Three Spanish terracottas in the Victoria and Albert Museum

THREE SPANISH TERRACOTTAS IN Tffi VICTORIA AND ALBERT MUSEUM Marjorie Trusted Polychromed terracotta works produced in Spain from the late i ® onwards are generally assumed to be freshiy modelled, exhibiting imp plicitly the «fingerprints of the Sculptor»^. Many of them are partly cast from moulds, and by analysing the way they were ma e, inferred about workshop practices. Three Spanish terracottas in e Albert Museum, dating from different periods, ¡Ilústrate throug sity the complexities of modelled and cast terracottas. . The earliest to be discussed is a polychromed terracotta fi gure o a ' ed by Pedro Millán (active 1487-1506)^. A youthful saint is s creature mour, under a cloak held by an elabórate clasp, trampling a „ . underfoot. His shoulder-length hair is held by a band adomed with ® towards his right, and originally held a lance in his right hand, whic en er dragón at his feet, and with which he appeared to support himself, t is ^ ing, and has been since the fi gure was fi rst published in 1884 . (Jn t e s i cross standing on a crown of thorns surrounded by a vine is depicted in re le . ' I am grateful to Paúl WilHamson and Anthony Radclíffe for their helpftil comments dur' g preparation of this article. . . - C. f. Charles AVERY. «Terracotta: the fingerprints of the Sculptor», Although pean Terracotta Sculpture fr om the Arthur M. Sackler Collection, Washington, . Pthe piece cited here discusses Italian modelled terracottas, works from other countnes a frequently assumed in other publications to be modelled rather the cast, not always wit jus ' ' 3 Victoria and Albert Museum. inv.no. A6-I943. Ht. 91 cm. Said to have come fr om the Dominican Convent of S. Florentina at Ecija (Seville) (see J. HERNANDEZ DI e .. Arqueológico y Artístico de la Provincia de Sevilla, III, Seville, 1951, p. 327, note VON Irueta Goyena Collection, 1884-before 1923; «Prívate Collection, Seville» (so descnbed J" ' loga, Spanische Plastik, Munich, 1923, pp. 14-15) (perhaps Irueta Goyena Collection) 1927; whereabouts unknown 1927-1943 (J. PUOAN, History of Art (translated by R. L. KU h Barcelona, 1927, p. 446, says it «has now been carried off to a foreign land»), so y • ® Sotheby's, London, 16th April, 1943, lot 98; bought by Dr. Burg ( £ 60); bought from Dr. i^g y the National Art-Collections Fund in 1943 ( £ 400) and presented by the Fund to the Victoria an ^ Museum in the same year. In María Elena GOMEZ-MORENO, Breve Historia de la Escu tura pttti (2nd ed.) Madrid, 1951, p. 62, the piece was mistakenly said to be in América. See J. GESTOSO Y PEREZ, Pedro Millán, Seville, 1884, frontispiece. Zli 322 MARJORIE TRUSTED armour is thought to be based on Itaiian, probably Milanese, armour of about 14805. Previously described as St. Michael, the iconography suggests the more likely indentification is St George, primarily because the saint is without wings^. The work is signed in gothic lettering: «p5 milla» on the lower part of the shield. The same signature appears on other works by Pedro Millán: on the girdles of two seated monks near the jambs of the West Door of Seville Cathedral, and on the Virgin of the Pillar also in Seville Cathedral'. Millán was almost certainly a native of Seville, and specialised in terracotta; other documented works by him in this material ¡nclude the glazed reliefs of saints for the portáis of the church of Santa Paula in Seville, whose indebtedness to contemporary della Robbia glazed terracottas, as well as to Iberian ceramic tiles, has been remarked upon^. Although the present piece has been dated c 1480-86, none of Millán's works is securely dated, and this figure can only be put within his documented period of activity (1487-1506)^. The rich detail on the shield and armour, the elegance of the pose, and the finely featured face of the saint are all stylistic indications that Pedro Millán is the author of this piece, even were it unsigned Although all but the later layers of polychromy have been largely lost, minute surviving remains of pigment and gold leaf indicate the original colours included gold and ultramarino, and the work must have seemed a radiantly coloured chivalric fi gure, perhaps designed to stand in a niche, or against a pillar in a church, as it is unworked at the back''. Despite its fresh appearance, there are indications that the Saint George was almost certainly cast from a mould: through the large opening in the fíat slab of ' I am grateful to Anthony North for his comments on the indentifícation of the armour. ^ See W. BRAUNFELS (Ed.), Lexikon der Christlichen Ikonographie, VI, Rome, 1974, p. 380: «Unter d. ca 55 DrachekámpferhII. nimmt Georg n. d. Erzengel Michael i. Abendl. d. I. Platz ein, alie anderen hil. Reiter verdrángend». See also the whole entry on St George in this volume (pp. 366-390), and E. KIRSCHBAUM (Ed.), Lexikon der Christlichen Ikonographie, III, Rome, 1971, pp. 256-266, for the entry on the Archangel Michael. The figure