. 166 n), and 1923, pp. 125 now super- by R. Klibansky, E. Panofsky F. Saxl, Saturn (Wittkower, op. cit., p. 173, pl. 42b, our Fig. 6, reproduced seded and and Melancholy; Studies in History Natural Philosophy, from a photograph kindly supplied by the Warburg Institute in the of Religion Art, London New York, (here London); and Pal., fol. 75 v (Wittkower, p. 174, pl. 42e, our and and 1964 p. 309, 89, Fig. 7). Apart from minor discrepancies in the sequence of the n. and passim). IV HERCULES AGRICOLA: THE PROBLEM OF THE ILLUSTRATED HRABANUS MANUSCRIPTS 23 series of compositions for the sole purpose of illustrat- evidence in its favor by examining the text as well as ing a contemporary or nearly contemporary text. We the pictures. But in doing so he soon found himself should find it hard, moreover, to imagine how an artist confronted with a problem which, instead of clarifying probably active at Fulda between, say, 85o and 950 the situation, complicates it even further. He wishes, should have gained accessto so many different pictures, for this very reason, to submit the whole question to all dating from the sixth or seventh century (that is to the judgment di color the sanno. say, precisely from the lifetime of Isidore of Seville), for Trinity which could serve as models the as well as III for the pagan gods; for the celestial bodies and such In Hrabanus' De this (XV, 6) meteorological phenomena as clouds, hail and thunder- chapter gentium the demi-gods (beginning Bel, Beel- storms as well as for an infinite variety of serpents, pagan gods and with Beelzebub the `Pilosi' worms, insects, plants, and minerals; for all kinds of phegor and and ending with and Faunus Ficarius) in the public and domestic buildings (most of them late- are enumerated exactly same described in largely identical in 'classical' in character) as well as for technological order, and terms, as the Isidore's Etymologiae devices from glass ovens to carts and sleighs. Whereas corresponding chapter of (VIII, There is, however, a scriptorium active in the very humanistic, `sub- i i). one notable exception: the Hercules. From Isidore's list antique' environment of Isidore of Seville-whose section on extensive Hercules is i$ With Hrabanus, Etymologiae, we recall, begins with the liberal arts- conspicuously absent the is different would have had no difficulty in producing all the illus- situation quite and very complex. As in P. L., C, trations required by the text on the basis of a tradition printed col. 430 as well as, with minor differences in in the still very much alive in the Iberian Peninsula. spelling and punctuation, editio princeps fol. Against the second alternative, however, there militates of c. 1465, o1vf., the text contains an loosely des- the ineluctable fact that we do not have a single manu- elaborate though somewhat constructed inserted between those Mars Apollo script of Isidore's Etgmzologiae (and their number is cription, of and (the in brackets by legion) which contains illustrations even remotely emendations are supported the `Herculem [ed. pr.: Hercolem] comparable to those encountered in the manuscripts manuscripts): crede- bant deum dicitur Hercoles [should of Hrabanus Maurus' De universo. virtutis: autem Heracles] Graece, heris id litis In view of this negative evidence such eminent scholars read: quasi cleos, est, heris, id lis, as Paul Lehmann14 and Adolph Goldschmidt16 are gloriosus, ab est, et cleos, gloria; vel quasi herocleos [should Heroncleos Heron Cleos, understandably reluctant to postulate an illustrated read: or i)pcLwv Latine fortium Isidore manuscript as the prototype of the `Hrabanus viz., KV OS], quod virorum famam dicimus. Fuit (ut Sextus Pom- Cycle'. But there are those, particularly Fritz Saxl, autem scribit ideoque Anchei [should Augei, for whom the improbability of a `Hrabanus Cycle' peius) agricola: read: freely assembled from many different yet stylistically our Figs. 3-5 and referred to in note 8. There is in the homogeneous them the nothing sources, all of antedating text of this chapter (literally taken over from Isidore, Etymo- middle of the ninth century by more than two hundred logiae, xvni, 42) to suggest so sanguinary a subject. Its presence however, be for by years, outweighs the difficulty presented by the absence might, accounted the assumption that the composition now serving to illustrate Hrabanus' chapter De illustrated Isidore It is possible, of manuscripts. quite theatro was originally intended to illustrate the related and Saxl thinks, that there existed an illustrated edition de nearly adjacent chapter De tragoedia in Isidore's Etymologiae, defined luxe of the Etymologiae, unique and now lost (just as XVIii, 45, where the tragedians are as those who : antiqua gesta atque facinora luctuoso Aratea the Psycho- sceleratorum regum the models of the manuscripts, carmine, spectante populo, coneinebant' (italics mine). This machia, the Notitia dignitalum, the Comedies of chapter, like the intervening chapters De scena and De orchestra Terence, lost), to the (Etymologiae, xviii, 43 and 44), expatiates on the chapter De etc., are which was accessible but by Hrabanus Maurus. In `Hrabanus Cycle' 16 theatro was omitted other words, originator of the we seem to be confronted with a picture purporting to illustrate in Inclined to accept what may be called the `Isidore a text Hrabanus' De universo while actually illustrating a found in Isidore's Etymologiae. hypothesis', 17this writer attempted to look for tangible text only 18. In Isidore's Etymologiae Hercules is mentioned only in other contexts and only in passing. First, he occurs in the chapter De discretione (v, in 14. `Fuldaer Studien, u', p. 46- temporum 39,11, not Hrabanus), where he Enzyklopädien', is have died in `third 15. `Frühmittelalterliche illustrierte p. 219. said to the age of the world'; and, second, Saxl's is in 16, Lectures, 1, pp. 233-9. opinion shared, for example, the two nearly-identical passages concerned with the Hydra: by Hanns Swarzenski (personal communication) and E. Schenk in De portentis, Isidore, XI, 3,34 (hence Hrabanus, vii, 7; P. L., Hrabanus Maur Kapitel Glase', B), in De Isidore, zu Schweinsberg, 'Des us vom col. 198 and serpentibus, xii, 4,23 (hence Glastechnische Berichte, xxxvii, 