UC-NRLF C 2 b73 b33 RAITS OF DANTE BY RANK JEWETT MATHER, JR. THE PORTRAITS OF DANTE K. T>AHTKS^ALljTGHS%tVf^ FLOXBHTJHV^. AUnari Dante in Palatine Ms. No. 320, Biblioteca Nazionale, Florence. A copy from Taddeo Gaddi's lost Fresco and the most Authentic Portrait of the Poet. Princeton Monographs in Art and Archaeology. X THE PORTRAITS OF DANTE COMPARED WITH THE MEASUREMENTS OF HIS SKULL AND RECLASSIFIED BY FRANK JEWETT MATHER, JR., PH.D., L.H.D. MARQUAND PROFESSOR OF ART AND ARCHAEOLOGY IN PRINCETON UNIVERSITY PRINCETON UNIVERSITY PRESS PRINCETON LONDON: HUMPHREY MILFORD OXFORD UNIVERSITY PRESS 1921 Princeton Univebsitt Press Printed by Princeton University Press Princeton, N. J., U. S. A. Published, 1921 TO THE MEMORY OF THE GREATEST OF CHRISTIAN POETS ON THE SIXTH CENTENARY OF HIS DEATH SEPTEMBER 21, I92I ^52570 PREFACE Whoever in this centenary year would study the history and bibhography of the portraits of Dante must perforce read Dr. Holbrook's standard book. Accordingly I have not thought it necessary to repeat unduly the matter provided so richly and agreeably in his pages. Save for the scrupulous use of the measurements of Dante's skull and their graphic application to the problem of the portraits, I can claim little originality for my w^ork. I hope, however, to have put beyond dispute the fact that the Palatine Minia- ture is the most authentic likeness, and to have given a more acceptable account of the sources of Signorelli's and of Raphael's Dante than we have had hitherto. I have tried also to give an improved classification of the minor and eccentric types. Beyond this the Riccardian Portrait and the so-called Death-mask are shown to be of identical profile. While I have sought to give due credit to my numerous predecessors — my debt to Hol- brook especially is great and gladly acknowledged — I have shunned the un- gracious task of noting old errors, natural at the time, the correction of which is now obvious to any careful reader. To Mrs. Emily A. Murray of the Photostat Department of the Princeton University Library I am under deep obligations for her patient and accurate work in reducing the skull and diagrams to precise measurements. Without her aid I could hardly have made my tribute before the Centenary. I wish also to thank my friend Professor Irving Babbitt for reading the proofs. To have had my rough diagrams redrawn by a professional hand would have given them and this book a more workmanlike aspect. But I feared falsification of the delicate measurements in such a process, and have preferred to pre- sent my graphic material in its crude integrity. For this and the graver errors I may have incurred let me plead with the Poet' — Vagliami il lungo studio e il grande amore. The Author. ix TABLE OF CONTENTS PAGE Chapter I. Contemporary Records and Portraits i Limited scope of the study — p. i ; Boccaccio's description of Dante — pp. 1-2 ; was Dante bearded and light haired — p. 2 ; discovery of Dante's bones — p. 3; the reconstruction of the skull from the measurements — p. 4; missing measurements unimportant — p. 5 ; use of the skull diagram for checking purposes illustrated with the so-called death-mask — pp. 5, 6 ; Giotto's Dante, its date and authenticity — pp. 6-8; why is Dante painted so young — p. 9; the discovery of Giotto's Dante in 1840 — p. 10; Seymour Kirkup's service in making copies before the restoration — pp. 10, 11; the color of the hood wrong in all copies — p. 11; close agreement of skull and Giotto's portrait — p. 11; small influence of Giotto's Dante because not grim enough in character — p. 1 1 ; The Palatine Miniature as a copy of Taddeo Gaddi's lost Dante portrait, pp. 11-15; Taddeo might have seen Dante between 1312 and 1321, — p. 15; Taddeo's portrait probably before 1330 — p. 16; comparison of Giotto's and Taddeo's portraits — pp. 16, 17; reconstructions of Dante's features from the skull-plan, very similar to the Palatine — p. 17 ; influence of Taddeo's portrait and its destruction by Vasari — p. 18. Chapter II. Other Fourteenth Century Portraits 21 A warning against Dantes in old frescoes as a class — p. 21 ; Orcag- na's Dante, its toothlessness probably based on oral tradition, apparent influence on Raphael and the Naples Bust — pp. 21, 22; a dubious por- trait in the same chapel — pp. 22, 23 ; a possible Dante in the Spanish Chapel — p. 23 ; alleged Dantes in S. Francesco, Ravenna, one echoed in Pietro Lombardi's tomb relief — p. 24; a possible Dante of 1323 by the Bolognese notary, Ugguccione Bambaglioli — pp. 25, 26. Chapter III. The Fifteenth Century, the Iconographic Tra- dition and its Growth 29 On the importance of little things in iconography — p. 29; tabulation of iconographic points of the Dante portraits of Giotto, Taddeo Gaddi, and Orcagna — p. 29 ; sterility of Giotto's tradition — p. 30 ; probable derivatives in Andrea Bonaiuti, a miniature, and Vasari — pp. 30, 31; Dante(?) in a Christ Church drawing of about 1400 — pp. 31, 32; The Riccardian Por- trait, a Renaissance revision of Taddeo Gaddi's — pp. 2^, 33; possibly by Paolo Uccello and a replica of the portrait made for the Cathedral in 1429 — pp. 34, 35 ; derivatives — the bronze Medal, the Death-mask, the Trivulzian and Eugenian Dantes — pp. 36, 37 ; Botticelli's illustrations — p. 38; The Naples Bust, its fidelity to the Taddeo Gaddi tradition, probably Florentine early i6th century, other views — pp. 39-41 ; con- clusions reached in the chapter — p. 41. xii CONTENTS PAGE Chapter IV. Fifteenth Century, Three-Quarter-Face Types AND THEIR DERIVATIVES 45 Three-quarter-face types in early Mss. — p. 45; Andrea del Castagno — p. 46; a Yale cassone — pp. 46, 47; Domenico di Michelino, 1465, for the Cathedral — p. 47; source of a Bronzino(lost) but represented by copies — p. 48; which was popularized by Raphael Morghen's print of 1757 — pp. 48, 49; copy of a lost Dante drawing by A. Pollaiuolo, at Christ Church — p. 49; a "Signorelli" drawing at Berhn and a Manuscript Miniature at Turin — p. 50. Chapter V. Irregular Types 53 On the variety of Dante types in Ms. miniatures and in cassoni — P- 53 ', two types in a single Florentine miniature — one haggard, the other young, straight nosed and toothless — pp. 54, 55 ; Giovanni dal Ponte's Dante at Harvard — pp. 55, 56; the Vatican Urbinate Ms. and Signorelli at Orvieto — p. 57; Pesellino, the painter of the Uffizi jardeniere, an in- tarsia by Benedetto da Majano: all grim types — p. 57; Filippino Lippi's Dante in the Brancacci, straight nose, round ear-tabs — pp. 57, 58; Justus of Ghent at Urbino — p. 58; derivative the "Morris Moore Raphael" — p. 60; Raphael's three Dantes, their very composite origin and iconographic irregularity — pp. 61-63; the Venetian woodcuts of 1521 and 1529 — pp. 63, 64; Bernardino India's portrait, its character and popularity, a deriva- tive ascribed to Giambellino — pp. 64, 65 ; the Munich drawing and the Sessa woodcut of 1564 — pp. 65, 66; Conclusion — the three Dantes yet to be created — p. 66. Bibliography 67 Appendix I. On the Reconstruction of the skull from the Official Measurements, Front and Side Views 69, 71 Appendix II. Tacca's "Head" of Dante 72, 73 Genealogical Tables of the Portraits 74, 75 A Summary Catalogue of Dante Portraits 79-85 LIST OF ILLUSTRATIONS The Palatine Miniature Frontispiece FIGURE PAGE I' — Skull of Dante, side view 3 2 — Skull of Dante, front view 4 3 — The Death-mask checked by the Skull 6 4 — Giotto's Paradise in the Bargello 7 5 — Seymour Kirkup's copy of Giotto's Dante 8 6 — Giotto's Dante, Kirkup's small color sketch 9 7 — The Arundel Print 9 8 — Faltoni's copy of Giotto's Dante lO 9 — Giotto's Dante as now repainted by Marini . , lo lo — Giotto's Dante checked by the Skull 12 1 1 — Head by Taddeo Gaddi 13 12 — The Palatine Miniature 13 13 — Head by Taddeo Gaddi 13 14 — The Palatine Miniature checked by the Skull 14 15 — The Palatine Miniature overlaid on Giotto's Dante 16 16 — Portrait of Dante reconstructed from the Skull 17 17 — Portrait of Dante reconstructed from the Skull 17 18 — The Palatine Miniature 17 19 — Orcagna's Dante 22 20 — Alleged Dante in Orcagna's Paradiso 23 , 21 — Dante ( ?) by Andrea Bonaiuti, Spanish Chapel 23 22 — Alleged Dante in S. Francesco, Ravenna 24 23 — Dante, Pietro Lombardi's marble relief on the Tomb 24 24 — Dante (?) Ugguccione Bambaglioli's sketch of 1323 25 25 — Grassa Bologna, from the same document 25 26 — Dante by or after Vasari at Cleveland 30 27 — Drawing of Dante( ?) at Christ Church, Oxford 31 28 — The Riccardian Portrait, Early Renaissance Dante 33 29 — The Renaissance Medal 34 30 — The Riccardian Portrait compared with the Death-mask 35 31* — Botticelli, illustration for the Paradiso 37 31' — Botticelli, illustration for the Paradiso 37 32 — The Naples Bust, front view 38 xiii XIV ILLUSTRATIONS FIGURE PAGE 33 — The Naples Bust, side view 40 34, — The Naples Bust superimposed on the Palatine Miniature 41 35' — Dante from a Laurentian Ms. of about 1400 45 35"^ — Andrea del Castagno's Dante 46 36 — Dante from a cassone front in the Jarves Collection, Yale University 47 37 — Domenico di Michelino's Dante, Cathedral, Florence 48 38 — Head of same 49 39 — The Tofanelli-Morghen Print 49 39" — Dante, old copy of a drawing by Antonio Pollaiuolo 46 40 — The Yale Dante 49 41" — Dante ( ?), drawing by Piero di Cosimo, Berlin 50 41"^ — Dante from a Ms. in the University Library, Turin 50 42 — Dante before the Muse Calliope, Ms. Miniature at Venice 53 43 — Dante in the selva oscura, Ms. Miniature at Paris 54 44 — Giovanni dal Ponte's Dante at Harvard University 56 45 — Dante with Paolo and Francesca from a Vatican Ms 55 46 — Signorelli's Dante at Orvieto 56 47 — Dante from Pesellino's Triumph of Fame, Fenway Court, Boston 57 48 — Benedetto da Majano's Dante, intarsia, Palazzo Vecchio 58 49 — Dante from a Triumph of Fame, furniture panel, Uffizi 57 50 — Filippino Lippi's Dante, fresco in the Brancacci Chapel 59 51 — Dante by Justus of Ghent, Louvre 60 52 — Alleged Dante, fresco in Sant'Agostino, Rimini 61 53 — Dante, "The Morris Moore Raphael" 60 54 — Dante, fresco from Raphael's Disputa 61 55 — Drawing for Raphael's Dante in the Parnassus 62 56 — Raphael's Dante, fresco, in the Parnassus 61 57 — Woodcut from the Convivio of 152 1 63 58 — Woodcut from the Divine Comedy of 1529 63 59 — Engraving after Bernardino India's Dante 64 60 — Engraving after a lost portrait ascribed to Giambellino 64 61 — Dante, sixteenth century drawing in the Munich Print Room. . 65 62 — Woodcut from Sessa's "Divine Comedy" of 1564 65 63 — Terra cotta head of Dante in the Galletti collection, Florence ... 72 DIAGRAM I — Dante's Skull reconstructed to scale from the official measure- ments, side view 68 II — Dante's Skull, front view, reconstructed to scale from the offi- cial measurements 70 CHAPTER ONE CONTEMPORARY RECORDS AND PORTRAITS CHAPTER ONE CONTEMPORARY RECORDS AND PORTRAITS Boccaccio's Testimony. The Reconstruction of Dante's Skull from the Measurements. Giotto's Dante. Taddeo Gaddi's Dante the Most Authentic Portrait. It is not my purpose to review the numerous Dante portraits. That has been well done by Paur, Franz Xaver Kraus, Richard Thayer Holbrook, Parodi, and lately by Passerini. I wish rather to find among the many, the portrait or portraits that actually resemble the poet, and then to con- sider the origin of the finer artistic types. The two classes do not neces- sarily cover. One may agree with Dr. Holbrook that only Giotto's youthful Dante at Florence and the austere bust at Naples are "Dantesque." That only means that they satisfy our imagination. They may or may not look like Dante. Indeed we shall find that one of the more ugly and artistically negligible portraits actually preserves the true proportions of the face. That need not surprise us. Charles Willson Peale's and Gilbert Stuart's Washington are so different that we hardly recognize the same individual. Peale's, according to all evidence and probability, presents the Virginia squire as his contemporaries actually saw him, but Stuart's masterly head is the only one that posterity has accepted as a worthy likeness of the father of our nation. There exists no portrait painted in Dante's lifetime, but fortunately his devout admirer and commentator Boccaccio consulted the memories of certain friends and contemporaries. Boccaccio visited Ravenna in 1346, twenty-five years after Dante's death, and again in 1353. On both occa- sions he saw men who knew the poet in his last years. Boccaccio writes, (I use Dr. Holbrook's translation) : "Our poet was of middle height, and in his later years he walked some- what bent over, with a grave and gentle gait. He was clad always in most seemly attire, such as befitted his ripe years. His face was long, his nose aquiline, and his eyes rather big than small. His jaws were large, and his lower lip protruded. His complexion was dark, his hair and beard thick, black, and curly, and his expression ever melancholy and thoughtful." These somewhat vague but welcome details Boccaccio got from friends of Dante including Dino Perini, Andrea Poggi, and Pier Giardino. Boc- 2 ' ' TliK" ^OiRTRAITS OF DANTE caccio oddly fuils to nieiiti{>n:' any portrait of Dante. Since the two con- temporary portraits do not show the protruding under lip, we may suppose that this trait was originally not prominent, but became more marked with old age and toothlessness. Boccaccio's informants saw Dante only in his last years. Dante speaks of himself as being bearded and as having tawny hair in youth. In the Paradiso, xx, Beatrice tells him to "lift up his beard," that is to look at her. The natural inference is that when he was writing these lines, probably about 1320, he wore a beard. It seems to me over-ingenious to suppose with several critics that Dante was bearded at the imaginary time of his vision, 1300, All the portraits of any authority, and at least one seems to represent him after 13 12, show him smooth shaven. We must suppose that he occasionally wore a beard and did so in his last years when Boccaccio's informants knew him at Ravenna. For that matter Boc- caccio's tradition that Dante let his beard grow after Beatrice's death may be true. Generally he went beardless like most patricians of his age. In a Latin correspondence with Giovanni del Virgilio, in 13 19, Dante speaks of his whitening hair as once tawny.^ This is apparently at variance with Boccaccio's account. But this may mean nothing more than that his hair, originally rather light, darkened with age — a very common case. Such is the scanty historical evidence for Dante's looks. Plainly it does not go far as a check upon the various old portraits. To select the most resemblant one on any such basis would be to make a very subjective choice. Happily we are not reduced to that. In the measurements of Dante's skull we have objective evidence of a most valuable sort — a real check upon the various portraits. My purpose is simply to make the fullest use of this evidence, which has been strangely neglected. 1 Dante, Eclogue i, to Giovanni del Virgilio, 11. 33 flf. "Me vocat ad frondes versa Peneide cretas. Quid facies? Meliboeus ait. Tu tempora lauro Semper inornata per pascua pastor habebis? O Meliboee, decus vatum quoque nomen in auras Fluxit, et insomnem vix Mopsum Musa peregit. Retuleram, quum sic dedit indignatio vocem: Quantos balatus colles et prata sonabunt, Si viridante coma fidibus paeana ciebo ! Sed timeam saltus, et rura ignara deorum. Nonne triumphales melius pexare capillos, Et, patrio redeam si quando, abscondere canos Fronde sub inserta solitum flavescere Sarno?" Meliboeus is Dante's friend Dino Perini, Mopsus is Giovanni del Virgilio. Dante values the offer of the poet's laurel but fears to go to Bologna and asks if it would not be better to be laureated in his fatherland, wreathing the white hair which once was tawny on the Arno. For a careful discussion of this doubtful passage see Holbrook and Parodi. THE PORTRAITS OF DANTE 3 On May 2y, 1865, certain masons working in a close adjacent to the Church of San Francesco at Ravenna accidentally hit a rude coffin. As they cleared it, they read Dantis ossa denuper revisa die j Junii i6y'/'. That is "Dante's bones again inspected 3 June 1677." As a result of this dis- covery, the tomb of Dante was opened and found empty save for a few dried laurel leaves and some small bones missing from the skeleton. In- deed for nearly three hundred and fifty years the venerated shrine had been empty. What had happened was this : After negotiations repeated through the fifteenth century to recover the ashes of her exiled poet, Flor- FiG. I. Side view of Dante's Skull drawn from the official measure- , ments. Teeth and lower jaw restored. ence found in the election of the Medici Pope, Leo X, the necessary po- litical support. In 1 5 19 he ordered Ravenna to surrender the skeleton to Florence. The plan was thwarted by the friars who had custody of the tomb. Cutting through the cloister wall against which the sarcophagus stood, they removed the bones and buried them secretly. In this hurried translation, a few small bones were still left in the sarcophagus and the lower jaw was lost. For a hundred and fifty years the friars seem to have had the custom of occasionally verifying their treasure. After the last inspection recorded on the coffin in 1677, the memory of the afifair seems to have died out in the convent. There remained only the persistent rumor at Ravenna that Dante's tomb was empty. ^ The discovery of the skeleton on Dante's birth year naturally created an immense enthusiasm at Ravenna. The bones were reassembled into a skeleton which lay in state in a glass coffin until the formal reinterment in the original sarcophagus. The skeleton was that of a man five feet five 2 All this most interesting story is fully told in Ricci, pp. 338 ff. For all matters con- nected with the discovery of the skeleton and the measurements consult Ricci "L'ultimo Rifugio" passim. 