was described as St. George in the Sotheby's sale catalogue of 1943 (see note 3), and a letter from Sir Eric Maclagan to Sir Robert Witt, Chairman of the National Art-Collections Fund, of 7th June 1943, a copy of which is held at the National Art-Collections Fund, notes, «It is generally described as St Michael, but I am inclined to think in spite of the long hair that it must be intended for St George». In subsequent publications it is nevertheless described as St Michael, e. g. M. THRING, «A Statuette by Pedro Millán of Seville», Buriington Magazine, LXXXni, 1943, pp. 253-5; A. DURAN SANPERE and J. AINAUD DE LASARTE, Ars Hispaniae, VIII, Escultura Gótica. Madrid, 1956, p. 368; Florentino PEREZ-EMBID, Pedro Millón y los orígenes de la Esculura en Sevilla, Madrid, 1973, pp. 41, 42, 54, 66, 73-76. M. E. GOMEZ-MORENO, op. cit., loe. cit. in note 3, calis the piece St George or St Michael. ^ PEREZ-EMBID, op. cit. , figs. 2, 3, and 10. » See PEREZ-EMBID, op. cit., pp. 67-73, and fi gs. 16-18. ' GESTOSO, op. cit., pp. 48-57 dated the piece c. 1480-86, but PEREZ-EMBID, op. cit., pp. 73-76 rightly said that adequate reasons were not given for this dating. A particularly cióse comparison of "the facial type, and the leaning slender fi gure can be found in the signed Man ofSorrows in the Museo de Bellas Artes in Seville. See PEREZ-EMBID, op. cit. , fi g. II. " Cf. the same artist's Virgin of the Pillar in Seville Cathedral mentioned above (illustrated in PEREZ-EMBID, op. cit., fi g. 10). 323 THREE SPANISH TERRACOTTAS clay attached to the back, it can be seen that the fi gure is hollow. The clay is of an even thickness of approximately 2.5 centimetres, something whic been achieved had the fi gure been modelled and then hollowed out e origina model is likely to have been of solid, modelled clay, probably constmcte over an armature support. A plaster mould must have been made from this mo e , e un er cutting necessitating a mould made in pieces'^. The original clay mo e ^ been spoilt by this process (fragments of the plaster would have become ® ® in it, and if brass fences or «shims» were used, cuts would have scarre e su face), and would have been discarded. The mould too was probab y a an on ^ use'"*. When the sepárate cast pieces of clay were removed from eir moulds, they would have been joined together using wet clay to ma ® ® fi gure, and the joins disguised by modelling. Other surfaces, suc as e shield, might have been touched up by modelling at the same time. is ca would then have been fi red, and fi nally painted. In this case, the reason or c was almost certainly to reduce the risks of damage during the fi ring process, ra than for reproductive purposes. As suggested abo ve, the technique o cas bably employed in this case meant that both the original mode an t e mo only used once; certainly no other versions of the fi gure are known. m terracotta, even if it has been hollowed out, is likely to be of an uneven i ' and in a fi gure of this comparatively large size there would be a strong pos that it would crack during fi ring'^. ., . • l The second piece has not been previously published, despite its evi , quality. It is a polychromed terracotta relief depicting The Lamentation, in a painted and gilt wood frame The Virgin supports the supine y ? , . at the foot of the cross. She leans back, her head sideways over er e s ou , Christ's body is twisted, with his left arm stretched back behind him, and his legs slightly bent. In the landscape behind, a man is seen carrying away a a er Plaster infill at the back of the figure perhaps indicates a break which occurred I am grateful to John Larson. Anne Brodrick and Richard Cook for their comments on the way m wn.cn this piece was ¡nade. ,, . .. For a brief description of the making of a mould, see [Giorgio VASA ], ^ isao que, translated by L. S. MACLEHOSE. ed. G. BALDWIN BROWN. New York. »960 pp. 158-9^ '■* For this reason, such a mould is known as a waste-mould. For a description o is . , using brass fences, see J. W. MILLS, The Technique ofSculpture. London, 1976, p. 53^ee aiso LARSON, «A Polychrome Terracotta bust of a Laughing Child at Windsor Castle», The Bur tng Magazine. CXXXI, n." 1038, September, 1989, pp. 620-621. See LARSON, op. cit. in note 14, p. 620. For the technical aspects of firing terracotta, se also M. GREENACRE, «A Technical Examination of Some Terracottas by Michael Katharine EUSTACE (ed.), Michael Rysbrack Sculptor 7(594-/770 (exhibition catalogue), pp. 50-51. Greenacre cites in addition {op. cit., p. 59, note 20): MILLS, op. cit., pp. 4 ' o » ' John LARSON, «Carrier-Belleuse, a technical study of his terracotta sculpUire», French Sculpture. Bruton Gallery, Somerset (exhibition catalogue), London, 1981. Victoria and Albert Museum, inv. no. 91-1864. Measurements with frame: Ht. 45 cm.; W. 55 cm.; Depth: 6 cm. Measurements without frame: Ht. 30.5 cm.; W. 40 cm. Bought for £.4 on 22nd December, 1863 by John Charles Robinson in Madid (on behalf of the then South Kensington Museum) from don José Calcerrada. (Victoria and Albert Museum records, J. C. Robinson report Reg. no. 2931, dated 15th February, 1864). 