1964, pp. 129 f. Hrabanus, Viii, 3; P. L., col. 233 A), In both these passages the Isidore hypothesis is `locus i7. One small point in favor of the may per- Hydra defined as a evomens aquas' and both contain haps be found in the fact that Hrabanus' chapter De theatro the sentence: 'quod Hercules videns, loca ipsa exussit, et sic (xx, 36) is illustrated by the Decapitation Scene reproduced in aquae clausit meatus. ' 24 ERWIN PANOFSKY IV (our Fig. 22 These do include a by-form of Augeae which already occurs in Seneca, fol. iii i o) manuscripts Hercules furens, line 248] regis stabulum stercoribus a textual description of Hercules, accompanied by a Cas. Pal. intended be purgasse refertur, quia proprie agricolarum est ster- miniature in and and to accom- in Reg. But description corare agros.' This reads in English: `Hercules was panied by a miniature this held to be the god of virtue. In Greek, however, he is differs from the `printed version' in three respects. Hercules is discussed between Mars called "Heracles", heris cleos, as it were, which means Firstly, not and "famed by strife", from heris, viz., strife [heris, need- Apollo but between Mercury and Mars so that the less to say, is a faulty transliteration of gptg], and sequence is not, as in the `printed version', `Mercury, Mars, Hercules, Apollo' but `Mercury, Hercules, cleos, fame; or else "Heroncleos", which we would call in Latin "the fame of brave men". But according Mars, Apollo'; and it is this order which is observed to Sextus Pompeius, he was a husbandman; and for in the illustrations. Secondly, there are some differ- this reason he is said to have cleaned King Augeas' ences in spelling, one of them crucial; to this we shall stables of dung, because it behooves the husbandman shortly revert. Thirdly, and most importantly, the to fertilize the fields with manure. ' derivation of the name `Hercules' from hens (strife) is I am unable to name the manuscript or manuscripts absent so that the text in Cas., Pal. and Reg. is limited i ný from which this `printed version' is derived and must to what follows (abbreviations expanded and proper ýý-ý} leave its or their identification to the experts. The names capitalized): `Herculem [Cas., Reg. and Stutt- manuscripts which have come to my very limited gart: Erculem] credebant deiim virtutis. Dicitur autem knowledge fall into two classes. Hercules [Cas. and Reg:: Ercules] grece [Stuttgart: In the first class-beginning with what may be the grate] Eracles quasi eron cleos [Pal.: heroncleos; Reg.: earliest extant manuscript but also including a manu- croncles], quod latine 'uirorum fortium faman dicimus. script as late as the fourteenth century-no mention is Fuit autem, ut ýscribit Festus [Pal. and Stuttgart: made of Hercules at all; the sequence of divinities is, Sestus] Pompeius, agricola ideoque Augei [sic in Cas., precisely as in Isidore: `Mercury, Mars, Apollo'. This Pal., Reg. and Stuttgart] regis stabulum stercoribus class is represented, for example, by a tenth-century purgasse refertur, quia proprie agricolarum est ster- manuscript in the Badische Landesbibliothek at Karls- corare agros [the clause beginning with ideoque in- ruhe (Cod. Aug. perg. 68, p. I 13); "' a twelfth-century advertently omitted in Reg. ]. ' manuscript in Paris (Bibliotheque Nationale, MS. lat. 16879, fol. Igo v); i0 a very beautifully written manu- IV script of the same century in the British Museum In Hrabanus' De have (MS. Royal 12 G. XIV, fol. 196) ; 2" and a fourteenth- chapter this gentium, then, we distinguish, far Hercules is century manuscript, probably of Italian origin, in the to as as the treatment of between A', University Library at Leiden (Cod. Voss. lat. F 5, fol. concerned, three strata. In `Stratum by Paris 121). 21 represented : the manuscripts in Karlsruhe, Of the manuscripts the I (Bibl. Nat,, MS. lat. 16879), London, and Leiden, of second class with which am _i familiar, ranging, from I022-1023 to the fifteenth there is, as in Isidore's Etymologiae, no reference to Hercules Cas., century, all but one (Stuttgart, Würtembergische at all. In `Stratum B', represented by Pal., Stuttgart, Hercules I Landesbibliothek, M. S. theol. 2° 45, fol. 202) happen Reg. and a section devoted to is inserted between devoted Mercury to be illustrated: Cas., p. 387 (Amelli, Pl. CX, our those to and Mars but is following (1) Fig. 8); Pal., fol. Igo v-191 (our Fig. 9); and Reg., confined to the statements: `Hercules was the god of virtue'; (2) his name derives from heron [T]pthwv `the fame heroes'; 1g. The Karlsruhe manuscript is divided into two parts, the first cleos KAeog], of (containing Books 1-xii) bearing, somewhat paradoxically, the (3) according to `Festus' or `Sestus' Pompeius, the Aug. For information (Manitius, signature `Cod. perg. g6'. this cleaning of the Augean identifies him Geschichte der lateinischen Literatur stables as a .... 1, p. 300, cites only husbandman (agricola) fertilizes his fields by Cod. Aug. perg. 96) as well as for a photostat of the pertinent who Lauts, Director Staatliche `Stratum page I have to thank Dr. Jan of the manuring them. In C', finally (known to me Karlsruhe. Kunsthalle at only from the printed editions), the section on Hercules A kindly to by Dr. 20. photostat of this page was presented me is inserted between Mercury Mars but be- Robert Klein, to whom I am also indebted in other respects (see not and note 32). Warner J. P. Gilson, British For in 2oa. This manuscript (G. F. and 22. the photograph reproduced Fig. gI am indebted to Old Museum, Catalogue of Western Manuscripts in the Royal the Warburg Institute at London. The fragments in Berlin (Cod. London, f. ) brought fol. lat. Paris (Bibl. Nat. MS. lat. and King's Collections, 1921,11, pp. 73 was 93o) and 17177) do not con- Greenhaus (now Mrs. Lord) folios. The to my attention by Miss Carla to tain the pertinent unillustrated manuscript in Stutt- (fifteenth brought whom I wish to express my gratitude. gart century) was to my attention by Dr. kindly by Florentine Mütherich 21. A photostat of this page was presented to me who also provided me with a microfilm; University. I her both Professor H. van de Waal of Leiden am most grateful to on counts. IV HERCULES AGRICOLA: THE PROBLEM OF THE ILLUSTRATED HRABANUS MANUSCRIPTS 25