4 THE PORTRAITS OF DANTE and one half inches high. Unhappily the Commune of Ravenna regarded its relic with jealousy and declined either to have the skeleton photographed or a cast made of the skull. Instead they commissioned local physicians to measure the bones and draw up a report. It was dated June 12, 1865, two weeks after the discovery. There had been time for careful work, and it is not to be supposed that so important a task was slighted. The measure- ments have been used by Welcker and Holbrook (pp. 46-47) to settle the case against the so-called Death-mask of Dante. Fig. 2. Front view of Dante's Skull drawn from the official measure- ments. Teeth and lower jaw restored. The trouble with skull measurements, as Dr. Welcker pointed out, is that they "do not enable us to see a skull." As I studied the measurements, however, I was satisfied that the skull could be modelled from them with entire accuracy. So I tried to interest that remarkable sculptor-ethnologist, Dr. McGregor of Columbia, whose restorations of prehistoric skulls and heads are famous. To interest him was easy, but he was going to Europe. So I had to give up the best plan — that of modelling the skull, and set myself to the task of putting the measurements into a profile drawing. As I worked away, it became clear that measurements made graphic can go far towards making us see any skull. What I got as the reader may judge (Fig. i) was a reasonably accurate diagram of the profile, accurate enough to serve as a check upon the transmitted portraits; and I also put up a front view (Fig. 2), which, although the measurements do not permit an THE PORTRAITS OF DANTE 5 equal accuracy, is also reproduced. The method of draughting off the measurements is fully explained in Appendix I. The results were reassuring. The outlines joined properly; hence the measurements were correct. They did not produce, as Welcker had feared, a monstrous skull, but one entirely normal. Indeed I found in the Anatomi- cal Museum of Princeton University a skull so close in all the facial meas- urements that I used it to supply smaller measurements omitted in the Ravenna report. In fairness, all such theoretical measurements are here avowed. The skull of Dante was found toothless and without the lower jaw. Since most portraits represent Dante with teeth, these with the lower jaw have been supplied from the Princeton skull, from which also the cheek bone has been drawn in. The small measurement from the point betwixt the eyebrows to the suture at the base of the nasal bone is lacking in the official report. From the Princeton skull I have taken it as a centimeter. The possible error is two or three millimeters. Again the indentation of the nose beneath the brow is uncertain. Apparently it should be a little greater than I have made it, if we may accept the testimony of the best portraits. A somewhat more serious lack is the measurement from the bottom of the nasal orifice to the upper jaw (alveolar). This again I have taken from the Princeton skull at 2.5 centimeters. The uncertainty of this measure- ment produces a similar uncertainty in the slope of the face (facial angle of Camper, 79° 34')- However such inaccuracies at most amount to an upper lip and forehead a sixteenth of an inch or so different from the plan. There is nothing that would change the general aspect. I may note a dis- crepancy in Dr. Nicolucci's remarks. In his general description of the skull he speaks of the forehead as rising vertically. Of course no forehead liter- ally does this. But since the facial angle is 79° 34' — indicating a normally slanting forehead — we must imagine either that he was holding the skull wrongly oriented when he was making the observation, or writing care- lessly; and piously but unscientifically crediting the altissimo poeta with a brow of corresponding altitude. Before applying the diagram of the skull to the portraits, I wish to re- peat, that everything in the reconstruction is either from the official measurements or from the Princeton skull. The portraits have been en- tirely disregarded. The method of comparison between the skull-plan and the portraits is the simplest. The profiles are taken off in outline, and re- duced to nearly identical scale (the skull being made a shade smaller). The agreement or disagreement of the matched profiles tells the story. Natu- rally an exact agreement between a freehand drawing and a measured drawing of the same head is not to be expected. We must be prepared to find small discrepancies between the reconstructed skull and the authentic 6 THE PORTRAITS OF DANTE portraits. But these discrepancies should not be large. On the scale of our reproductions (about >^ of a diameter) differences of a sixteenth of an inch are serious. Let us test the matter with the so-called Death-mask, which as a fine piece of modelling and a characterful thing still holds a certain authority despite the observations of Welcker and Holbrook. We make the join at the mouth, neglecting the reconstructed lower jaw of the skull. The eye, the root and the base of the nose of the Mask are all over Fig. 3. The Skull (dotted line) over- laid on the Death-mask. The misfit about the nose, eye and brow is obvious. half an inch lower than the corresponding points on the skull (Fig. 3). The maker of the mask shortens the chin and upper lip and heightens the fore- head. The Death-mask in short is not a cast from the face nor directly based on one but an artistic improvement (and perversion) of the tradi- tional portrait type. Having thus illustrated the method of comparison, I may now apply it to the two portraits which were painted by contemporaries of Dante, — Giotto and Taddeo Gaddi. Unhappily we know these portraits only through copies, but we have every reason to think the copies faithful. Giotto's portrait was painted in the chapel of the Podesta (Military Gov- ernor of Florence) probably between 1334 and 1337. It represents the poet as a man in the late twenties amid the saved souls in Paradise (Fig. 4). The portraiture is of notable ideality and beauty. It should be unnecessary to enter into the needless controversy that raged about the authorship and date of this famous portrait. Amateurs of such battles long ago may find the strategy and tactics spiritedly recounted by Dr. Holbrook (Chapter IX). We have the best contemporary evidence for believing the portrait to represent Dante and to be by Giotto.^ Moreover the admirable style of 3 Our firm witness to Giotto's authorship of the Dante in the Bargello is the town crier and poetaster Antonio Pucci, who was about Florence when the portrait was painted. A general disposition to doubt so good a contemporary record is merely an in- heritance of the queasiness that Milanesi and his generation introduced into the criticism of Italian painting. I quote the biographical part of Pucci's rimes from the best text, THE PORTRAITS OF DANTE Fig. 4. The Paradise in the Bargello from Holbrook, sketched by Kirkup. Dante is seen behind the figure kneeling at the right. this and the accompanying portraits in the great fresco is beyond the ca- pacity of any other known Florentine artist of the time. Nothing is to be made of the fact that most of the frescoes of the chapel, representing that of Magliabecchiana, Var. 1145, as printed by Paolo d'Ancona in "Scritti Danteschi," Florence 1913, p. 54i, note. "Questo che veste di color sanguigno Posto seguente alle merite sante Dipinse Giotto in figura di Dante Col braccio manco avvinchia la scrittura Perfetto di fatteze e qui dipinto Com' a sua vita fu di carne cinto." "The one who dresses in blood-red color, placed behind the holy saints, Giotto painted in Dante's likeness. . . . With his left hand he grasps the book. . . . He is here painted perfectly as his features were, as he was in life when still in the flesh." It will be noted that this description entirely corresponds to Dante's action and position in the great fresco of the Bargello. Pucci's testimony as to the likeness of the portrait is not very valuable for he may never have seen Dante, and surely not at the age repre- sented by Giotto. Pucci was born about 1300. THE PORTRAITS OF DANTE Alinari. Fig. 5. Seymour Kirkup's copy of Giotto's Dante in the Bargello. A combination of the original tracing from the fresco and the small color sketch. The basis of the Arundel Print. THE PORTRAITS OF DANTE legends of the titular Saint, Mary Magdalen, and St. John the Baptist, are palpably by a weaker hand. Giotto, busy at the moment with the Campanile and other tasks, apparently turned over the mass of the work in the Magdalen Chapel to an assistant who had already aided him in the Lower Church at Assisi. Such is Mr. Berenson's view. ("Florentine Painters of the Renaissance" 3rd Ed.) I differ from him only in the stronger conviction that the portraits in the lower rank of the Paradise were painted by Giotto himself. Even the radial arrangement of this throng is From Holbrook Fig. 6. Kirkup's small color sketch made on the inside vellum cover of his copy of the "Con- vivio," 1521. Historical Museum, Florence. Fig. 7. The Arundel Print. thoroughly characteristic of Giotto, being like that in the famous Baron- celli altar-piece, the design of which, at least should be Giotto's. Why Giotto painted Dante so young is uncertain. He may have meant to represent a soul in heaven where there is no old age. He might have used a sketch made years before. Giotto surely saw Dante at Padua in 1304-5 when the poet was forty, and probably he had also known him at Florence some ten years earlier. In the interval Giotto had been working at Assisi and Rome. If Giotto had a sketch for this portrait — and we shall see that it is probable he had — it was pretty certainly made before he left Florence, at a time when Dante was not older than thirty, and just coming into prominence as a politician and lyric poet. For the youthful appearance of Dante Professor Venturi has advanced a characteristically genial hypothesis, Storia, vol. V, p. 448 f. Since Dante had been banished 10 THE PORTRAITS OF DANTE under penalty of death in 1302, it was unseemly to represent him in a public building at Florence except at an age prior to his condemnation. It seems strange that in view of Dante's rising fame the old ban should have held fifteen years after his death in 1321. Yet the mediaeval Florentines were sturdy haters, and may well have hit on this way of honoring their poet without discrediting their own act of proscription. Its intrinsic beauty and the romance of its discovery have made Giotto's Dante one of the most famous portraits in the world. Somewhere about the end of the sixteenth century the Palace of the Podesta became a prison, the chapel was turned into a storeroom and its frescoes whitewashed. In 1840 the Piedmontese patriot and refugee, Aubrey Bezzi, with the Ameri- From Holbrook Fig. 8. Faltoni's Copy made before the Restoration. Fig. 9. Giotto's Dante in the Bar- gello as repainted by Marini. can man of letters, Richard Henry Wilde, and an eccentric English artist, Seymour Kirkup, joined forces to uncover the portrait. Bezzi engineered the difficult formalities with the Grand Ducal administration. A wretched painter-restorer was assigned to them, one Marini, and the fading frescoes of the chapel everywhere testify to the brutality with which he did his job. At length they found the portrait minus the eye through ]\Iarini's care- lessness. Soon Marini put back unskilfully the eye he had already put out and smeared the beautiful face with unnecessary and disfiguring repaints, altering as well the form of the hood for good measure of vandalism. Before this chagrin befell the discoverers, however, Kirkup had saved the day. Foreseeing the restoration, he bribed his way into the chapel and made two copies of the fresco. One was a tracing on mica, the other a small color sketch on the inside vellum cover of his copy of Dante's ''Con- THE PORTRAITS OF DANTE ii vivio" (Fig. 6). He combined these copies in at least two versions (Fig, 5), one of which was carefully facsimiled in lithography for the Arundel Society (Fig. 7). This is the standard version of Giotto's ''Dante." Kirkup wrote to Gabriele Rossetti that he had put "nothing of my own" into these copies, and Dante lovers, who owe him an inestimable debt, will take his word for it. Moreover, the fidelity of Kirkup's work is at- tested by the independent copy which the sculptor Perseo Faltoni (Fig. 8) made before Marini (Fig. 9) had defaced the portrait. The agreement of Kirkup's two copies, the Arundel Print and Faltoni's copy should set at rest the uneasiness of such as find the evidence for Giotto's Dante second- hand. Second-hand it is, but all the same trustworthy, as the reader may judge from the reproductions here supplied. All the copies have a slight error in representing the hood as light with a red reversed band. It was once uniformly red, as Calvalcaselle has shown. ^ Marini scraped most of the red off, exposing the gray preparation. It remains only to say that Marini's repaints could probably be removed. To do so would be a finer tribute to Dante than all the speechmaking which is being planned for the approaching sixth centenary of his death. Giotto's Dante matches so closely the reconstruction of the skull, as is shown in the accompanying plate (Fig. 10), that we may suppose it to have been based on an early sketch from life. It is a most precious embodiment of the poet of the "Vita Nuova." But the portrait, though famous until Vasari's time, was singularly without influence. Only two or three Dante portraits before the discovery of 1840 show a trace of Giotto, and that doubtfully. We may attribute this strange case partly to the fact that the little chapel of the Podesta was not a place of public worship or resort, partly to the fact that people habitually thought of Dante not as the vision- ary young lover of the "Vita Nuova" but as the austere poet of the "Inferno." Something more grim than Giotto's Dante was expected, and about the same time as Giotto's a competing portrait of the desired grim type was painted by Taddeo Gaddi in one of the most popular churches of Florence, Santa Croce. As the most accessible and characteristic likeness, it naturally became the source of all subsequent grim Dantes. We have, I shall try to prove, a faithful fifteenth century copy of Taddeo Gaddi's Dante in a miniature prefixed to the Palatine Manuscript No. 320 of the Laurentian Library. (Frontispiece.) This miniature, weakly drawn with a pen and lightly washed with color, is inserted before the * "The original color was not white and red, and this is obvious from a close inspection of the bag and of the repainted red part. The scraper, in removing the whitewash, took out the color of a portion at the back of the head and of the pendent part, which may now be seen gashed by the razor; but, here and there, a red spot by chance remains even in the pendent portion, showing that the bonnet was red all over." Crowe and Cavalcaselle, "A History of Painting in Italy," (Hutton Ed.) Vol. I, p. 224. 12 THE PORTRAITS OF DANTE text of a collection of Dante's lyrics. The writing is not earlier than 1450. The volume belonged to Francesco Sassetti, treasurer for Lorenzo the Magnificent. Francesco fled from Florence in 1488, and as early as 1472 had a considerable library. (H. Hauvette, "Ghirlandaio," Paris, pp. 66, ff.) The manuscript is likely to have been purchased by the latter date. It cost ten broad gold florins. As for the Miniature, it is not an original drawing but a weak copy of something else. Those who have had experience of old Fig. 10. The Skull (dotted line) overlaid on Giotto's Dante. The fit is fair but the nose and eye come a little too low. drawings will see this without argument, and it is useless to labor the point with others. I may only say that the entirely purposeless character of the line, which cannot be matched in any original drawing of the fourteenth or fifteenth centuries, is conclusive. It is not a copy of a Renaissance original — for that it shows too little knowledge — ^but of a Gothic original. On general principles it is likely to be a copy of the most famous fourteenth century portrait of Dante in Florence, Taddeo Gaddi's. Elsewhere^ I have tried to show that such is the case, and have now only to develop the proof more fully. Quite independently, Parodi and Passerini have arrived 5 In the Romanic Review for 1912, page 119, reviewing Dr. Holbrook's book, I wrote of the Palatine Miniature : "The drawing is palpably a true copy of a fourteenth century original. It has distinct characteristics of Taddeo Gaddi's manner and should be a true sketch copy of that famous portrait which he painted in a Franciscan miracle on the choir-screen of the church of Santa Croce." The same year, before my review reached him, Parodi, in the Marsocco for July 28, in an article called "I Ritratti di Dante," expressed the same view more cautiously as THE PORTRAITS OF DANTE 13 i ^' '^ /^^:i-' \ \ — ^- • X Oh t/) •-3 U ■£ >> . .3 y w -s :S "« J^ *** 5 8 °i :« S ^ ■-3 '"i £ O « J3 O ^^ > TJ jn ^ 03 rt oj H^ « - . E -o il — ' OS rt a (u j: 4) J3 2 w .2 '^ . T3 ni "^ -a •5 2 J2 U rt V i:ih •a H H U-i 14 THE PORTRAITS OF DANTE at the same conclusion. As long as 1897 Franz Xaver Kraus (p. 177) was near the truth when he declared the Palatine Miniature represented a lost original which is the source of most of the grim Dantes. That the Palatine Miniature, or rather its original, is by Taddeo Gaddi, is more positively suggested by its peculiar artistic defects. It is a poor thing, but a poor thing of considerable character. Its chief fault is the absurdly short meas- ure of the head. Many painters of the first half of the fourteenth century Fig. 14. The Skull (dotted line) overlaid on the Palatine Dante. The fit is close except for the absurd proportions of the top and back of the head in the Palatine. shared this fault, but none save Taddeo Gaddi, so far as I know, exagger- ated it to the extent that we find in this copy. My plates (Figs. 11, 12, 13) which show alongside the Palatine Miniature two heads from Taddeo Gaddi's frescoes in the Baroncelli Chapel, in Santa Croce, make the case follows : "E lecito immaginare che la Miniatura Palatina abbia avuto, non diro per originale ma per punto di partenza il ritratto di Gaddi." ("It is permissible to imagine that the Palatine Miniature had, I will not say as its original but as its point of departure, the portrait by Gaddi.") Parodi maintained the same position in his elaborate and instructive review of Holbrook and myself in the Bulletino della Societa Dantesca Italiana N.S. Vol. XIX. (1912), pp. 88-106. Passerini in his recent brochure, "II Ritratto di Dante," pp. 11-13, takes the same position with more emphasis, apparently without knowledge of my review of Holbrook, which he does not cite. It is a pleasure to have the confirmation of these learned and distinguished Italian col- leagues. As for the identity of Gaddi's portrait and the Palatine miniature, the newly applied evidence of the skull brings something like objective demonstration to what in 191 2 was only an intuition in style and in historic probability. THE PORTRAITS OF DANTE 15 clear. We have to do then with a very faithful copy of the most promi- nent Dante portrait in Florence and with the fountain head of the grim tradition. Despite its artistic defects, it is the indispensable document. Overlaying the skull plan on it (Fig. 14), we find the fit is precise. Only the nose bone of the skull projects too much, and here surely the recon- struction, which rests on measurements, is more correct than the portrait. The correspondence is more exact than it is in Giotto's portrait, and here is the reason for denying the theory sometimes advanced that the Palatine Miniature is based on Giotto. For the matter of that it represents a more emaciated and an older man, not a man in the twenties, but in the late thirties or the forties. We must next ask what chance Taddeo Gaddi would have had of mak- ing or consulting a life portrait of Dante, for this portrait ultimately rests on an accurate study of Dante's features. Born about 1300, Taddeo Gaddi was Giotto's godson, and for twenty-four years his assistant, and finally his artistic executor. Presumably he served Giotto in the last twenty-four years of the master's activity, from 13 12, when Giotto after a long absence registered in the Painters' Guild at Florence, to 1337, when Giotto died. We know little of Giotto's movements during this period save that after 1328 he was for some time at Naples. But Vasari tells that Giotto worked at Verona, Padua, Ferrara, Ravenna, and elsewhere in the north, also spe- cifically that Giotto visited Dante at Ravenna.® Dante was at Ravenna from about 13 19 to his death in 1321. Before that, he was some years at Verona. It is possible enough that Giotto accompanied by his young as- sistant saw Dante between 13 12 and 1321 and that young Taddeo then drew his feeble but faithful sketch of the poet. The probabilities are that ^ Giotto, writes Vasari, after a long absence returned to Florence in 1316 (the real date being 1312). "But he was not allowed to stay long in Florence, because, brought to Padua by means of the Lords of the Scala family, he painted in Sant' Antonio, ... a very beautiful chapel. From there he went to Verona, where for Messer Can Grande he made some pictures in his palace . . . When these works were completed, in returning to Tuscany he had to stop at Ferrara, and he painted in the service of the Este Lords. . . . Mean- while, when it came to the ears of Dante, Florentine poet, that Giotto was at Ferrara, he managed that he should be brought to Ravenna, where he himself was in exile; and he had him make in the Church of San Francesco for the Polenta Lords some stories around the church which are excellent." Translated from Milanesi's edition. Vol. I, p. 388. Of course it is wholly the fashion to discard all unsubstantiated statements of Vasari, but this is as uncritical as was the earlier fashion of accepting him in toto. Vasari's statement begins with a fact — that Giotto painted in the Santo at Padua, why should we suppose that it ends with a shameless fabrication? In short it is likely enough that Giotto and Taddeo Gaddi met Dante in 1319 or a little earlier at Ravenna, and equally possible that the meeting was still earlier at Verona. The Palatine Miniature suggests a man rather under than over fifty, the age which Dante reached in 1315. i6 THE PORTRAITS OF DANTE the meeting would have been not very late within this span, for Dante, who lost his teeth in his old age, is not represented as toothless. Mere possibilities these! — the sceptical reader will remark. All that is important is that whoever drew the original of the Palatine Miniature had before him a head and face the measurements of which very closely correspond with those made on Dante's skull. Who made the original drawing is of small account. Taddeo Gaddi himself is the best guess, for a superior work would have kept something of its quality even under his slack transcription. Just when Taddeo painted the portrait is uncertain. From 1332 to 1338 he was working on the Life of the Virgin in the Baroncelli Chapel. Lor- FiG. 15. The Palatine Miniature (dotted line) overlaid on Giotto's Dante. The fit of the faces is close, though Giotto has some- what extended the nose at the expense of the other fea- tures. The comparison shows strikingly the false proportions of the back-head of the Palatine. enzo Ghiberti in his Commentaries tells us this about the Dante portrait: Taddeo "painted in the Church of Santa Croce about the middle of the church the miracle of the resuscitated boy. There is the figure of Dante Alighieri, where there are together three figures taken from life, and his is the middle one. They are standing." The lines cutting off the Palatine bust at the lower left suggest that it was taken from a standing figure, Vasari says in his first edition that the two other figures were Giotto and the poet Guido Cavalcanti; in his second, that some say the third figure was Taddeo himself. The subject was a natural sequel to Giotto's Mira- cles of St. Francis in the Bardi Chapel, which were done no later than 1328 and probably some years earlier. Taddeo's work in continuation is likely to have followed his master's closely. In that case his portrait of Dante will have been painted some ten years before Giotto's. Giotto's and Taddeo Gaddi's portraits differ greatly in expression but are very similar in proportions as may be shown by overlaying one on the other (Fig, 15). Giotto slightly shortens the upper lip and raises the in- dentation above the nose making that feature longer in span. The differ- THE PORTRAITS OF DANTE 17 ences, on the scale of life would be about a quarter of an inch in the upper and an eighth in the lower discrepancy — enough to discredit the view that the Palatine Miniature was copied from Giotto's Dante. If additional confirmation of the view that the Palatine is the sole author- ity for the looks of Dante in maturity be needed, it can be furnished by a novel experiment. When I reconstructed the skull, I intended to use it only as a check on the portraits. But when the outline lay clear before me, I could not resist the temptation to add features. In so doing I tried to work objectively, disregarding all the portraits. The result was appallingly like the Palatine Fig. 16. Fig. 17. Fig. 18. The Palatme Miniature (Fig. 18) compared with two Reconstructions traced upon the Skull by two artists who did not know the Skull was Dante's. Miniature. Naturally I attributed this to unconscious "wishful thinking," for I already believed the Palatine portrait to be the most accurate. So I again repeated the experiment with two of my colleagues, with the same result. But they knew they were working on the plan of Dante's skull, and, though both were ignorant of the Palatine portrait, might again have unconsciously echoed the very similar traditional type represented by the Naples Bust. Evidently the crucial experiment must be with men who did not realize that Dante was in any way in question. Accordingly I asked a painter and a sculptor to trace features upon the skull, merely saying that the face should be thin, the eye large and the lower lip projecting (Figs. 16, 17). The accessories, cap and ear-tabs, were added only after the face had been drawn. These reconstructions naturally differ, but both have a strong family likeness and both resemble the Palatine Miniature more than any other portrait. We seem then to have a nice cross verification both of the rightness of the reconstruction of the skull and of the authority of the Palatine portrait (Fig. 18). It has seemed unfair to these artist friends i8 THE PORTRAITS OF DANTE to make them responsible for these casual sketches, and unfair to the reader to tamper with the evidence ; so I take occasion to thank Messrs. Augustus Vincent Tack and Mahonri Young, for their kindness and leave to the reader the by no means uninteresting critical game of guessing who did which drawing. The development of the still older, sterner and more artistic portraits of Dante from Taddeo Gaddi's crude but faithful likeness is another story which I will soon try to tell. Most of it is accessible enough in Dr. Hol- brook's excellent book. It is enough to say that Gaddi's Dante held its place on or near the choir screen of Santa Croce until, in 1566, Vasari ruthlessly modernized the old Franciscan church. Two years before the demolition, at Michelangelo's funeral, among the decorations was a portrait of Giotto holding in his hands the head of Taddeo Gaddi's portrait of Dante (Vasari, Vol. VII, p. 307). The nameless artist was pardonably muddled as to the painter, but in the straight Florentine tradition as regards the choice of type. For over two centuries Florence had regarded Taddeo's likeness as the standard, and Florence was right. The up-shot of this chapter is twofold. Giotto's Dante in the Bargello must be regarded as an idealization based on a sketch from life; Taddeo Gaddi's as an uninspired but very faithful rendering of ^ro+trr's features in his maturity. The modern artist who wishes to do a Dante should consult the Palatine Miniature and the reconstruction of the skull. Finally we should be most grateful to that fourth-rate and nameless draughtsman of the fifteenth century who preserved for us the only portrait which tells us what Dante really looked like. XV^^ CHAPTER TWO OTHER FOURTEENTH CENTURY PORTRAITS CHAPTER TWO OTHER FOURTEENTH CENTURY PORTRAITS Orcagna's Toothless Dante. An Alleged Dante at Ra- venna AS Source of the Tomb Relief. Ugguccione Bambag- LiOLi's Sketch. Finding portraits of Dante in fourteenth century frescoes is so facile an indoor sport that the list is notable and increasing. Wherever an old face under the familiar tailed cap and ear-tabs appears, the intelligent tourist exclaims "How like Dante !" And many scholars who should have known better have yielded to this impulse. Thus we have alleged Dantes at Verona, Ravenna, Rimini, Siena, Florence and Assisi, to go no further. Now of course the tailed cap with a white under-cap showing ear-tabs was common wear from before 1300 to nearly 1400. Since the material may be found in Holbrook and Passerini it does not seem necessary to con- sider all these doubtful identifications. I shall limit myself to such as seem to me to have a chance to be right. Far the most impressive of these portraits is by Andrea Orcagna or Nardo his brother. The old and modern critics differ as to a point that can never be settled. In the Strozzi Chapel, in Santa Maria Novella, there is a Last Judgment on the window wall. Below the great Christ, at the spec- tator's left, a worn and ecstatic old face looks upward from a group of theologians and poets (Fig. 19). None of the early writers mentions this haggard head as a Dante portrait, but for nearly a hundred years modern scholars have been discovering and rediscovering it (Holbrook, pp. 161 f.). •We know from Vasari that Orcagna was "most studious of Dante," and the Inferno in the same chapel, which was unquestionably composed under Orcagna's direction, follows circle by circle and holgia by holgia the topog- raphy of the "Divine Comedy." Hence it is highly probable that we have to do with the portrait of our poet. The head has certain marked peculiari- ties, though generally it appears to follow the version of Taddeo Gaddi. It is much older than the man in the Palatine Miniature, the cap is peaked up in front — a feature which we shall later find in the Naples Bust — the ear-tabs are small and instead of hanging loose are tied under the chin. This trait prevails generally in the miniatured manuscripts of Dante and in most of the three-quarters portraits. A more remarkable peculiarity is that the mouth is toothless, or at least the upper jaw, so that the lower lip 22 THE PORTRAITS OF DANTE slips well behind and inside of the upper. When we remember that Dante was toothless in his last years, as the skull attests, there is a temptation to hope that a third genuine portrait underlies Orcagna's. Born about 1308, and never so far as we know in Northern Italy, it is virtually impossible that Orcagna should ever have seen the poet. What is likely is that he merely made Gaddi's portrait older, and intelligently added traits which he had from oral tradition. Since the portrait has great intensity of char- FiG, 19. Dante from Orcagna's fresco of the Last Judgment in the Strozzi Chapel, Sta. Maria Novella. acter and seems to have served as inspiration both for Raphael and his unknown contemporary who modelled the Naples Bust, the student can- not neglect it. It anticipates the conception of Dante as haggard and tragic which ruled in the Renaissance. And if Raphael and the maker of the Naples Bust took Orcagna's yearning type for a Dante, why should we be ashamed to imitate them? In the same chapel, in the Paradiso, under the great throne at