tween Mars and Apollo; `Festus' or `Sestus' Pompeius ever, directly appropriated it from Hrabanus. 27 But has become `Sextus' Pompeius; and the text is aug- who is the author credited with the interpretation of mented by another derivation of the hero's name: Hercules as a husbandman and the inventor of `Hercules' means hens cleos, `famed by strife'. manuring ?28 `Stratum A' presents no difficulties. Omitting Hercules Even if the reading 'Sextus Pompeius', adopted in the altogether, and thus refraining from any alteration of `printed version' of Hrabanus' De universo as well as the Isidore text, it maybe held to represent the primary in Boccaccio's Genealogia deorum, 29 were correct, the redaction or `Urfassung' of Hrabanus' chapter De author referred to could only be Sextus Pompeius this gentium. Festus, the well-known Roman lexicographer flourish- `Stratum B' does not, however, seem to be much later. ing near the close of the second century A.D. His De Since it is exemplified by Cas. (dated 1022-1023) as verborum significatu is accessible to us only in one in- well as Pal., Reg. and Stuttgart, this shorter form of complete and sadly damaged manuscript, the so-called the Hercules section must have been present in the `Farnesianus' (Naples, Bibl. Naz., Cod. IV A 3). It `Carolingian archetype' from which the extant illus- had already been deprived of its first seven quaternions ''thetrated manuscripts derive. Nothing militates against (which included the `Hercules' entry) when it was dis- assumption that the insertion was made by covered in 1476 and later lost three further quater- Hrabanus himself; " and this assumption would agree nions which are, however, preserved in Renaissance with the fact that all the sources of this shorter text are transcriptions; for the portion already missing in 1476 classical or late-antique. we depend on scattered quotations and on a much That Hercules was held to be the `god of virtue' was a abbreviated digest (Epitome) composed by Paulus commonplace ever since Xenophon had popularized Diaconus of Montecassino in or shortly after 782 and Prodicus' story of Hercules at the Crossroads, wherein dedicated to Charlemagne. But the whole work was the young hero prefers Virtue to Pleasure. The deriva- available to the Middle Ages from pre-Carolingian tion of his name from the Greek equivalent of virorum times to at least the thirteenth century; what was pre- fortium fama, `the fame of brave men', is found in served of it enjoyed a genuine revival in the Renais- Fulgentius' Mythologiae, II, 2: `Hercules enim Eracles sance, and its author is indiscriminately referred to as [should probably be followed by something like `id est, `Sextus Pompeius', `Festus Pompeius', `Festus', or eron cleos'] Grece dicitur, quod nos Latine virorum even `Pompeius' tout court. " fortium famam dicimus' 26 hence it ; and passed not only 27. Boccaccio, Genealogia deorum, V. Romano, ed., Bari, 1951, into Hrabanus' De universo but also into the mainstream p. 638: 'Rabanus autem in libro de origine rerum dicit, quod Herculem deum dici Herun- the mythographical tradition. It is paraphrased, for cum crederent virtutis, cum quasi of deos, quod Latine fortium famam dicimus. ' Boccaccio by 'Mythographus III', (whether virorum example, the who quotes Hrabanus' De universo no less than twenty-six times. he be identical with Alexander Neckham, with a 28. In a somewhat different form the connection between the Augeas incident invention is implied `Magister Albericus Lundoniensis' or with a great and the of manuring also by Pliny when he writes (Nat. hist., xvi, 50): 'Augeas in Anonym) in latter twelfth rex was active the part of the Graecia excogitasse traditur [scil., the idea of fertilizing the soil century and may be considered the most authoritative by means of dung], divulgasse veto Hercules in Italia, quae regi Stercuto, Fauni filio, the High Middle Ages: `Hercules suo ob hoc inventum immortalitatem mythographer of tribuit. ' ? )Leog, fortium inter- Boccaccio's igitur quasi tp6wv i: virorum gloria, 29. text, quoted in note 27, continues with another in to Hrabanus: 'Et ipse Rabanus Sexto pretatur. '25 And this paraphrase is repeated the specific reference scribit a Pompeio Herculem fuisse agricolam... ' I had Libellus de imaginibus deorum of c. 1400: `Hercules no opportunity to check the manuscripts used in Romano's edition; but 'Sextus' Hercleos, fortium. 72C invariably enim quasi quod est virorum gloria appears in the early printed editions of the Genealogia Boccaccio in his Genealogia deorwn (XIII, i), hoi%- deorum: Venice, 1481, fol. x, 1 v; Venice, 1494, fol. 96; Venice, 1511, fol. 96. 3o. For Festus' De verborum significatu and its transmission, form the by W. M. Lindsay, Sexti Pompei 23. For the reappearance of this shorter of Hrabanus' see admirable edition Festi de Hercules section in another text, the Scholia in Isidorum Valli- verborum significatu quae supersunt cum Pauli Epitome, f. Leipzig, by in cellianä, see below, pp. 26 1913, revised and augmented the same scholar Fulgentii V. C. R. Helm, Leipzig, Glossaria latina iussu Academiae Britannicae Paris, 24. Fabii Planciadis opera, ed., edita, Iv, by Lindsay What- fragments from 1898, p. 41 (apparently overlooked and 1930, pp. 73-467, where many absent the included is for mough, cited in notes 30 and 31). edition of 1913 are and where a good case made in North Italy 25. G. H. Bode, cd., Scriptores rerum mythicarum latini tres, the availability of a pre-Carolingian manuscript (p" For Festus described Celle, 1834, P. 246 (xin, 1). 