the left, is a forceful young face with the familiar attributes (Fig. 20). While it is not far from the true proportions, it seems to me rather a generalized Gothic type than surely a portrait of any one. The upward cut of the base of the nose is entirely unlike that of any Dante in the true tradition. It resembles alleged Dantes at Siena and Rimini which may be consulted in Passerini and Holbrook. I reproduce this portrait merely to show how THE PORTRAITS OF DANTE 23 seductive these casual attributions are. The Strozzi chapel was decorated after 1354. Quite apart from the Dante portrait, its frescoed decorations with the Last Judgment, Heaven and Hell, make it the most Dantesque in- terior in the world. In the same decade Andrea Bonaiuti was covering the chapter house of the same church, later called the Spanish Chapel, with the most elaborate Dominican allegories. Here plausible Dantes so abound that the critics have been moderate indeed in fixing the name only upon the bent, writing figure which sits at the feet of the impersonation of Grammar. There is Fig. 20. Alleged Portrait of Dante from the fresco of the Paradise by Nardo di Gone in the Strozzi Chapel, Sta. Maria Novella. Fig. 21. Dante(?) from Andrea Bonaiuti's fresco of the Church Militant in the Spanish Chapel, Sta. Maria Novella. not much for the identification. The scribe resembles the true portraits in no way, and the traditional champion of Grammar should in any case be Priscian or Donatus. If there is a Dante in the Spanish Chapel, it is, I think, the hooded figure facing towards the Emperor, among his cham- pions, from the right, near an ermined justice (Fig. 21). As the author of the "De Monarchia" Dante would have deserved such a position, and by 1360 or so the pious Dominicans may have forgotten or ignored the fact that some thirty years earlier, in 1329, the treatise on the monarchy had been condemned and burned as an heretical book. Such possibilities are alluring, but, in the absence of the traditional headgear of Dante, should not be insisted on. If a Dante, it is one of the few that rest upon Giotto's portrait. In January of 1920 the friars of San Francesco at Ravenna removed the whitewash from two frescoes^ in which they recognized Dante's portrait. ^ The reader is referred to Michele Barbi's cautious and thorough discussion of these 24 THE PORTRAITS OF DANTE In one case he appears as a witness of the crucifixion and the other as a pensive seated figure with hand on chin apparently reading intently before a lectern (Fig. 22). The latter fresco is near the side door that opens into the little graveyard where Dante's tomb originally stood. Pietro Lom- FiG. 22. Fresco by Giovanni Baronzio of Rimini, before 1350, recently uncovered in the Church of S. Francesco, Ravenna. May have been intended to rep- resent Dante, but is not a portrait. Fig. 23. Marble Relief made in 1483 by Pietro Lom- bardi for Dante's Tomb at Ravenna. bardi who enclosed the sarcophagus with a shrine and provided a por- trait relief in 1483 clearly believed this fresco to be a Dante, for he adopted its pose for his sculpture (Fig. 23). Both the relief and the (fresco are characterless and far from the good tradition. The same painter, Giovanni Baronzio of Rimini, according to Mr. Berenson (per- portraits in "Studi Danteschi," Florence, 1920, Vol. I, pp. ii3ff. Here also is the bibli- ography. With Barbi's skeptical conclusions I entirely agree, but it seems to me he minimizes unduly the relation between the seated "Dante" and Pietro Lombardi's relief (pp. 121 f.). THE PORTRAITS OF DANTE 25 sonal communication), repeated the figure in Santa Maria in Porto outside of Ravenna. The same face and figure now represent Anti-Christ, which is perhaps a sufficient commentary upon their value as portraiture. (See Abb. I. in A. Brach, "Giotto's Schule in Romagna," Strassburg 1902.) We must deeply regret the loss of frescoed portraits of Dante and Petrarch by Lorenzo Monaco. Vasari places them in an Ardinghelli Chapel in the Trinita. Milanesi (Vol. II. p. 20, n. i), thinks they were rather in a chapel of the same family in the Carmine. Their date will have been not far from 1400. Probably, since character was not Lorenzo's forte, these were rather fine works of idealistic Gothic painting than fine portraits. With the exception of the illustrated manuscripts, which are of small >'/.V v-^ :ci Fig. 24. Dante (?). Fig. 25. Bologna Grassa. Sketches by Ugguccione Bambaglioli, Notary of Bologna, between two Documents of 1323. These possibly represent Dante receiving the Poet's Laurel from Bologna. importance and may best be separately treated, this closes our Gothic chap- ter. I may, before passing to the next century, mention a fascinating pos- sibility of a sketch done shortly after Dante's death. Among the notarial records of Ser Ugguccione Bambaglioli of Bologna, in the narrow space between two documents of 1323 (two years after Dante's death) is lightly sketched a curious scene of a Coronation. A very fat woman in the right hand margin offers a wreath to a figure kneel- ing in the left hand margin (Figs. 24, 25). The little kneeling figure is fairly Dantesque. The draughtsman was a fellow townsman, kinsman, and associate of Ser Grazioli Bambaglioli, one of the earliest commentators of Dante. In 13 19 Giovanni del Virgilio had invited Dante to come and re- ceive the laurel at Bologna (see note i). The fat figure is probably the traditional Bologna grassa crowning some one. This circumstantial chain is developed with charm and modesty by Giovanni Livi in the Nuova 26 THE PORTRAITS OF DANTE Antologia for April i, 1904. He imagines Ser Ugguccione in an idle mo- ment thinking of a laureation that ought to have been and scribbling in the scene between documents. I reproduce the tiny and spirited sketches enlarged to two diameters, not wishing to omit what may be, however negligible as portraiture, the earliest extant representation of the poet. I may add that the abbreviation Ug^ above the kneeling figure is probably neither Ugguccione's signature nor yet a label implying that he himself ought to be crowned. It is, I believe, simply the catch-word for the fol- lowing document, which begins "Ugolinus." CHAPTER THREE THE FIFTEENTH CENTURY, THE ICONOGRAPHIC TRADI- TION AND ITS GROWTH CHAPTER THREE THE FIFTEENTH CENTURY, THE ICONOGRAPHIC TRADI- TION AND ITS GROWTH The Three Iconographical Traditions. Traces of Giot- to's Dante. The Gaddi Portrait. Palatine Miniature as Source of the Grim Dantes, Modified in the Riccardian. Faithfully Transmitted in the Naples Bust. The Riccar- dian Portrait Possibly a Replica of the Lost Portrait of 1429. Source of the Medal and Death-Mask. Whoever has seen a great collection of portraits of a historic person, say a Washington or Franklin, will have been amazed at the variety of ex- pression and physiognomy. On the contrary, the accessories of costume and equipment are singularly uniform. Thus in tracing the genealogy of portraits buttons are often more important than eyes. The study of such fixed traits is called iconography. I wish to trace the early iconographical tradition of the three oldest portraits of Dante — Giotto's, Taddeo Gaddi's and Orcagna's. The distinguishing traits may best be presented in tabu- lar form. Giotto in the Bargello Kirkiip's Copy i) An ample hood. 2) A white under-cap shows obtuse rounded tabs over the ears, nq ties. 3) Cassock has lapels. 4) Hair is hidden. 5) Upper lip is short. 6) Nose is aquiline. Taddeo Gaddi in Santa Croce ( Palatine Miniature ) i) A cap with a pointed tail hanging behind. 2) White cap has acute angled tabs from which ties hang loose. 3) Lapels (from Giotto). 4) Hair shows at three points : at forehead, before ears, and in tuft at nape of neck. 5) Upper lip is long. 6) Nose breaks angle. 29 at an Orcagna in Santa Maria Novella i) A cap cocked out in front, the baggy tail falling somewhat fore- ward. 2) Acute angled tabs tied under chin. 3) No lapels. 4) Hair shows at two points before and be- hind ear tabs. 5) Upper lip is short, mouth toothless. 6) Nose aquiline. 30 THE PORTRAITS OF DANTE Of these traits numbers one to four are most important for classifying derivatives, particularly number two, the form of the ear-tabs and dispo- sition of the strings. The tradition of Giotto is singularly sterile. In his own century I detect no influence of his delightful portrait unless it be in a figure by Andrea Bonaiuti in the Spanish Chapel (Fig. 21). This head is similar to Giotto's Dante both in form and in the ideality of its senti- ment. But it may not be a Dante at all. In the fifteenth century I find Fig. 26. Dante by or after Vasari, date after 1544, in the L. E. Holden Collection, Museum of Fine Arts, Cleveland, O. the trace of Giotto only in a rather nondescript miniature portrait adorn- ing the great initial N of a vellum copy of Landino's edition of the "Di- vine Comedy," dated 1481. The volume is in the Magliabecchiana at Flor- ence, and the portrait has been well reproduced in color as the frontispiece to Dr. Holbrook's volume. Only the youthfulness of the face and the short rounded forms of the ear-tabs associate it with Giotto's Dante. Were it detached from its text, no one would suspect that Dante was the person represented. Probably Giorgio Vasari had Giotto's version in mind when in 1544 he painted Dante with his fellow poets, Guido Cavalcanti, Cino of Pistoia, Guittone d'Arezzo, Petrarch and Boccaccio. Vasari tells us that the heads were "taken accurately from the early copies," and that many copies of THE PORTRAITS OF DANTE 31 his group were made. In the latter statement, at least, he seems to have been correct, for a copy of this group was in the Orleans Gallery, and a somewhat larger one, sometime in the Imperial Austrian collections, now hangs on the chimney piece of the Senior Common Room at Oriel Col- lege, Oxford (Holbrook, p. 157 f.)- Through the courtesy of the Cleveland Art Museum I am able to offer an unpublished variant of Vasari's Dante (Fig. 26). It has been somewhat rearranged and the ear-tabs changed to a longer form like those in Raphael's Dantes. The pose of Dante in the original group may, as Dr. Holbrook suggests, have been borrowed frora Signorelli's Dante at Orvieto. But the relative youthfulness of the face, Fig. 27. Drawing formerly ascribed to Gaddo Gaddi in the Library of Christ Church, Oxford. Date about 1400, identification as a Dante uncertain. By permission of the Corporation its soft expression, aquiline nose, and form of ear-tabs (in the Oriel ver- sion) betray Giotto as the chief inspiration. The Cleveland Dante, which is not much later than 1550, is a careful example of the bad, sleek execution of the time, and if not by Vasari himself, is at least a contemporary ver- sion of his type. From now on for nearly three hundred years Giotto's Dante is completely forgotten, and even since its discovery in 1840 it has, outside the inner circle of Dante lovers, been singularly neglected. The book publishers have continued to prefer the inferior types drawn from sixteenth century origi- nals or from Marini's restoration. Virtually all the Dante portraits of any character descend directly from Taddeo Gaddi's. The endeavor of the fifteenth century was either to make three-quarter aspects out of the original profile, in which process much character was lost, or to create an older and more austere type of profile. Perhaps the most remarkable and earliest of these derivatives is a sadly rubbed and retouched drawing in the famous Christ Church collection (No. V. i) at Oxford. Through the kind permission of the Corporation 32 THE PORTRAITS OF DANTE I am here able to publish it for the first time (Fig. 27). In general it seems to follow the lines of the Palatine Miniature, but the brow is bulged and the forehead sloped back. One might suspect that some old version different from Taddeo Gaddi's underlay this grim portrait. Against that is the good and relatively modern indication of the eye. No real Gothic painter did ihe thing so well. Iconographically it is unique. Instead of the hood or Phrygian cap, we have the loose turban common from 1400. The ear-tabs oddly combine all the early types. We have the rounded form of Giotto covered with the sharp points of Taddeo Gaddi, which are fastened under the chin as in Orcagna. It is this which makes me pretty sure that we have to do after all with a composite Dante and not with a portrait of some unknown fifteenth century Florentine of Dantesque look. The drawing belonged to Carlo Ridolfi, the historian of Venetian paint- ing, who died in 1648, and bears a worthless eighteenth century ascription to Gaddo Gaddi, Taddeo's father. On its face, the work is that of a Flor- entine trained in the Gothic technique but commanding something of the new knowledge of the Renaissance. As work it could hardly be earlier than 1 410 or later than 1440. In the whole matter we are frankly in a field of uncertainty, and there is no reason apart from its strongly Dan- tesque look and the slight circumstantial evidence given above for calling this portrait a Dante. We rejoin the true iconographic tradition with the frontispiece of the Riccardian Manuscript 1040 (Fig. 28), and here we also find the first Renaissance Dante. The manuscript is a small in-folio written on paper and containing a selection of Dante's lyrics with others (see Holbrook and Parodi passim). The portrait is boldly executed in colors on the front of a vellum guard leaf (295 by 200 mm.), hence may be of different date from the text. We should not speak of it as a miniature, for it is only slightly under the scale of life. It is, moreover, the work not of a miniaturist but of a man accustomed to work in fresco. Dr. Holbrook's facsimile in colors makes any elaborate description superfluous. It agrees in all six icono- graphic points with the Palatine Miniature, but is a modernization of that type. The attempt is to introduce a reasonable truth of anatomical de- tail, and to make the expression somewhat older and more grim. These improvements have led to certain infidelities to the exemplar, Taddeo Gad- di's fresco. The cap is cocked forward after Orcagna's fashion, the brow is made more bulging through the indentation of the nose, the upper lip is considerably shortened as is also the chin. These changes make the nose relatively larger. They are perhaps due to Giotto's portrait which offers these peculiarities in a less marked fashion (see Fig. 5). Besides, the lower lip projects slightly (a trait probably drawn from Boccaccio's report), the tuft of hair at the nape of the neck is covered by a bag, the cassock shows THE PORTRAITS OF DANTE 33 Fig. 28. The Early Renaissance Dante. Painting- inserted in front of Riccardian Ms. No. 1040, Florence. By Paolo Uccello or an imitator and possibly a variant of the lost Dante done in 1429 for the Cathedral at Florence. 34 THE PORTRAITS OF DANTE no standing collar, is more exposed than in Taddeo Gaddi's portrait and is completely closed in front by numerous buttons, finally the ties of the under cap are rigid and perfectly straight. In nearly all these peculiarities the Riccardian Portrait forms a group with a nearly contemporary bronze medal and the so-called Death-mask. Unfortunately we have no evidence as to the age and authorship of the Riccardian Dante. The manuscript is later than 1450. The frontispiece might be an earlier thing bound in, or it might be an addition by a later ^^ From Heiss. Fig. 29. Bronze Medal, Florentine, about 1465. owner. On the basis of style alone, I am confident that such a portrait could have been made no earlier than 1425 and is not likely to have been made later than 1470. More narrowly, it is the work of an artist trained before 1450, for it reveals the rigidity and painful detail of the first gen- eration of Florentine realists. Too inert in its contours for Andrea del Castagno, and too hard in its construction in light and shade for Domenico Veneziano, it is in every way like the work of Paolo Uccello and should be by him or by a close imitator. Paolo died in 1475. There are reasons to think that this portrait was well known, for its es- sential iconographic peculiarities are echoed in a bronze Medal of the period (Fig. 29), while the so-called Death-mask of Dante shows identical facial proportions, indeed the two profiles match precisely (Fig. 30). This identity has escaped previous observers because of the forward tilt of the Death-mask. The Naples Bust, which is generally associated with the Riccardi Portrait, has much truer projxDrtions and the general similarities are to be laid to a common origin in Taddeo Gaddi's portrait of Dante. THE PORTRAITS OF DANTE 35 Since the Riccardian Portrait, the Death-mask and the Medal are not likely to be by one hand, and indeed all seem different in period, we must seek an explanation for the relationship. The easiest hypothesis is that the Riccardian Portrait represents a well known version displayed in a public place and of contemporary authority. The portrait is precisely what a young and progressive artist might have executed in 1429 when Fra An- tonio Neri, public Dante reader, commissioned a Dante for the Cathedral (Holbrook 172 ff.). We may note that Paolo Uccello did the equestrian Fig. 30. The Riccardian Portrait compared with the Death-mask. effigy of Sir John Hawkwood for the Cathedral in 1437. It is possible enough that eight years earlier he did the Dante for Fra Antonio. As the only Dante in modern style accessible in Florence, it might readily have gained a temporary authority during its thirty-six years of public exhibi- tion.^ When, in 1465, it was superseded by Michelino's Dante it was still preserved for three hundred years in the Cathedral precincts and was available for any active searcher. Naturally I do not wish to give to a pure hypothesis more than its due weight. At least it may serve as a pro- * Since Fra Antonio's picture represented Dante and an old man engaged in a colloquy in a Florentine street, it is natural to suppose that they faced each other presenting their profiles to the observer. Indeed anything but a profile portrait was unlikely as early as 1429. The scene, a city street, suggests again that the painter was a progressive of Uccello's type, and interested in the then new art of linear perspective. For date see Catalogue, No. 8". 36 THE PORTRAITS OF DANTE visional explanation, pending a better one, of a curious chain of facts and circumstances. The fifteenth century Medal of Dante exists in several sizes. I repro- duce the one chosen by Armand Heiss in his "Medailleurs Italiens," t. VHP pi. XVII. No expert has attempted either to date it precisely or to name its author. Somewhat classicized through the addition of a wreath and the softening of the features, it agrees fairly both in facial proportions and iconographical traits with the Riccardian Portrait. The reverse shows Dante standing in profile facing the mountain of Purgatory and the en- trance to the Inferno. In a general way this composition is similar to Michelino's, but the motive is different. Michelino makes Dante turn away from the scene of his poem and face the public. I think both the obverse and reverse of this medal may be drawn in part from the lost picture of 1429, which in turn may well have furnished precedents to Michelino. The date of the medal is rather likely to be 1465, the second centenary of Dante's birth. As for the Torrigiani Death-mask in the Uffizi,^ the parent of nu- merous copies, we have no history of it before the eighteenth century. Its false proportions are to a hair those of the Riccardian Portrait. On the theory that it is a falsification, nothing is more likely than that the fabri- cator should have hunted up the discarded and virtually unknown Dante of 1429 in the storerooms of the Cathedral. Such is probably the case. But it also remains possible that this version was never even intended as a death-mask, being merely so labelled by some credulous owner. In that case we should have to do with a plaster sketch for a memorial medallion which was presumably never executed in more enduring materials. To date the work would be hazardous. Its generally austere and unmannered modelling suggests a sculptor still influenced by Donatello and active about 1500. It may of course be merely an imitation of such work. The re- semblance to the probably contemporary Naples Bust, which has universally been overstressed, seems to be more or less casual and due to the persistence in both works of the traits of Taddeo Gaddi's portrait. The proportions of the bust are far more faithful. That the Riccardian Portrait represents a well known original is once more suggested by the kindred fifteenth century panel in the collection of Prince Trivulzio at Milan. It is reproduced in Count Passerini's excellent ^ Perhaps the most discreditable chapter in Dante scholarship is the credulous ac- ceptance of the Death-mask by such scholars as Cavalcaselle, Norton, Paur, Maria Ros- setti and others. Surely there should have been caution in considering an object without early history or credentials. The mere fact, accessible to all, that Dante's skull had been found toothless should have suggested that the Torrigiani Mask was not based on a mould taken from the dead face. THE PORTRAITS OF DANTE 37 little brochure as plate 9. Its peculiarities are that the face is slightly turned towards the spectator while the long tabs are not straight but curved. One may suspect a similar origin for the rather nondescript fifteenth century miniature in a Codex Eugeniano of the Palatine Library at Vienna, Pas- serini, plate 7. It shows the same misproportion of the features, the row of l\. ,,yj)v • l< ' « . '■ ''■ '^ * ^ ( / ' '' .. ' ' * " . ';' '':^I 1 i 1 1 " -^ 1 ■ • i V ( j •; 1 i Fig. 31°. Dante and Beatrice, Para- dise XXVIII. Note the difference between this Dante and the one in the adjoining cut. From Holbroak Fig. 3I^ Paradiso III. Here there seems to be an influence o£ the Medal (Fig. 29) in the face and shortened form of the cap-tail. buttons and the sharp tabs. Conformably to Boccaccio's account, a trace of beard is added. As for the Riccardian Portrait, it seems not a copy but an original work. If it is a version of the lost portrait of 1429, as seems to me very likely, we must regard it as a study or a replica. It by no means deserves the confidence that Luigi Passerini and Gaetano Milanesi (on the strength of its similarity to the Death-mask), bestowed upon it, for it gives a quite false account of Dante's features. But it is interesting as 38 THE PORTRAITS OF DANTE the earliest and perhaps artistically the most characterful of the early Renaissance Dantes. Botticelli, in the Dante drawings made after 1481 for Lorenzo di Pier Francesco de'Medici, continues the iconographic tradition rather fitfully. In his examination of these vellum sheets now divided between the Print Room at Berlin and the Vatican, Dr. Holbrook (page 189) found only ten figures which have a Dantesque character. Botticelli's acquaint- \ Alinari. Fig. 32. The Bronze Bust in the Naples Museum. ance with Taddeo Gaddi's portrait is shown by his tendency to let the cap-ties hang loose and crinkly. Naturally in these swift and imaginative sketches he depended on memory. I do not think that the influence of any portrait except Taddeo's, and possibly the Medal, is traceable in this won- derful series of illustrations (Figs. 3i\ 31"). Botticelli's habit of varying his Dante type capriciously, however, had a precedent in many early illu- strated manuscripts of the "Divina Commedia." The engravings in the famous Landino edition, Florence, 1481, depend on Botticelli's drawings and naturally repeat his type of Dante. THE PORTRAITS OF DANTE 39 The true iconographic tradition reaches its culmination and end in the tragic bust at Naples (Figs. 32, 33). This great masterpiece in bronze may well have been modelled and cast in the early years of the sixteenth century (see p. 64), but it carries so strongly the stern accent of Dona- tello's age that I must include it here. As to its history we have no ade- quate record. It came into the Naples Museum in 1738 with the famous collections of Ottaviano Farnese. In 1545 he had married Margaret of Austria, the widow of the infamous Alessandro de' Medici. Guglielmo Becchi, a learned editor of the "Museo Borbonico," Naples 1852, Vol. XIV, tav. XLIX, makes the suggestion that the bust may have come from the old collections of the Medici through this marriage. Ultimately we are reduced to stylistic evidence. Corrado Ricci (pp. 279 flf.) has sug- gested that this Bust, with the so-called Death-mask, represents a lost origi- nal by Tullio Lombardi which once stood on Dante's tomb at Ravenna. It was probably brought to Florence about 1480. A Ravennese Archbishop is said to have given the precious relic to Giovanni da Bologna, who in turn left it with his pupil Pietro Tacca, from whom it was snatched by an eager Sforza countess, in the early years of the seventeenth century. This strange story is supported by certain circumstantial evidence which may better be discussed in an appendix (II) than here. It has the merit of bring- ing the Bust to Florence some time before the Death-mask turns up, and also of accounting for the absence of all influence of so noble a work in Renais- sance Florence. It is supported also by general if not striking similarities between the Naples Bust and the tomb effigies of TuUio Lombardi, also by the general probability that he might have added a portrait to the sepulchre designed by his father at Ravenna, Such considerations have brought ad- herents to Ricci's view. Against it there are, it seems to me, especially strong circumstantial argu- ments and the stronger evidence of the Bust itself. If so splendid a thing stood for three-quarters of a century on the much visited tomb of Dante, surely it would have left some trace of itself at Ravenna and throughout the north. On the contrary, wherever we find an echo from Ravenna, it is of Pietro Lombardi's feeble marble relief. It would be strange too that there should be two sculptural portraits on one tomb, and Dante's tomb had its portrait slab from 1483. More difficult yet is it to reconcile the robust and searching modelling of the Naples Bust with the sentimental and rather timid handling of Tullio Lombardi. The bronze clearly speaks of Padua or Florence rather than of Venice as its birthplace, and its artistic back- ground is unmistakably the stern naturalism of Donatello. Its iconographical peculiarities point to Florence. It transmits with singular fidelity the main proportions of Taddeo Gaddi's portrait (Fig. 34), with which it agrees in all essentials. But it is a composite work and 40 THE PORTRAITS OF DANTE also reflects the influence of the Riccardian Portrait. This is shown in the sloping of the forehead, the indenting of the nose and projecting brow; in the protruding underlip, in the disposition of the cap and in the rigid ties. Again the very large eye of Taddeo Gaddi has been reduced to more agree- able if less correct size. Generally speaking the proportions are from Taddeo Gaddi and the accessories from the Riccardian type. It combines in an audacious and impressive synthesis all that was best in the Gothic and the Renaissance Dante. As an imaginative conception of the poet it is in- Alinari Fig. 33. The bronze Bust at Naples, front view. comparable, yet I doubt if it is quite a true portraiture. When Dante looked as old and worldworn as this he was toothless. If the Naples Bust were made a little younger, more like the Palatine Miniature, by a sculptor of genius we should have a true semblance of the haughty exile who "made a party by himself." The date and origin of this Bust remain doubtful. I fully agree with Count Passerini — and the opinion has the weighty support of my colleague Professor Allan Marquand — that the bust is Florentine. It is made on models which were readily accessible only at Florence, it betrays the in- fluence of Donatello. In its general handling it suggests Donatello's pupil and successor Bertoldo, but there is nothing specific enough to link it to his name. As to date, the probable limits would be from say 1475 to 15 15, with a probability for the later years of the span. If I am right in thinking that the wood-cut portrait in the "Convivio," 1521, reflects the influence of the Bust (see p. 64), then the Bust may be set a few years earlier, say 15 ^5" 1520. In any case we must suppose that this masterpiece was jeal- THE PORTRAITS OF DANTE 41 ously guarded in some private collection, for it is almost without influence on the contemporary iconography of Dante. The main conclusions of this chapter are as follows : The Riccardian Portrait may represent the lost Dante made for the Cathedral in 1429; the Fig. 34. The Palatine Miniature (dotted line) overlaid on the Na- ples Bust. The facial proportions are nearly identical, but the sculptor has exaggerated the nose, made the brow and lower Up project and the forehead recede, and has also diminished the size of the eye. taste of the Renaissance objected to the big eye and long upper lip of the Gaddi Dante, and attenuated these features ; the learning of the Renaissance took over from Boccaccio the tradition of the projecting under lip — a trait possibly marked only in Dante's last years (p. 2) ; while most por- trayers of Dante followed these innovations, the unknown genius who modelled the Naples Bust fortunately respected the true proportions of Taddeo Gaddi 's portrait. CHAPTER FOUR FIFTEENTH CENTURY, THREE-QUARTER FACE TYPES AND THEIR DERIVATIVES CHAPTER FOUR FIFTEENTH CENTURY, THREE-QUARTER FACE TYPES AND THEIR DERIVATIVES Miniature and Cassone Types. Castagno. Michelino, the Lost Bronzing. Morghen's Print. When Italian portraiture shifts from a Gothic to a Renaissance basis, from about 1460 on, the aspect changes from profile to three-quarter view and there is a corresponding improvement in structure and anatomy. The more progressive depictors of Dante followed this tendency, generally with unhappy results as regards character. There was early precedent in the Fig. 35*- Dante from a Laurentian Ms. of about 1400. 45 46 THE PORTRAITS OF DANTE manuscripts. A Laurentian manuscript, Plut. 40, No. 7, shows a rudely drawn Dante standing in three-quarters aspect and awaiting Charon's barque (Fig. 35"). The date is before 1400, and the characterization is fairly Dantesque in a caricature fashion. The earliest life-size Dante in this position is probably that which An- drea del Castagno did for a Carducci villa at Legnaia (Fig. 35"). It formed Fig. 35^ Fresco by Andrea del Cas- Fi<- 39'- Contempo- tagno, Museum of Sta. Apol- rary copy of a Ionia, Florence. 1440-1457. Drawing by Antonio Pollaiuolo, Library of Christ Church, Oxford. About 1475. part of a series of famous men and women, and with its fellows has been transferred to the Ex-Convent of Sant' Apollonia at Florence. The figure seems to follow Taddeo Gaddi's type rather closely in facial proportions (note the long upper lip), but the cap strings are tied after Orcagna's pre- cedent and the cap has a peculiarity in a furred reverse band. The pose and expression are serious, not to say stiff, and as characterization it is hardly adequate. Andrea has been more concerned with placing, balance, and modelling than with character. We have no means of precisely dating the work within the span of Andrea's activity, 1420 to 1457. Generally the three-quarter types partake of the mildness of this early example. We have a very gentle and idealistic portrait in a cassone front at New Haven (Fig. 36).