74). two manuscripts in high-medieval lat. fol. The is library 'Great Catalogue' Cluny, 26. Cod. Vat. Reg. 1290, 5a. text transliterated catalogues, the of written there in H. Liebeschiltz, Fulgentius meta/oralis; Ein Beitrag zur between 1158 and 1161, and the Catalogue of Glastonbury, in M. Manitius, 'Zu Geschichte der antiken hfythologic im Mittelalter (Studien der written there 1247, see Pompeius Festus', Bibliothek Warburg, iv), Leipzig and Berlin, p. 124 (with Hermes, xxv11,1892, pp. 318 if. In the Cluny Catalogue, Festus' is Herclos, obviously a depraved form of Heroncleos, misprinted work described as Liber Festi Pompeii ad Arcorium Rufum, into Hereleos) and reproduced in p1. axcv1, Fig. 45. and in the Glastonbury Catalogue as Liber Pompei de signifi- ERWIN PANOFSKY IV ,26 V There is, moreover, evidence that the Hrabanus manu- `Sextus'. scripts originally had `Festus' rather than While there is every reason to believe that the shorter in This is the way in which he, is referred to, we recall, form of the Hercules section ('Stratum B') goes back to Pomp. ' Cas. and Reg. (see our Fig. z o, here with `Fes. the ninth century-possibly to Hrabanus Maurus `Festus' added in margine by a humanistic hand). And himself-the `printed version' ('Stratum C') must be known is also what we read in the marginal glosses, as of considerably later origin. Scholia Vallicelliana, which are found in an eleventh- As will be remembered, `Stratum C' differs from in century manuscript of Isidore's Etymologiae; these `Stratum B' not only in that Hercules appears-between in scholia Hrabanus' Hercules section, as it appears Mars and Apollo rather than between Mercury and B' (from Herculem to Mars, but in it in- `'Stratum credebant ... ster- also, and more significantly, that corare agros), is taken over-lock, stock and barrel- cludes, in addition to the derivation of his name from Grauso 'by a glossator possibly identical with a certain cleos (fame) and heron (of heroes), the stranger and, as Grausus, supposedly Bishop of Ceneda (not far is from the inter- or We evident awkward repetition of autem, from Bellurio) in 998. are thus faced with the polated derivation from cleos (fame) and hens (strife). fact-apparently unnoticed by the Latinists- So far I know, this derivation be .curious as second cannot that one of the precious Festus fragments found neither traced back beyond the twelfth century where we find in the `Codex Farnesianus' nor in its Renaissance it in two nearly contemporaneous and probably inter- transcripts nor in Paulus Diaconus' Epitome (where dependent texts: William of Conches' Glosses on Hercules is mentioned only as an astrologer), was pre- Boethius and Bernardus Silvestris' Commentary on Six " served for posterity by Hrabanus Maurus. Books of Virgil's Aeneid. `Hercules vero', says William of Conches, `ponitur pro sapiente. Qui bene Hercules cacione verborum. But Paulus Diaconus in his dedicatory letter dicitur, id est lite gloriosus, [sic] enim lis, to Charlemagne (quoted, e. g., in Manitius, Geschichte der -"her" "cleos" lite lateinischen Literatur des Mittelalters, 1, p. 264) refers to the gloria, quia sapiens et pugna contra author as 'Sextus Pompeius' (as in the `printed version' of sapientes gloriosus apparet' ('Hercules, however, Hrabanus' Hercules section and in Boccaccio's Genealogia signifies the wise man, and he is aptly called deorum), and so does Pietro Vettori (Victorius) in Variae "Hercules", lectiones, xvII, 2 (Florence, 1553), as quoted in Lindsay's edition that is to say, a man glorious by strife- of 1913) pp. x11 f. By the middle of the sixteenth century the work of Festus was familiar enough to be quoted in Vincenzo thing which follows stercorare agros representsan elaboration on Cartari's popular handbook of classical mythology, Imagini dei Festus' agricultural interpretation of Hercules rather than a de Dei gli antichi (first published in 1556), with reference to the direct quotation. Be that as it may, certain it is that the first half belief that the chariot of the moon goddess was drawn by the (from Herculem a of scholium credebant ... to ... stercorare mule (Festus, p. 135, line 11, in Lindsay's edition of 1913; agros) is literally identical with the Hrabanus Maurus text and Cartari, p. 106, in the edition of 1571); and, after him, by that the interpretation of Hercules as 'virorum fortium fama' Annibale Caro in his program for the bedchamber of Alessandro comes from Fulgentius. But this identity confronts us with Cardinal Farnese, composed in November 1562. But what jean another problem''. Seznec, to whom credit is due for having observed Caro's That the Isidore scholiast and 'Hrabanus Maurus' should inde- extensive Cartari (The Survival Pagan Gods, New i have identical both use of of the pendently produced two absolutely texts, York, 1953, Pp. 291 ff. ), fails to mention is that Cartari, as so ushered in by a statement to the effect that Hercules was the often, merely transcribes L. G. Giraldi's Syntagmata, first pub- `god of virtue' and both composed of the same excerpts from lished (under a different title), in 1548: L. G. Gyraldus, Opera Fulgentius and Festus, is, of course, unthinkable. We have omnia, Leiden, 1696,1, col. 358. therefore to consider three possibilities: (i) the Hrabanus