^° The poet accompanied by Petrarch and Boc- 1^ In my review of Holbrook I heedlessly suggested that this figure might derive from the lost Dante of 1429, in the Cathedral. Consult note 8 for my considerate view of the matter. The Dante on this cassone front is as original as it is charming and uncharacter- istic. THE PORTRAITS OF DANTE 47 caccio stands by a fountain in a garden of love. The panel has been vari- ously attributed, but seems of Mas/accian type and no later than 1440 (W. Rankin, Burlington Magazine XI ('07) p. 339). In the soft and youthful beauty of the poet we may have a remote echo of Giotto's Dante in the Bargello. Far the most famous portrait of this type was painted by Domenico di Michelino in 1465 (Fig. 37). It replaced the Dante given in 1429 by Fig. 36. Dante (right hand figure} with Petrarch and Boccaccio. From a Florentine cassone front of about 1440 representing a Garden of Love. Jarves Col- lection, Yale University Antonio Neri. The contract, dated January 30, 1465 (new style 1466) provided that the design should be by Alesso Baldovinetti and the price 100 pounds. Domenico was a pupil of Fra Angelico and in his work imi- tated the bright Gothic coloring and idealistic sentiment of his master. The type and accessories are similar to Castagno's. The rose and green colors of the costume are from Giotto. The decorative beauty of this picture and its conspicuous position on the north wall of the cathedral gave it great authority and popularity. Soon the obscure name of Domenico was for- gotten and the great name of Orcagna substituted. No portrait of Dante has been so often seen. It is generally mentioned in old guide-books which > 48 THE PORTRAITS OF DANTE overlook Giotto. Altogether, its mildness satisfied that multitude of Italians whose knowledge of Dante, then as now, was confined to the touching episode of Paolo and Francesca. The head (Fig. 38) was revised before the middle of the sixteenth cen- tury by a painter of talent, possibly Agnolo Bronzino." Old copies of this version are in the Uffizi and the Yale Art School (Fig. 40). (See Dr. Siren's Catalogue No. 85.) After the middle of the eighteenth century a Fig. 2>7- Dante painted in 1465 by Domenico di Michelnio for ihe Cathedral, Florence. picture of this type was copied in line by Tofanelli and engraved from his drawing by the illustrious master Raphael Morghen, notably in the beauti- ful Zatta imprint, Venice 1757 (Fig. 39). His print was separately sold and often reduced to serve as frontispiece of many editions of Dante; thus this gentle and, to tell the truth, somewhat oldwomanly likeness of Dante became in sole competition with Bernardino India's portrait, the 11 Vasari (ed. cit. Vol. VII, p. 595) writes of Bronzino "for Bartolommeo Bettoni, to fill certain lunettes in one of the rooms, the portrait of Dante, Petrarch, and Boccaccio, half-length figures and very beautiful." THE PORTRAITS OF DANTE 49 standard effigy for the late eighteenth and most of the nineteenth century. In iconography too often the less fit survives. Artistically the best of these three-quarter types is a drawing at Christ Church (Fig. 39"). It represents the poet tall and gaunt holding a book and is more or less reminiscent of Castagno and Michelino. Mr. Berenson in his ''Drawings by the Florentine Painters," catalogues this Dante as a contemporary copy of a lost original by Antonio PoUaiuolo. I doubt Fig. 38. Head of Dome- nico di Michelino's Dante. Fig. 40. The Yale Dante (cut down at sides). Fig. 39. The TofanelH- Morghen Print (cut down at sides). if it was ever executed on a large scale, for in such case so distinctive a thing would probably have left some trace in the iconographic tradition. There is in the print room at Berlin a vigorous drawing of a middle aged man ascribed to Signorelli (Fig. 41"). It is near enough the Michelino type for some sixteenth century owner to have scribbled "Dante" on the sheet. He may well have been wrong in the identification. Mr. Beren- son in "Drawings by the Florentine Painters," has proposed the correct attribution, Piero di Cosimo. The date is not very far from 1490. I may add here the rugged figure in a manuscript of the University Li- brary at Turin (Fig. 41"). It is of Ferrarese (Cossa) type and about 1475. 50 THE PORTRAITS OF DANTE It merely shows how freely the painters worked when uncontrolled by the Florentine tradition. As portraiture it is fantastic. So ends a brief and inglorious chapter which has little to do with the authentic development of Dante iconography. •r^ Fig. 4i\ Dante from Ms. N, VI, ll University Library, Turin, Lom- bard School, about 1475. \ ■>.. ? k.. . .k /-* P Fig. 41". Dante (?). Drawing by Piero di Cosimo, ascribed to Sig- norelli, in the Berlin Print- room. CHAPTER FIVE IRREGULAR TYPES CHAPTER FIVE IRREGULAR TYPES In Miniatures and Cassone Paintings. Toothless Type. Giovanni dal Ponte and Signorelli. Justus of Ghent. FiLiPPiNO Lippi. Raphael. Wood-cuts of 1521 and 1529. Bernardino India, The Munich Drawing. Sessa's Wood- cut OF 1564. Outside of the true iconographic tradition which stems from Taddeo Gaddi there are a number of Dante portraits of a greatly modified or purely- fanciful sort. One such is the recently discovered fresco at Ravenna (Fig. 22). Since artists of the note of Signorelli and Raphael produce or fol- O laJwUtu: uiadliannaxctucomcdu TDromcc .Uc«urn udlacjiulc n.«a> xkI \»xr ifxtozio cTxUmiiu. I Fig. 42. Dante before the Muse Cal- liope, Purgatorio. I. 9., Miniature from a manuscript of about 1350 in the Marcian Library, Venice. From Bassennanu low eccentric types, I must treat in brief fashion a matter of rather minor interest to the true Dante lover. The manuscripts at all times, and the cassone paintings generally, offer irregular and most various portraiture of our poet. These portraits are of some importance if only because they represent a constant possibility of variation. Some eighty manuscripts out of eight hundred of the "Divina Commedia" contain either portraits or illustrations or both. The matter has been studied summarily by Volkmann, pretty thoroughly by Franz Xaver Kraus, reduced to instructive resume by Holbrook, and sumptuously .53 54 THE PORTRAITS OF DANTE illustrated by Bassermann. But the results from these studies are small. Generally the miniaturists worked at best from vague memory of the estab- lished types, often they took no pains to be consistent with themselves. Most of them adopted the pointed ear-tabs of Taddeo Gaddi as sufficient Fig. 43. Miniature for Inferno, i, from Ms. Ital. No. 74 Bibliotheque Nationale, Paris, date before 1400. Note two types of Dante in one picture. mark of identification, sometimes making assurance doubly sure by labelling the poet with a big D. Outside of Florence pure fancy reigned. We find in the Marcian Codex, cl. IX., p. 270, at Venice a bearded Dante^^ kneeling 12 Florence produced at least one bearded Dante in the Miniatures of Laur. Cod. Tempi I.C.Z. The chin beard is a mere fuzz as in the Palatine Miniature at Vienna [see Pas- serini pi. 7, and Ancona, "La Miniatura Fiorentina" (Florence 1914) Tav. LI.] THE PORTRAITS OF DANTE 55 before the muse (Fig. 42). The date will be about 1350. Elsewhere in the manuscript the type differs wildly. Evidently there was no standard outside of Florence, and not much for miniature painters at Florence itself. We have for example in a single miniature of Italian Ms. No. 74 of the Bibliotheque Nationale, a Florentine work of about 1400, the meeting of Dante and Virgil, and Dante fleeing from the three beasts (Fig. 43). The smaller figure shows a very aged and heavily lined face with the accessories of Taddeo Gaddi's portrait. The larger figure is younger, the nose is straight, and the lips are drawn in as if upon toothless jaws. Oddly enough Fig. 45. Dante and Virgil with Pa- olo and Francesca, Cod. Urbin, No. 369. Vatican. About 1480. From Bassermann. this single miniature contains the two influential manuscript types — One a very decrepit and aged face which recurs in Pesellino's Triumph of Fame at Fenway Court, and the other a young, straight nosed and toothless type, which occurs in a painting by Giovanni dal Ponte at Harvard University, in the famous Urbinate Ms. No. 369, in the Vatican, and in Signorelli's frescoed portrait at Orvieto. Raphael in the Stanze blends the two un- orthodox traditions in an aged and toothless Dante with a straight nose. We may briefly trace these two traditions. The little panel by Giovanni dal Ponte which I have just mentioned was painted probably about 1440 (Fig. 44). It represents Dante being crowned by a genius, with a second figure which may be either Virgil, as seems most likely to me, or Petrarch. The picture has been pub- lished with a learned commentary by Mr. F. Mason Perkins in Art in America for June 192 1, and also by Giacomo di Nicola in an appendix to 56 THE PORTRAITS OF DANTE Passerinl's booklet. It is the end of a cassone, the front of which prob- ably showed an array of famous poets. In a cassone front by the same painter (Spiridon Collection, Paris), a figure nearly identical with this Fig. 46. Signorelli's Dante, fresco painted after 1500 for the Chapel of S. Brixio, in the Cathedral, Orvieto. Fig. 44. Dante by Giovanni dal Ponte, from the painted end of a cassone, date about 1440, in the Fogg Museum, Harvard. Dante appears as the representative of Grammar. Probably the type served the artist indifferently for a Dante and a Priscian. The type shows a huge, nearly straight nose, a short lower face with projecting upper jaw and ap- pearance of toothlessness, ear-tabs tied under the chin, a bag rather than a tail falling behind from the cap, the reverse of the cap not in self stufif THE PORTRAITS OF DANTE 57 but in fur. Most of these traits recur in the Vatican Urbino manu- script (Fig. 45) and in Signorelli (Fig. 46). The idea of the projecting upper lip and toothlessness may have ultimately been drawn from Orcag- na (Fig. 19), but the type is so unlike his that one would rather credit or discredit the miniaturists and cassone painters with this not very happy in- vention. Fig. 49. From a Jardiniere in the Uffizi, historiated with the Tri- umphs of Petrarch. This head of Dante is in the Triumph of Fame. The date is after 1450. Fig. 47. From the Triumph of Fame in a cassone front by Francesco PeselHno, about 1450, in the col- lection of Mrs. John L. Gard- ner, Boston. The grim and aged Dante which occasionally appears in the manuscripts and in Pesellino's Triumph of Fame, about 1450 (Fig. 47), is based chieliy on the Riccardian type with perhaps an infusion of Orcagna. It is prob- ably the earliest portrait to emphasize the projecting under lip. It has pe- culiar narrow ear-tabs which show the ears. The most impressive version before Raphael is an intarsia (Fig. 48) of about 1500 by Benedetto da Ma- jano in the Palazzo Vecchio on the door of the audience hall. The long, round ear-tabs are novel traits which need explanation. Another ex- ceptional type is on a painted jardiniere in the Uffizi in a Triumph of Fame (Fig. 49), of about 1450 (Schubring "Cassoni" taf. LVI). The facial type with straight nose and short upper lip foreshadows Raphael's. The peculiarity of long round ear-tabs appears earliest in a stern figure in the fresco of St. Peter before the Proconsul in the Brancacci Chapel (Fig. 50). The painter is Filippino Lippi and the date about 1484. No old authority mentions it as a Dante, but Raphael took it for such and modern critics have generally agreed with him, Melcly^ Missirini^^ (in "Delle Memorie di Dante," 1830, p. 10), being perhaps the first to make the identification. The severe face with the projecting under Hp is em- 13 He writes : "Besides, the Great Masaccio [mistake for Filippino] painted the face and figure of Dante in one of the personages of the picture of the Martyrdom of St. Peter in the splendid chapel of the Carmine." The Dante is really in the adjoining sub- ject, St. Peter before the Proconsul. Missirini was content to get the wall right. V 58 THE PORTRAITS OF DANTE inently Dantesqiie and one of the finest portraits of the poet. In its pro- portions it generally "follows the Riccardian type, but it is essentially an original invention both in accessories and expression. The nose is straight. Filippino has endued with Dante's features the personage who plays the part of St. Peter's advocate in the legend. Incidentally he has given the inspiration for Raphael's type. A quite feeble and characterless Dante generally ascribed to Justus of Fig. 48. Benedetto da Majano, In- tarsia (detail) in the Palazzo Vecchio. From Passerini. Ghent and now in the Louvre (Fig. 51) demands our attention if only because it is presumably the first Dante that Raphael saw and because it left a certain influence on his type. This Dante was painted about 1475, by a Fleming, Joos van Wassenhove, who came to Urbino in 1473, ^^^ ^^^" cuted a series of famous writers ancient and modern for Federigo da Montefeltro's library. The best account of the matter is that of Venturi (Tom. VII, part 3, p. 124 fif.). Giusto da Gand, as the Italians called him, was little of a humanist and had presumably never read a line of Dante. For his portrait he apparently had no good Florentine exemplar. The Medal (Fig. 29), may have given him the wreath and other accessories. He seems to have taken the meaningless gesture of the raised hand from a fresco in Sant' Agostino at neighboring Rimini (Fig. 52). This fresco contains a figure, which while without authority of any sort, is thought by some to represent Dante. Justus seems so to have regarded it, and its straight nose may have seemed preferable to the ugly, hooked feature in the medal. In any case Justus arrives at a rather neutral and glum type, in its poor way quite his own. We find it echoed in still more woebegone THE PORTRAITS OF DANTE 59 Fig. 50. Dante by Filippino Lippi from the fresco of St. Peter before Nero, painted about 1484 in the Brancacci Chapel, Church of the Carmine. 6o THE PORTRAITS OF DANTE guise in an early sixteenth century portrait of Dante in the Pisa Museum (Holbrook i68). There is a better offshoot in the portrait of a wreathed and very youthful Dante formerly in the Morris Moore collection (Fig. 53). Mr. Moore wrote a monograph on the picture in which he ascribed it Braun Fig. 51. Dante by Justus of Ghent, now in the Louvre, from a series of famous authors painted about 1475 for the Library of the Palace at Urbino. From Berthier Fig. 53. The "Morris Moore Dante," present where- abouts unknown. Apparently of the Urbino School about 1500. to Raphael and traced it to Cardinal Bembo's collection. This view while not convincing is by no means impossible. The little panel is just what a boy of talent refining on Justus's portrait might paint. The relation which has already been noted by Koch and Holbrook is confirmed not only by the features but by the peculiar twist of the extra cloth at the front of the cap. Whether the "Morris Moore Dante" be by Raphael or by some more THE PORTRAITS OF DANTE 61 obscure Urbinate of his time, it at least illustrates the influence of a por- trait publicly shown in the famous library of a celebrated palace. Bernardino India's Dante which Dr. Holbrook associates, I think er- roneously, with that of Justus, will be considered later. Fig. 52. Alleged Portrait of Dante from a fresco in the Church of Sant' Agostino, Rimini, date about 1350. From Passerini. Raphael's Dante, like his work as a whole, is highly composite and eclectic. Unquestionably he had seen the Dante of Justus of Ghent in childhood; in young manhood at Florence he no doubt studied with all Fig. 56. Dante from Raphael's fres- co of the Parnassus, Vatican. Fig. 54. Dante from Raphael's fres- co of the Disputa, Vatican, care the profiles by Giotto, Taddeo Gaddi and Orcagna, the full length versions by Michelino and Filippino Lippi, and possibly the immensely aged effigy in intarsia by Benedetto da Majano. He probably knew the Riccar- 62 THE PORTRAITS OF DANTE dian type in the Medal, but I cannot see that it in any way influenced him, any more than Signorelli's Dante did. Raphael did three Dantes all of which, while differing in age and cer- tain details, have a consistent character. The earliest perhaps is the ravaged face which appears in profile, at the right of the earthly group in the Dis- FiG. 55. Raphael's Pen-drawing for the Dante in the Parnassus, date about 151 1, Albertina, Vienna. puta (Fig. 54). It is the face of a haggard and toothless sage. The nose is almost straight, the whole face slopes sharply forward to a very long chin, there is a slight projection of the lower lip. The cap has no hanging point but rather a fold, the ear-tabs are long and rounded. I doubt if it be possible to assign exact sources for this very original and imposing head. I believe Justus of Ghent and Filippino Lippi cause the rejection of the tra- ditional hooked nose. Raphael had very carefully studied the frescoes by Filippino and Masaccio in the Brancacci, and the whole port of his full- length Dante in the Parnassus seems to be suggested by Filippino. The THE PORTRAITS OF DANTE 63 character of the Dante of the Disputa is, I believe, inspired by Orcagna. Orcagna and Filippino supply the peculiar form of the head gear. I can find no trace either of the Death-mask or of the Naples Bust in Raphael. Indeed it is unlikely that they existed as early as 151 1. A little later perhaps, Raphael had to repeat the figure of Dante m the Parnassus. Appropriately the painter created for the Elysian fields of poesy a younger and less tragic type. We have it in a preliminary pen sketch in the Albertina at Vienna (Fig. 55) and in the fresco (Fig. 56). The Fig. 57. Title Vignette, woodcut, ' of the Convivio of 1521, Venice. From Holbrook. Fig. 58. Title Vignette of the Divina Commedia of 1529, Venice. posture and accent recall Filippino in the Brancacci Chapel. None of Raphael's Dantes preserve the true proportions of the authoritative types. It is clear that he found both Giotto and Taddeo Gaddi negligible for his purpose. Such was the common view of the time. While in artistic quality these Dantes in the Vatican are inferior only to the Naples Bust, they lack its insight, and a sound instinct has prevented the use of these much ad- mired Raphaels as a source for later Dantes. In printed books of the sixteenth century there are only three portraits of Dante interesting enough to excite curiosity as to their source. The earliest and best is in the Venetian edition of 1521 of the "Amoroso Con- vivio" (Fig. 57). The face seems to me to imply study of the Naples Bust or its equivalent, the form of the hood is that of Pietro Lombardi's relief 64 THE PORTRAITS OF DANTE on the tomb at Ravenna. A mountainous landscape is indicated as back- ground and the whole layout suggests a portrait of Giovanni Bellini's school executed before 15 15. The wood-cut may be based on such a lost portrait or may merely simulate the case. We have to do with the compo- sition of a man trained in the style of the Venetian early Renaissance. We learn from this portrait that the Naples Bust was extant at least as early as 1 52 1 and accessible to a Venetian designer. The wood-cut of Dante in the Divina Commedia of 1529, Venice (Fig. Fig. 59. Heylbrouck's engraving of Bernardino India's Dante from the Padua edition of 1727. Fig. 60. Dante from a picture as- cribed to Giambellino, but plainly derivative from India's, in the Zatta edition, 1757. 