text b31. For Paulus Diaconus' reference to Hercules as an astrologer, depends on the Isidore scholium; (2) the Isidore scholium see Lindsay's edition of 1913, p. 8g, lines 22 f. As far as the depends on the Hrabanus text; (3) both depend on a common Scholia Vallicelliana are concerned, those which are or seem to source, presumably an Isidore manuscript already containing be based on Festus were incorporated in Lindsay's revised the shorter form of the Hercules section. Possibility number three edition in Glossaria latina, loc. cit.; the Hercules passage, pp. is not too probable because, even assuming that such an Isidore 221 f. They were published in toto by J. Whatmough, 'Scholia manuscript existed (which may or may not have been the case; in Isidori Etymologias Vallicelliana', Bulletin du Cange, 11,1926, cf. below, p. 28), it would have been unique and could hardly PP. 57-75,134-69 (the scholium containing the Festus quota- have wandered from Fulda to an obscure little diocese in North tion added in margine to Isidore, Etymologiae, VIII, II, 50-55). Italy. Possibility number one is even less credible because the A problem is posed by the personality of Grauso (called shorter form of the Hercules section, occurring as it does in Cas. 'Guasone' in P. Gams, Series episcoporum, Ratisbon, 1873, as well as in Pal., Reg. and Stuttgart, must have been present in p. 783) as well as by the continuation of the text after stercorare the `Carolingian archetype' and therefore antedates the Isidore from fact Vallicellian agros: 'Quod mala ab Hesperidibus petisse fertur, pecorum per scholium-quite apart the that the manu- have hoc cura signatur, quae graece mila (j4Aa) dicuntur. Item script does not seem to been exploited or even noticed until We left armenta, cum Geryonis boues abegisse narratur. Per aprum the twentieth century. are thus with possibility number fecisse is hypothesis Vallicellian autem quod supinum portasse fingitur, sues mansuetos !!two, that to say, with the that the his interpretation demonstratur. Per canem tricipitem uenandi studium gessisse scholiast owes acquaintance with Festus' of Hercules husbandman to Hrabanus Maurus' De ostenditur. ' Lindsay accepts the statements referring to the I as a universo Lindsay Whatmough Festus' apples of the Hesperides and to the oxen of Geryones as coming and not, as and seem to assume, to De itself. This hypothesis is from Festus; but then the reference to the Erymanthean boar, significatu verborum concordant Hercules fact Hrabanus' which also attempts to support the interpretation of as with the that work was easily accessible and in Italy times, including the a husbandman, should not be excluded. I suspect that every- widely read at all eleventh century. IV HERCULES AGRICOLA: THE PROBLEM OF. THE ILLUSTRATED HRABANUS MANUSCRIPTS 27 for her means strife, cleos glory-because the wise of William of Conches, Bernardus Silvestris or any man achieves glory by strife and fight with [other] other writer of the past who might have advocated the wise men'). And in Bernardus Silvestris we read: `Per heris derivation; second, that this derivation was not Herculem intelligimus sapientem. Unde nomen con- included in the Hrabanus manuscript or manuscripts gruit. Dicitur enim graece Hercules, latine gloria litis. at Boccaccio's disposal; third, that it must have been "Her" [sic] enim lis, "cleos" gloria' ('By Hercules we interpolated into the `printed version' of Hrabanus' mean the wise man, which is consistent with his name, Hercules section ('Stratum C') either on the basis of for he is called "Hercules" in Greek, and in Latin the William of Conches' Glosses on Boethius (rather than Glory of Strife; for her means strife, cleos glory') 32 Bernardus Silvestris' Commentary on Virgil) or in Neither William of Conches nor Bernardus Silvestris spite of the different spelling-on the basis of Boc- seems to have found immediate followers. 33 Boccaccio caccio's Genealogic deorum itself. " (who, we remember, conscientiously cites Hrabanus Maurus as his source for the derivation of `Hercules' VI from cleos and heron as well as for the passage from The strange eventful history of Hrabanus 'Maurus' Festus) does mention the derivation of the name from Hercules section-originally absent from his chapter eptg, lis; but he explicitly, and doubtless truthfully, De this gentium, later inserted into it in a form which asserts that he owes this second derivation to an oral fuses a passage from Fulgentius with one from Festus communication from Paulus Perusinus (that old, gruff, (thereby saving the latter from oblivion), and finally ugly, and unkempt but immensely learned ex-librarian augmented by the heris derivation appropriated from of Robert of Anjou, who was linked to Boccaccio by a William of Conches or even Boccaccio-may not be lasting friendship): `Paulus [Perusinus] autem dicebat without interest to students of the mythographical Herculem dici ab erix [sic], quod lis, et cleo, gloria, et tradition in the Middle Ages; but for the art historian, sic gloriosus litium' (`Paul [of Perugia], however, used anxious as he is to determine the origin and filiation of to say that Hercules received his name from erix, the Hrabanus illustrations rather than the evolution which means strife, and cleo, which means glory, his of the Hrabanus text, it makes the problem even more name thus signifying "glorious by strife"'). " The baffling than it was before. inferences are, first, that Boccaccio had no knowledge As will be recalled, both Cas. and Pal. contain the earlier, shorter version of the Hercules text as well as a The from William Conches (commenting 32. passage of upon the figure the hero himself, Boethius' De consolatione philosophiae, iv, Metrum 7) is found miniature which shows of in E. Jeauneau, ` "Integumentum" chez Guillaume de Conches', characterized by the lion's skin, the club and, curiously d'histoire Archives doctrinale et litteraire du Moyen-Age, xxxii, enough, a serpent coiling around his right leg. S" And if., from Bernardus 1957, PP- 35 particularly p. 40, n. 5; that in both these he Mars-as he Silvestris (cited by Jeauneau) in W. Riedel, ed., Commentum miniatures precedes Bernardi Silvestris super sex libros Eneidos Virgilii, Greifswald, should according to `Stratum B' but not according to 1924) p. 71 (its content condensed into 'dicitur enim Hercules `Stratum C'-while Mars is followed by Apollo. graece gloria litis latine' on p. 87). As to the relationship between the two passages,the acquaintance with which I owe to the 35" It may be added that the 'printed version' of Hrabanus' kindness of Dr. Robert Klein, I feel that Bernardus Silvestris, Hercules section has. litis gloriosus, which is a little closer to professor at Tours in 1136, used William of Conches (xo8o- both William of Conches' lite gloriosus and Boccaccio's gloriosus 1153 or 1154) rather than vice versa; cf. note 34- litium than to Bernardus Silvestris' gloria litis. A connection 33. See note 34- between the 'printed version' and Boccaccio would seem prob- 34. Genealogia deorum, xui, i (in Romano's edition, p. ' 638). able because an interpolator interested in the pagan gods was For the personality of Paulus Perusinus and his relationship with perhaps more likely to consult a book devoted to this very sub- Boccaccio, see ibid., xv, 6 (in Romano's edition, pp. 761 Q. ject than a Commentary on Boethius; but it goes without saying Here Boccaccio states that he had never seen anything written that it would have to be excluded should a Hrabanus manuscript by Paul of Perugia (although the latter 'ingentem scripsit librum containing the heris derivation turn out to antedate the last quem Collectionum titulaverat') but that he had enjoyed his quarter of the fourteenth century. friendship and instruction for about three years and was thus 36. This serpent presents a special little problem. It may be able to quote his oral communications: 'Huius ego nullum vidi thought to allude to the Hydra (which is, however, hard to opus, sane quicquid ex eo recito, ab eo viva voce referente reconcile with the fact that it has only one head); it may allude percepi. ' Whether Paul of Perugia thought up the a"pigderivation to the two serpents throttled by-Hercules when he was still an himself or, learned as he was, had borrowed it from either infant (which is, however, at variance with the fact that the William of Conches or Bernardus Silvestris is difficult to decide. miniatures show only one snake); or it may be a general symbol If he had borrowed it his more likely source would have been of evil emphasizing the role of Hercules as deus virtutis. I learn William of Conches whose life gloriosus agrees with Boccaccio's from Professor F. Brommer that there are only five or six (that is to say, Paul of Perugia's) gloriosus litium more closely classical examples of Hercules in conjunction with a single- than does Bernardus Silvestris' gloria litis, and whose Glosses on headed snake (one of the nicest in Encyclopedie Photographique Boethius seem to have been more popular than Bernardus de 1'Art, 111,Paris, p. 82 C) in which the significance of the snake Silvestris' Commentary on Virgil. According to K. Heitmann, is equally obscure. Some classicists (e. g., P. Friedländer, `Orpheus im Mittelalter', Archiv für Kulturgeschichte, XLV, Herakles, Berlin, 1907) p. 125) accept the animal as the Hydra; William's 1963, PP- 253 if., esp. p. 276, treatise still exists in C. Robert, Die griechische Heldensage, fourth edition, Berlin, Bernardus thirty-one manuscripts, that of Silvestris only in three. 1921, p. 527, categorically rejects this interpretation. IV 28 ERWIN PANOFSKY Hercules illustration; but is In appearance, the miniature in Cas. (our Fig. 8) is of text and the there nothing justify be faced course much closer to the `Carolingian archetype' to such a postulate, and we should still homogeneous than the thoroughly modernized miniature in Pal. with an apparently miniature composed from different (our Fig. 9), where Mars is equipped with a 'Bohemian two sources. The be based shield', where Apollo has lost his solar halo, and where third solution would upon the premise (as `Hrabanus Cycle' derives from Hercules, gratuitously demonized by animals' ears that the whole an illustrated Isidore's Etymologiae; but in can be observed in the rendering of several other manuscript of divinities, including Saturn, Jupiter and Neptune as this case we should be forced to assume that this lost edition de luxe became Fulda well as the Pluto in our Fig. i), 3' carries a pronged accessible at only after `Urfassung' Hrabanus' De stick instead of the club. But Pal. retains the idea- the completion of the of differed from known suggested by the text and therefore hardly to be universo and that it all the illustrated credited to the illuminator of 1425-of