58), follows its predecessor in general arrangement but eliminates the land- scape and considerably changes the features. The forehead becomes nearly straight, the indentation of the nose disappears, the lower lip projects more noticeably, the cap-ties are treated as in Justus of Ghent's version at Urbino. We have possibly merely an attempt to age and revise, according to Boc- caccio's account or some Florentine portrait, an original rather close to Justus. Noteworthy is the huge, deeply cut eye which again is likely to come from Boccaccio. Sometime about 1570 Bernardino India of Verona made a bust portrait of Dante (Fig. 59), which had considerable vogue as a frontispiece in the eighteenth and early nineteenth century. It is remarkable for a great angry THE PORTRAITS OF DANTE 65 eye (from the Gaddi portrait?) and for a very pendulous nose, the chin is also puffy and heavy. The original painting has disappeared and is small loss, but in 1727, when it was engraved for the Comino, Padua, edition, it was still in the collection of Count Daniele Lisca of Verona. Its not very recondite origin seems to me to be the woodcut portraits of 1529 perhaps revamped on a portrait of the Riccardian type. The same ugly and sin^ter countenance appears, in reverse direction and provided with a meanly drawn wreath, in the fine Zatta edition, Venice, 1757. It is ascribed pre- posterously to Giovanni Bellini (Fig. 60) and located in the library of Fig. 61. Brush drawing, Florentine about 1540, in the Print Room at Munich. Fig. 62. Woodcut title Vignette of Sessa's "Divina Commedia," Venice, 1564. Marchese Jacopo di Dionisio Can. It is merely a variant of India's bad type. Dr. Holbrook dates the India portrait in 1472, on what authority I do not know. I shall tax the patient reader with only two more portraits and then our course together will have been run. In the Print Room at Munich is a brush drawing of very fine quality (Fig. 61), originally in the collection of Sir Thomas Lawrence and then ascribed to Raphael. Ernst Forster when he published the drawing in the "J^^rbuch der Deutschen Dante-Gesell- schaft," Vol. II (1869), pp. VII f. only bettered the matter slightly when he changed the attribution to Domenico Ghirlandaio. My own feeling was for an able Florentine artist of the transitional type of Ridolfo Ghirlandaio, who died in 1561. Mr. Berenson, who has kindly communicated his authoritative opinion, dates the drawing about 1540 and suggests, for type 66 THE PORTRAITS OF DANTE an early imitator of Michelangelo like II Rosso. We have to do with a masterly work, full of learning and delicacy. The profile seems to me al- most exclusively based on the Death-mask. The heavy and fleshy upper eyelid betrays the relation. The unknown artist has emphasized the massiveness and break of the nose. Possibly the sentimentalized head on the title page of the Fratelli Sessa's Dante, 1564, Venice (Fig. 62), depends on something similar to the Munich drawing. Otherwise I cannot trace any influence from a type which deserved more attention than it secured. Our story of the Dante portraits has its sadder side in the persistence of bad over good types. Happily Taddeo Gaddi's correct but uninspired record of Dante's features dominated Florence long enough to perpetuate itself in the Palatine Miniature and to produce the noble bust at Naples. It should be clear that we still lack a satisfactory semblance of Dante as he actually appeared in youth, middle age and old age, to his contempo- raries. Giotto's portrait has too much sacrificed the sharpness and char- acter of the actual features. In Taddeo Gaddi's portrait we may divine the irascible, political Dante, but the work is ugly and unskilful. The superb bust at Naples represents Dante with teeth at a time when we know he was really toothless. There remains a pious work of reconstruction for some modern painter or sculptor of genius to achieve. His firm data would be the Palatine Miniature and the reconstructed skull. His first task might well be to convey the aggressive character of the Palatine Miniature while amending its defects. We should then have the political Dante who ate the bitter bread of exile while yearning for an Italy unified under the secular regent of God, the Emperor. His next endeavor might be to give us the poet- mystic of the '-Vita Nuova," not in the celestial transfiguration of Giotto but with the keen, alert features faithfully transmitted by Taddeo Gaddi. His final emprise might be to figure forth the exile of Ravenna ; and poet of Hell, Purgatory and Heaven. Here would be necessary a very sensitive rehandling of the Naples bust, reducing it to toothlessness without making it merely decrepit or pitiful, preserving the fortitude and fire that burned to the end in the soul of the Divine Poet. BIBLIOGRAPHY It has not seemed necessary to list all books consulted in this study, but only such as have really aided me. Nor have I repeated here the numer- ous citations made only once and fully in the text and notes. Bassermann, Alfred, "Dante's Spuren in Italien." Heidelberg, 1897. Dante, "Tutte le Opere di Dante Alighieri," Ed. E. Moore, 3 vols., Ox- ford, 1895. HoLBROOK, Richard Thayer, "Portraits of Dante from Giotto to Raph- ael," London and New York, 191 1. Kraus, Franz Xaver, "Dante sein Leben und sein Werk," Berlin, 1897. Mather, Frank J., Jr., Review of Holbrook in Romanic Review, Vol. Ill (1912), pp. 1 16-122. Parodi, E. G., Bulletino delta Societa Dantesca Italiana, Vol. XIX (1912), pp. 89-106. Passerint, G. L., "II Ritratto di Dante," Florence, 192 1. Ricci, CoRRADO, "L'ultimo Rifugio di Dante Alighieri," Milan, 1891. Venturi, Adolfo, "Storia dell' Arte italiana," Tom. V. & VII, Milan, 1907. Volkmann, Ludwig, "Iconografia Dantesca," Leipzig, 1897. o a 1/3 C Q o Appendix I On the Reconstruction of the Skull from the Official Measurements There is no reason to doubt the official measurements as Welcker and Holbrook (p. 34) have done. Any error in the main curves and diameters would plainly show itself in the failure of the outline to join. It does join perfectly. Presumably the refinements of modern craniometry were not observed in 1865, and errors of 2 mm. or less are to be expected both in the longer measurements and in my reconstruction therefrom. Diameter AC ^187 mm. — from brow to occipital eminence. BI = 144 — from highest point to great foramen. Curve ABC = 317 — from eminence at brow to occipital eminence. CID =188 — from occipital eminence to centre of alveo- lar (under skull). Vertical JD = 85 — from root of nose to alveolar. Slope JH = 22 — length of nasal suture. Diameter EF = 39 — vertical diameter of orbit. " XG = 43 — from centre of base of orbit to alveolar of first bicuspid. " KC =189 — from tip of nose bone to occipital eminence. Angle L = 79° 34' — Camper's facial angle. All these dimensions are taken directly from Ricci's reprint of the "Re- lazione della Commissione governativa eletta a verificare il fatto del ritrova- mento delle ossa di Dante in Ravenna," Florence 1865. It will be seen that the minor measurement AJ is lacking. It has been supplied from the Princeton skull. Also the important diameter KC is not given as such in the official measurements. Upon it depends the projection of the nose bone. When I found that the nose bone projected outside of such contemporary portraits as Giotto's and Taddeo Gaddi's, I assumed that my reconstruc- tion of the nose from the Princeton skull was at fault, that it should be in- dented and less protruding. Then I managed to get the great horizontal curve 530 mm. which passes from the nose bone over the occipital eminence across the cheek bones. The coordinates were the width across the cheek bones 135 mm. and the transverse diameter from ear hole to ear hole 130. The longest diameter HC of this oval was 189 mm, which very gratifyingly corresponded to a millimeter with my theoretical reconstruction. In short the reconstruction is in this dimension more accurate than the old portraits. The significant measurement from the base of the nasal orifice to the alveolar — span of the upper lip — was not taken and had to be supplied, at 25 mm., from the Princeton skull. Since the base of the facial angle of Camper, 79° 34', is the line passing over the lower rim of the nasal orfiice from the condyles, my arbitrary measurement for the upper lip may have made an error in the angle. Such error, however, as the reader may satisfy himself by moving the base line up or down is really very slight and would not affect the general appearance. The forehead may well have been a trifle straighter than it is in my reconstruction. But again the fit with Giotto's and Taddeo Gaddi's portraits is so close (Figs. 10, 14) that any error must be insignificant. N \ '-- 1> Diagram II. Front view of Dante's Skull reconstructed to Scale from the Official Measurements. Front View The coordinates for this reconstruction are : BI = V^ertical diameter = 144 mm. NO == Extreme diameter on frontal suture 150 LM = Horizontal diameter above ear-holes 130 TU = Extreme diameter across cheek bones 135 PQ = Diameter outside of orbits 124 LBM = Curve from ear-holes over top 330 JD = Root of nose to alveolar 85 EF = Vertical diameter of orbit 39 RS = Horizontal diameter of orbit 41 XG = Base of orbit to alveolar of first biscupid 43 K = Nose bone is located from the profile The least satisfactory reconstruction is that of the curve LBM, for the reason that there is no sure way of locating the greatest diameter NO. These points have been fixed arbitrarily, considering the curve of the fore- head. They probably were not lower and they may well have been a half inch higher. Such a location would give the skull a much more wedge shaped look. The reader will realize that these elevations of the skull are to be re- garded as mere diagrams — graphic expressions of measurements, and that nothing like an anatomical reconstruction has been even considered. The diagrams are offered simply that my use of the measurements may be checked, and also in the hope that they may be useful to some modern artist who may wish to make a portrait of Dante. 72 THE PORTRAITS OF DANTE Appendix II — Tacca's "Head" of Dante It was Seymour Kirkup who dug up the old wives' tale of a stray skull (or "head") of Dante, in a manuscript volume of the Florentine antiquary Giovanni Cinelli (1623-1706). It is in the Magliabecchiana, N. IX, and contains a life of Dante. Therein is conveyed the following startling bit of information. I use Holbrook's translation (p. 57 f.). "His [Dante's] head was afterwards taken out of the sepulchre by order of the Archbishop of Ravenna and given to Giambologna, from whose hands, along with all the other things in small models and other materials, it passed into the possession of Pietro Tacca [died 1640], his pupil and heir. And this explains how, one day, as he was showing to the Duchess Sforza, among other pretty things and oddities, the head of Dante, she, with an imperious gesture, snatched it away and saw fit to carry it ofif, depriving at the same time Tacca and the city of so dear a treasure, to the very great sorrow of this Tacca, as I have often been told by Lodovico Salvetti, his pupil, and an eye-witness of this act. "This head was not very large in front, but constructed with a very great delicacy in the bones, and from the forehead back to the part called occiput, where the lamboid suture ends, it was very long. I mean it was not round like most heads but oval — a manifest proof of the profound memory of Fig. 63. Terra-cotta head of Dante in the collection of the late Count Paolo Galetti, Florence. this excellent poet — and because of its beauties it was often used as a pat- tern in drawing by Tacca's young men. The duchess, however, having put it in a green scarf, carried it away with her own hands, and God knows in whose possession and where so precious a thing may be now." I cannot agree with Holbrook that this delectable anecdote is to be sum- marily dismissed as a "legend." Something happened. Cinelli's informant was, or said he was, an eye-witness of the event and is circumstantial even to the sutures of the "head" and the hue of the enterprising Duchess's scarf. Holbrook and Parodi seem to me right in assuming that the actual head (skull) of Dante is intended. Otherwise the anecdote has little point. If as various critics including Ricci have thought, the treasure ravished THE PORTRAITS OF DANTE 73 from Tacca was not a skull but a sculptured head, it must have been of terra cotta. It was light enough to be carried off in a scarf by a noblewoman not accustomed to bearing heavy burdens. There is only one extant head of Dante that meets the case, namely a terra cotta head in the collection of the late Count Paolo Galletti at Florence (Fig. 63). This head was pub- lished by Passerini in the Giornale Dantesco, XXI (1913), p. 191, and again in his recent brochure. In both cases he treats this sculpture with becoming reserve. If it really antedates the last century — and I need to be convinced that it does — it is the sort of thing that might well have been done in the north in Giambologna's time. Indeed it somewhat resembles the Munich drawing (Fig. 61) and Sessa's woodcut (Fig. 62). It is also the sort of thing that a Giambologna might have admired and an archbishop need not have hesitated to give away. So much for bare possibilities. But Cinelli certainly seems to be talking of a skull. I think Tacca really had a skull which he and the Duchess Sforza believed to be Dante's. Of course it wasn't Dante's skull. That the Ravennese Franciscans had long ago prudently hidden beyond archiepisco- pal ken. It was in short some skull or other which the archbishop hoped or thought was Dante's. The age that has accepted the Death-mask is in no position to mock at the generation which believed Tacca's model skull to be Dante's. Ricci, who wished to see in the "head" a portrait by Tullio Lombardi and the source of the Death-mask, is fully aware of the weakness of his own position. In "Ultimo Rifugio," p. 28 1 f. he admits that Cinelli must be gossiping not about a portrait head but about a skull. c c c < I/) O 1—1 o «5 E^ 2 U ■" u a: O H < In U w ■J .W) O u b Q "—^ r~ ec! Q b 5^ < ^ •-J • i' w - ti, ' H ° O Co <2 OS H - C go <; CQ « be 1 ^ So"" ^....J ■pi OS OO i c ^ • Ow ® .S" PQ ,< - w J. <^^ > -^ E^ ^ .c-. r.; 5 « « E -tf ;^ o ^ ■= Eh N ft« — « o2^ *-! >■ S C U ■< (i — o S o - «U< >o Q <: < H •f r-t Q > •< > IE bi « o _ Z W o z "^ .2P X ■< "O z EC ;- o* -I— I >o WEh < w O •-SCJ O — N? r-- L-- S 1-3 .M) 1—1 «r5 1X1 - ^ 00 A SUMMARY CATALOGUE OF DANTE PORTRAITS A SUMMARY CATALOGUE OF DANTE PORTRAITS Only portraits of relatively true or early iconographic type or otherwise important are included. Uncertainty of identification is indicated by the interrogation point (?). 