representing Isidore manuscripts not only in that it was Apollo twice: not only as an archer as he appears in but also in that it contained textual interpolations- Cas. ('hunc [Pythonem] sagittarum ictibus sternens') among them the shorter form of the Hercules section by but also as a soothsayer and physician (`divnator et -and that these, too, were accompanied minia- medicus'), equipped with what looks like a medicine tures. " This third solution would account both for the jar but may originally have been meant to be a mortar Fulgentius-Festus combination in the text" and for the with its pestle. stylistic and compositional homogeneity which seems Since, as we have seen, Hercules does not occur in to unify the genetically different elements of the Isidore's chapter on the pagan gods at all and is still miniature; and it would be consistent with the fact absent from the original version ('Stratum A') of the that `Stratum B' invariably occurs in the illustrated corresponding chapter in Hrabanus' De universo, his but only sporadically in the unillustrated manuscripts. appearance in the illuminated manuscripts deepens But, like solution number two, solution number three rather than clears up the mystery surrounding the would pile one hypothesis upon the other. genesis of the `Carolingian archetype'. It confronts us Only a comprehensive and systematic analysis, com- in fact with a choice between three solutions, none of paring all surviving manuscripts and printed editions which is wholly satisfactory. of Hrabanus' De universo with each other as well as The first two solutions would be based upon the with all surviving manuscripts and printed editions of premise that the originator of the `Hrabanus Cycle' Isidore's Etymologiae, might lighten our darkness. To collected his material from all - kinds of different encourage such an analysis, for which our little Then sources. we should either have to assume that Hercules problem might prove to be a good starting the author of the text concocted the Hercules insertion point, is the chief purpose of this woefully inconclusive himself and asked his illuminator to provide a suitable essay. It states a problem rather than solves it. illustration; but in this it be difficult Museum at Cassel (R. Zazoff, 'Gemmen im Cassel', Archäolo 4, i- case would to 1: imagine sCher Anzeiger, LxxXVIII, 1965, p. 88, 48). why he should have confined himself to two no. c22- -c- ý ý ern disparate classical passages as as those selected from excursus inserted into the discussion of the goose (anser) and Fulgentius duck (anas) in Hrabanus' and Festus, and where his illuminator the chapter De avibus (VIII, 6). From the text as transmitted, e. g., in Cod. Leidensis Voss. lat. F. 5 could have found the model for Hercules an unusual (fol. 68 v), in the editio princeps (fol. H. 4), and in P. L. (col. figure which could be to those Mars B), between deprehensus added of and 248 this excursus-inserted ... est and Haec homines like loquacious Apollo without disrupting the stylistic and compo- providos ... and, the section on the woodpecker (Isidore, Etymologiae, XII, 7,46; hence Hrabanus, sitional unity of the original picture. Or we should vier, 6, P. L., col. 247 C), spiced with a quotation from Martial- have illustrated Latin to postulate the existence of an is absent; and so it is from the corresponding chapter in Isidore's Etymologiae (xII, The treatise on mythography, otherwise unknown, which 7,51-52). excursus on the duck-appar- ently another interpolation into the 'Urfassung'-is present, would have provided the model for both the Hercules however, inPaall., fol. 102 v, where it is distinguished by a rubric 37. The cynocephalous Mercury in our Figs. r and 2, described as well as a miniature (rig. 11): 'Anatis caro suavis est pectore Unde "tota as such by Isidore as well as Hrabanus, is, of course, an alto- et cervice. et poeta: tibi ponatur anal, sed pe9tore gether different phenomenon. He owes his dog's head to the tantum/et cervice sapit; cetera redde coco [Martial, x111,51]"' ('The flesh duck is breast Hermes Psychopompos with the Egyptian Anubis of the agreeable with regard to the and equation of hence "a duck (cf. W. H. Roscher, Ausführliches Lexicon der griechischen and the neck; the poet: whole may be placed before but is breast römischen Mythologie, Leipzig, r, r886-go, cols. 2300 ff. ), and you what tastes good only the and the neck; the In Berlin Cod. fol. lat. representations of this dog-headed 'Hermanubis', identified by rest you may return to the cook"'). the is I Cas. the caduceus, occur quite frequently in Hellenistic art. See (in 930 the pertinent page missing, and could not check Reg. beyond Cas., to the instances adduced but not illustrated in Roscher, and ascertaining that too, contains a minia- addition duck (listed but illustrated in Amelli, ff. ) the reproduced in R. Wünsch, ture showing a not p. xv). cols. 2311 magic ring It `Deisidaimoniaka', Archiv für Religionswissenschaft, xrr, r90g, 39. may be significant that the pre-Carolingian Festus manu- by Lindsay (cf. is have if., pp. rg ff.; the relief from the Ileum in the script postulated note 30) supposed to pp. i particularly been Spanish. Museo Capitolino (Stuart-Jones r, pl. gr); and a cameo in the I/ rý- lip., 3ý Icr(2 A!(. 6 ä ý . nt-fý-c mat` Vt. ýý Lte /c Q ä, '`ý-e', /ihAA) / a ý, w-. 'Le r ,l` ch-o 1 7a'-1e; a ýi ur 'J .Zce, c/en, )' c p-e-e--la:

C0-C IV Erwin Panofskv: Hercules -Uricola: The Illustrated Hrabanus Manuscrip/,

It uanlýtiýiaºm. fiqrbir'sui Maurrrýaäuýaýgiiý- a4muini: toll ý$iilC°6t __

ý i

I..

\V. 1), dii, Hrabanus Maui-us, De XV, 6 (D2 this gentium): i. Hrabanus \faurus, De unicerso. li ,, ntimn : universo, Pluto, Bacchus. Vercun". \lontecassino. Cod. 386 Fulcan, Pluto, Bacchus, Alercury. Vatican Library, Cod. Pal. lat. 291, Vulcan, . 13z. P. fol. Igo, right-hand column

1

Iý 11 ýI' ýý

i

ab--- - Itý't1 tiýLl ttýfiilýi" iCttll or Hrabanus Maurus, De Decapitation Scene. . tatiz, erso, 36 (1)e theatro) : `tlt 1.111i. VIontccassino. C.oCI. 1 q2. A\. 3. Hrabanu> Muuru,. D< uni: ý,,; ý. Decaßitation Scene. Destroyed miniature originally belonging to IF .. º -, -: A : tlttöutL . ý. C"P Rýýt-lirn. . Z, rL, ýaü v.;  .. _. .. r,".., ý ; . a'J. wph- ;ý, , +! "tý3Sý, ý. ý ý 5: ,..! t. oil-ttifirma, 'ý:i ,1ý

A\. 1): llralxtnuý Al. c. u;: ý. 1)- ;h tv<ýý1rýý ý" Pal. lat. Decaßilalion Scene. \ atican Library. God. 2w (here XX. 35'- fol. 248v.

6. Hrabanus \laurus, De tmi, ri: cu, VI I. Fifteen Spri irs 2De t)Orterttis: .

(ý Erwin Panofsky: Hercules Agricola: The Illustrated Hrabanus Manuscripts IN

id palptMttattttff. to of tint tnattl týll"1'Hl1h' fI? ttýnb'mrg?l1A t1r1 aü ttldtýcpbruttr buYtuý'tlar ý yunut crtrý.:srptttrýdutc ptltu _ durtttnls till dftrtntnpttdfutni ittfiritp r. rtttrban: tnnmttr'r'trbrutsat-- QtiPptlltdl fttftrtlp. 3411Ct1w fn'lpol ýfAýDi11Ct0?1t''tllttHtdltB nttPlýt tudru trrdtfcr'irittrtta tttntsdnttttd r ý gäfitrrtitý cm fit ttrttrttar itcirra a omw _t_ýtttgPC'4tttt ptte ttbtgptttlPttt ut tlptttltt udttidiu 60tttattb9fgrrrot'attttailcrn 4tttr-fitr uauuum autr u ttd aýiTt1ý'trtauttaý.ýIýä tathMýfý 4suutwuctutlt"ir ptfttit ittntt* fttaltt$ thtYtifrm rcami unuitttxam amm tm fitttwtttf nfaiattft. at. ý# fMrilWt'tggiMipGlttt.. catat* rotcbmrp tutm, uü ' ctý.it grturtýtru uuýo ttýrturý ututo trttdýrlrtldut. pým, crýflr rrý rOCtpundtttd, f; e(rtt>ý ttdfii. pfPtrntxdttlt_týgtta fr arttrnto up" nl'd dto bota IIIPßutütrU f td w. f>a f(b 1tttUIrtYimp m 11ý,ý1Üftltilt' U1pltit'iit'4' aý pbý'týr ý utittrtl" V031'110l tbtf t 0lbtttln. gt(tCtttAýttttrs' ß+ pf ýtq Irt+P4ino fimuv.ýýný: ýfbtg>

.6

I

AIaurus, De VI 1.2 De hunt Species Uonsters. 7. Fl raban us uniceiso, , portentis : tn of Vatican Library, Cod. Pal. lat. 291. fol. 75 v. IV

Illahai.., A1. 1), () 1): dii, 11ý°rýulaýc, liars, Apollo. IIa : 1'. grnlirua: . ýlonn caý: ino. (: od. 132, p. 387

ýt°_. ý'ý"ý''ý"ýa am nm nvttr csttttmfitlooraä ým äýa .ý ýarmtta camstAgs

"Týýty "#*I ý to

.,.. "M T4. " It

ýa-ý-ýcýr 9ýnýo ý-ýý? ýaýf

ýý. ý M ºAý ý 0e0 Stre+l8A .. freity+ _ý Aºfiýc"_0-.f lt1t! i ýttr; ;: r,: üiý: 1'Cttýn"ý, ý_ tü: n: i firrr,: ý ý::i, u Iýl: tnrtC irrnh . ýA Ii e, e . _'ýý ý'iur ausýrrt arýr t3ýattý'eýCa . ýý. Eliht ilu \hLtIrUS, Dr 11111i, A\ ti Ur fwbuiciamoýir 1111, , (Iii., ge'nhunr : it, thvkmaupitep Space intended for Hercules, ýý a miniature showing puqlaa uý gwadý Mars and : Apollo; below, the text referring to Hercules. Vatican Library, Cod. Reg. Iat. 391, fol. III Hral. Mali I). A\ 1). dii, ýý ai, i;, I ýi'. :.... < ,. . !. g

ýý ýý Iflowtv awto_ýýýsý+ lip" do ioaýº -ýft0.0 ý ý, ýmruýý aýo ;fat - #M fca#titrrattfcr. tu*tt. dlrtt- d9Oý. "aftPttu atuctalro ifýrtt''.;lsI simw 1 0, Attttx cr icn ý pem! tmu "rD flfldUd .a0ý

\ III. 1) -l)nrl. V; uican l. ibral Cod. Pal. lat. , 2q),