1300—1400 1. Dante(?) kneeling to receive a crown from Bologna, pen sketch o£ 1323 by the Bolognese notary Ugguccione Bambaglioli (Figs. 24, 25). p. 25 f. 2. Taddeo Gaddi's Portrait in fresco for Sta. Croce, Florence, probably painted before 1330. The fresco was destroyed in 1566, but it is almost certain that we have a faithful 15th. century copy in the Miniature frontispiece of Palatine Manuscript No. 320 in the Biblioteca Nazionale, Florence. (Frontispiece.) This miniature corresponds closely to the proportions of Dante's skull (Fig. 14) and is the most authentic likeness. pp. 11-18 3. Giotto's Dante in fresco in the Bargello, painted about 1336. Discovered in July 1840 by Seymour Kirkup, Aubrey Bezzi and Richard Henry Wilde. Promptly covered and disfigured by Marini's repaints (Fig. 9). Now best represented by Kirkup's chalk copy in the Vernon collection, England (Fig. 5), and in second order by his small color sketch (Fig. 6) and by the Arundel Print (Fig. 7). Before the restoration, the sculptor Perseo Faltoni (Fig. 8) made a copy now in the Berlin Print Room, which confirms the accuracy of Kirkup's work. pp. 6-11 4. Orcagna's Dante (Fig. 19), in the frescoed Last Judgment, Strozzi Chapel, Sta. Maria Novella. Contract 1354. A very old and toothless type with, con- trary to Boccaccio's evidence, a retracted under lip. The peculiar form of the cap before and behind influences the later tradition, as do the tied cap-strings. I think the haggard expression was in Raphael's mind when he created the Dante of the Disputa (No. 2']'). p. 21 f. 5 Dante(?) by Andrea Bonaiuti (Fig. 21) painted in fresco about 1360 for the allegory of the Church Militant in the Spanish Chapel, Sta. Maria Novella. Possibly a derivative of Giotto's Dante. pp. 23, 30 6. A LOST portrait by Lorenzo Monaco in an Ardinghelli Chapel, either in the Trinita (Vasari) or the Carmine (Milanesi). See Milanesi's Vasari Vol. II, p. 20, n. I. P- 25 79 8o THE PORTRAITS OF DANTE 1400—1500 7. A Head of Dante(?), drawing V. i, in the library of Christ Church, Oxford (Fig. 27) by a painter trained in the Gothic technique working in the early years of the fifteenth century. p. 31 i 8'. A Portrait of Dante in cx)lloquy with an old man in a Florentine street. Ord- ered by Fra Antonio Neri for the Cathedral about 1429. Replaced in 1465 by No. 18. For full particulars and inscription see Holbrook, p. 172 f. Since the inscription urges moving Dante's bones to Florence, and that city in 1849 formally memorialized Ravenna to that end, we have good warrant for the date 1429 as that of Neri's picture. P- 35 8*. The Riccardian Portrait. Colored frontispiece of Riccardian Ms. No. 1040 (Fig. 28). Probably a replica of the portrait ordered about 1429 for the Cathe- dral, and by Paolo Uccello or a follower. The first Renaissance Dante, freely altered from Taddeo Gaddi with the aid of Boccaccio's description, and pos- sibly with consultation also of Giotto's and Orcagna's portraits. pp. 32-34 9. The Trivulzian Dante. A panel in tempera in the collection of Prince Trivul- zio, Milan. A free variant of the Riccardian type, the head turned the other way and not quite in profile (Passerini, 9). Painted about 1475. p. 36 f. 10. The Eugenian Dante. In the Eugcnian Ms. of the "Divine Comedy" in the Palatine Library, Vienna (Passerini, 7). Chiefly based on the Palatine type but with features, in the cassock and ear-tabs, borrowed from the Riccardian. From Boccaccio the peculiarity of a slight beard. Probably drawn about 1450. P- 37 11. The Bronze Medal, exists in several sizes and versions (Fig. 29). A close derivative of the Riccardian, made possibly in 1465 for the second centenary of Dante's birth. P- 36 12. Botticelli's Dante. In the pen and silver point illustrations for the "Divine Comedy" made after 1481 (Figs. 31*, 31*). The splendid Ms. is in the Berlin Print Room, with a few stray sheets in the Vatican. The type varies freely, often recalls the Palatine Miniature, and is probably slightly influenced also by the Medal. The engravings in the Landino "Commedia" of 1481, which are commonly ascribed to Baccio Baldini, echo Botticelli with fair fidelity. For the entire Botticellian series consult the greater and lesser publications of Friedrich Lippmann, P- 38 13. Pesellino's Dante. In a cassone front in Mrs. John L. Gardner's collection, Boston, painted about 1450 and representing with other Petrarchan Triumphs the Triumph of Fame. The type is grim and old (Fig. 47) with a protrud- ing under lip. Probably based on the Riccardian or its original. P- 57 THE PORTRAITS OF DANTE 8i 14. Benedetto da Majano's Dante, an intarsia on a door of the Audience Cham- ber, Palazzo Vecchio. Full length, very old and grim with an enormous hooked nose. (Fig. 48 and Passerini No. 13). Somewhat similar to No. 13 and perhaps freely handled after the Medal or some other version of the Ric- cardian type. The long rounded ear-tabs, later a characteristic of Raphael's type, apparently are taken from Filippino Lippi's Dante (No. 25). Date of the intarsia uncertain but not far from 1490. P- 57 15. The Urbino Dante by Justus of Ghent (Fig. 51) now in the Louvre. One of a series of famous authors painted about 1475 for Federigo da Montefeltro's library in the Ducal Palace. The panels are now mostly in the Louvre and Barberini collection, Rome. The type is glum and characterless, and may be distinguished by the straight nose, and curving form of the ties of the ear- tabs. Facial proportions are closer to the Palatine than to the Riccardian. Raphael in boyhood undoubtedly often saw this Dante and its influence may be traced on his three portraits of the poet. The insignificant motive of the raised hand may have been borrowed from a fresco at Rimini (Fig. 52). p. 58 f . 16. The "Morris Moore Dante" (Fig. 53), formerly ascribed to Raphael and traced to Cardinal Bembo's collection by Mr. Moore in a monograph. Whereabouts unknown. Small panel with illegible cartellino, apparently of the Romagna school and earlier than 1500. It is simply No. 15 made youthful and idealized. While a rather feeble work, the ascription to Raphael's callow years is not preposterous. Nothing more definite can be said until the picture itself turns up. Excellent cut in Berthier's "Divine Comedy." p. 60 17. Andrea del Castagno's Dante (Fig. 35''). Full length in fresco for a Carducci villa at Legnaia. One of a series of famous men and women now in the Museum of Sant' Apollonia. Perhaps the earliest worthy example of the three-quarters type, though the date is uncertain. Andrea died in 1457. The facial proportions suggest that Taddeo Gaddi's portrait was the exemplar, but the changed aspect of Andrea's figure disguises any relation to earlier por- traits. Accessories freely handled. Color facsimile in Holbrook. p. 46 18. Domenico di Michelino's Dante (Fig. 37) in the Cathedral, Florence. Com- missioned Jan. 30 (old style), 1465. Partly designed by Baldovinetti. Repre- sents Dante standing before a gate of Florence, with the mountain of Purga- tory behind him, the portal of Hell at his right, and the seven circles of the celestial bodies traced in the sky. Dante holds the Divine Comedy in his right hand. The type is mild and melancholy in a gentle idealistic way. The posi- tion is guided by No. 17, the facial proportions suggest the Riccardian. Dome- nico may well have adopted some traits from the old picture of 1429 (No. 8") which his own replaced. Color facsimile in Holbrook. p. 47 f. 19. Antonio Pollaiuolo's Dante (Fig. 39"). Drawing in the library of Christ Church, Oxford. According to Berenson, "Drawings by the Florentine Painters," a contemporary copy after a sketch by Antonio. Excellent repro- 82 THE PORTRAITS OF DANTE ductlon. Dante is represented at full length with his book, in three-quarters aspect. Influenced by No. i8. Date probably after 1475. Antonio died in 1498. p. 49 20. PiETRO LoMBARDi's Dante, marble low relief (Fig-. 23) in the tomb at Ravenna, 1483. The type is nondescript. This reading Dante takes its motive from a fresco in S. Francesco (Fig. 22) at Ravenna. It is ascribed to Giovanni Baronzio, active about 1350, and has apparently from an early time been re- garded as a Dante though having no resemblance to the standard portraits. Holbrook's doubt (p. 54 f.) as to the originality and period of Pietro Lom- bardi's feeble portraiture seem to me unnecessary, for the influence of Pietro's version is shown in woodcuts before 1500. But, judging from the photograph, the marble may have been drastically refinished in a more modern manner during one of the various restorations of the tomb. p. 24 21. The Paris Miniature Dante, illustration for Inferno I, in Ital. Ms. 74, Bibl. Nat. (Fig. 43), date about 1400. The larger figure has a straight nose with the lips drawn back upon toothless jaws. This is the earliest example known to me of a straight nosed toothless type which ends with Signorelli (No. 26). Orcagna's Dante (Fig. 19) is toothless, but the later portraits which show this trait are so different and so far from any true tradition that one may suppose they rest upon some oral report of Dante's toothlessness in old age. P- 55 22. Giovanni dal Ponte's Dante (Fig. 44), on a cassonc end in the Fogg Museum, Harvard University. Dante with a fellow poet is represented at full length being crowned with a wreath by a flying genius. Date about 1440. Straight nosed and toothless, this head or its equivalent should be Signorelli's exemplar. P- 55 f- 23. Dante on a Jardiniere in the Uffizi (Fig. 49). In a Triumph of Fame later than 1450 in a style akin to that of the "Cassone Master." An aged straight nosed and toothless type so near Raphael's that one is inclined to imagine that he may have seen this piece in the old Medici collections. P- 57 24. Dante in the Urbino Codex (Fig. 45). This extraordinary Ms., Urbin. 369 of the Vatican Library, is not wholly consistent in its portraiture of Dante, but generally the face is toothless with a huge aquiline nose. Possibly we have to do with a fusion of the cassone type (No. 22) and the Medal (No. 11). Date about 1480. Many fine cuts in Bassermann. PP- 55. 57 25. Filippino Lippi's Dante(?) in the fresco (Fig. 50) of St. Peter before Nero in the Brancacci Chapel, Carmine. Date about 1484. This is not explicitly a portrait of Dante but rather a personage of the story with Dante's features. Raphael took it for a Dante, for he used it for his own type (Nos. 27-29). The iconographical peculiarities are the long rounded ear-tabs which here appear for the first time, and the straight nose. I can find no sure trace of earlier portraits, though the intentness of the expression and the facial pro- portions may echo the Riccardian (Fig. 28). p. 57 f- THE PORTRAITS OF DANTE 83 1500—1600 26. SiGNORELLi's Dante (Fig. 46) in the dado of the Chapel of S. Brixio at Orvieto. A reading type, straight nosed and toothless, derived from a miniature — cassone type similar to Giovanni dal Ponte's, No. 22. The several small portraits of Dante in the same decoration, in illustrations of the "Divine Comedy," are not characteristic, but often have the pointed ear-tabs with crinkled ties of the Palatine type. They are probably by an assistant. PP- 55, 57 27. Raphael's Dante in the Disputa (Fig. 54) fresco. Date about 151 1. A very composite type which draws from Justus of Ghent (No. 15) Filippino Lippi, the rounded ear-tabs, (No. 25), and possibly, for expression of eager senility, upon Orcagna (No. 4). It is drawn exclusively from eccentric types, including possibly No. 23, and is wholly out of the true tradition. pp. 61-63 28. Raphael's Albertina Drawing (Fig. 55) full length pen study for the Dante of the Parnassus. A younger version of No. 25. Date and derivation as above. p. 63 29. Raphael's Dante of the Parnassus (Fig. 56), Date and derivation as above. p. 62 30. The Naples Bust (Figs. 32, 33), nearly life sized, in the Farnese Collections, Museo Nazionale, Naples. Probably modelled by a Florentine imitator of Donatello in the early years of the i6th century. This is the artistic consum- mation of the grim type of Taddeo Gaddi. It agrees very closely in facial proportions and essential iconographical features with the Palatine Miniature (Fig. 34). But the artist also knew the Riccardian type, adopting the form of the cap, the rigid ties, small eye, retreating forehead, and protruding under lip. If I am right in detecting the influence of the Bust in the woodcut of 1521 (No. ;i2), we have a terminus ad qucm for the Bust. PP- 39-41 31. The Torrigiani Death-Mask, a very high relief in colored plaster, given in 1840 by Marchese Pietro Torregiani to the Uffizi. Not surely traceable before 1735- The many other versions are recent casts (Passerini, 24-29). The profile, except for the horizontal setting of the eye, agrees exactly with the Riccardian Portrait (Fig. 30). That is the main source, but the modeller may have also seen the Naples Bust or an equivalent. Probably a falsification of the second quarter of the i6th century, based on the discarded portrait of 1429 which was still in the Cathedral precincts. If the Munich drawing (No. 36) is derived from the "Mask," the latest date for the latter is 1540-1550. pp. 6, 36 2,2. Title Woodcut of the "Convivio" of 1521 (Fig. 57). Composed like an early Renaissance Venetian portrait with a mountainous background, and possibly based on such an original. Apparently chiefly influenced by the Naples Bust, for which it afYords a terminal date. The headgear may be taken from Pietro Lombardi's tomb relief (No. 20). p. 63 f. 84 THE PORTRAITS OF DANTE 33. Title Woodcut of the Divine Comedy of 1529, Venice (Fig. 58). An ambigu- ous type pretty freely invented with, possibly, consultation of the Medal (No. ,11), the Tomb Relief (No. 20) and Justus of Ghent (No. 15). p. 64 34. V.'^s.ARi's Dante in a Group of Poets, 1544. A version in the Senior Common Room of Oriel College, Oxford (cut, Holbrook, p. 154) and another now un- located, formerly in the Orleans Gallery represented by an engraving (cut, Holbrook, p. 156). This appears to be an effeminated and sentimentalized ver- sion of Giotto's Dante (No. 3) in the Bargello. The reading pose with head turned over the left shoulder from Signorelli (No. 26). p. 30 f. 35. Vasari's Dante the Head only (Fig. 26). Oil painting on panel, Cleveland Museum of Art. A nearly contemporary copy with the oval ear-tabs narrowed after Raphael's precedent. p. 31 36. The Munich Drawing (Fig. 61). A fine brush drawing in the Print Room. Berenson considers it by a follower of Michelangelo of Rosso's sort and dates it about 1540. It seems to be freely modified from the Death-mask, for which it perhaps affords an approximate latest date. p. 65f. 37. Oval Woodcut in Sessa's "Divine Comedy," Venice, 1564 (Fig. 62). Dante with a huge nose. Apparently drawn from an original similar to the Munich Drawing, and perhaps influenced by No. 34. p. 66 38. The Galletti Head in Terra Cotta (Fig. 63 and Passerini, 29, 30). Similar to the last two numbers but of entirely uncertain period. At best a work of about 1600. P- IZ 39. The Yale Dante probably after Bronzino (Fig. 40). Half length with book in right hand. A direct derivation of Domenico di Michelino's Dante (Fig. 38), the narrower look of the face being mostly due to the heavier shading in the later work. Regarded as an old copy from Bronzino by Dr. Siren in his "Catalogue of the Jarves Collection," No. 85. A variant without the hand is in the Uffizi (Holbrook, p. 184). In 1757 Raphael Morghen (Fig. 39) engraved the frontispiece of the beautiful Zatta Dante after Tofanelli's drawing from a picture of this type. This gentle and somewhat woebegone presentment of the poet has been unduly popular as a frontispiece almost till today. p. 48 40. Bernardino India's Dante engraved by Heylbrouck as frontispiece for the Comino, Padua, Dante of 1737 (Fig. 59). Then in the collection of Count Daniele Lisca, Verona, present whereabouts unknown. Dated by Holbrook in 1572. A morose and disagreeable type of uncertain origin. I think the Medal (Fig. 29) and the Venetian woodcut in the Commedia of 1529 (No. 33) may largely account for it, but the huge eye suggests the reading of Boccaccio. I THE PORTRAITS OF DANTE 85 list it only because of its great vogue in the i8th. and early 19th. centuries. On the margin of the example of the print in the Princeton University Library an outraged eighteenth century purist has scribbled the comment "blood, lust, dung." If the sentiment applies only to the portrait, I largely agree with the writer. p. 64 f. 41. The "Bellini-India" Dante, represented by an engraving in the Zatta Dante, Venice, 1757 (Fig, 60). Reversed from the India type with a wreath added. The original was absurdly ascribed to Giambellino and preserved in the col- lection of Marchese Jacopo di Dionisio Can, Venice. Present whereabouts unknown. Listed only because the lost original seems to have been of the 16th. century. p. 65 THE END THIS BOOK IS DUE ON THE LAST DATE STAMPED BELOW AN INITIAL FINE OF 25 CENTS WILL BE ASSESSED FOR FAILURE TO RETURN THIS BOOK ON THE DATE DUE. THE PENALTY WILL INCREASE TO SO CENTS ON THE FOURTH DAY AND TO $I.OO ON THE SEVENTH DAY OVERDUE. '^MMi M^ 30 1934 OCT 24 '9m 16iul'53KW ^•^-* 1953 L AUG 30 1960 INTER UBRARY LOAM tl^ON-RENEWABLE MAR 18 1255 15Ju)*66Wr REC D LD JUL 9 *65-5PM LD 21-50m-8,32