324 MARJORIE TRUSTED the right, while on the left a figure is bending over the tomb, seen through an opening ¡n the hillside. The crown of thorns and a cloth are resting on the trunk of a small bush to the left. Painted in the distance are the towers and buildings of Jerusalem. The frame is inscribed in gold: QVOS OMNES QVI TRANSITIS./ FERVIAM. ATENDITE ET VIDE[T]E SI EST./ DOLOR SIMILIS, SICVT. DOLOR MEVS / QUINET TVAMIPSIVS ANIMAM PENE[T]R[A]BIT GLADIVS (Is it nothing to you, all ye that pass by? Behold, and see if there be any sorrow like unto my sorrow. And why will not the sword pierce your soul also?) The bottom of the frame is elaborately shaped, with scroll-like corners at the lower edge, and shallow curves scooped out of the upper edge, giving the effect of a cartouche. There are breaks in the terracotta by the root of the tree on the right, and Christ's left foot is missing. Remains of sealing wax on Christ's left leg indícate an oíd repair. Minor breaks can be seen at the top of the terracotta. The wood frame is chipped in places and a sliver is missing on the lower left. The paint is flaking in places Fragments of the original polychromy lie beneath the now muted tones of the surface, and indícate that the relief was originally much brighter than its present appearance suggests. Traces of azurite, a brilliant blue, were found in the Virgin's robe, and gold leaf was certainly present elsewhere on the robe, on the trees behind, and on the ground beneath the Virgin's hand. A layer of red bole found over the preparatory gesso, and dating from the time when the relief was first polychromed, indicates that gold leaf may have been present over the whole surface, over which paint would have been applied and then scratched through to reveal patterns of gold (a technique known as estofado) The terracotta was overpainted at least twice, once probably in the seventeenth or eighteenth century, and once in the nineteenth century^®. The frame too seems to have been overpainted; one of the original gold letters of the inscription is visible^'. Other versions of this relief are in the Museo Arqueológico in León, in the Museo Diocesano y Catedralicio in Valladolid, and in the Camón Aznar Collection formerly in Madrid, and now Zaragoza. With the exception of the Camón Aznar relief, they are virtually identical in size^^. The mannerist style of the composiThe first part of the inscription (up to «sorrow») is from Lamentations, 1, 12. The present author has not been able to trace the source of the fi nal sentence, which is incribed along the bottom of the frame. The Latin appears to be corrupt and may be the result of a later re-painting of the lettering. Infrared photography did not reveal an altemaiive wording beneath the present surface; 1 am gratefiil to Stanley Eost for undertaking this. I am grateful to Caroline Elam and Rowan Watson for their comments on the Latin inscription. '* The piece was recently cleaned by Sarah Boulter. I am grateful to her for her comments on it. Red bole is normally used as preparatory layer for gold leaf. Samples of polychromy were analysed by Josephine Darrah, to whom I am grateful for her helpful comments. The fi rst re-paint contains at least one colour (bice) which was not generally used after the eigh teenth century. The second repaint, which is in parts thick, and of duller colour, must have been applied before the relief was acquired by the Museum in 1864. The inscription now visible appears to be the same as the earlier one. Juan José MARTIN GONZALEZ, Juan de Juni: Vida y Obra, Madrid, 1974, pp. 115-117. According to Martín González (op. cit., 1974, p. 378), the measurements of the relief in León are; Ht. 30 cm.; W. 39 cm. M. Gómez-Moreno states that the measurements are Ht. 31 cm.; W. 40 cm. (M. GOMEZ-MORENO, Catálogo Monumental de España: Provincia de León. (1906-1908), Madrid, 325 THREE SPANISH TERRACOTTAS tion, with its elongated, contorted fi gures, is typical of the work . (b.c. 1507; d. 1577), to whom the versions in the Museo Diocesano y Ca ra i in Valladolid and in the Museo Arqueológico in León respective y ® attributed^^. In particular, a Pietá in the Cathedral at Salamanca, a mos ce dating from around 1540, is closely comparable^'*. A provided by a fi gure group of the Virgin with the dead Christ attri u de Juni in the Museo Marés, Barcelona^^. j i • K;,v,cí»if Martín González has argued that the «original» versión by uan e uní is the ene in Valladolid, which may in tum be identical with e re ^ in the inventory of the belongings of Juana Martínez, of the o . , .u in Valladolid. This inventory was made on 10 January, 1613, ^ y a . of Juana Martínez. She was the widow of Isaac de Juni, Juan ® frnm son, himself a Sculptor. The inventory lists a relief in terracotta o e the Cross, although it must remain speculative whether this is in reliefs under discussion, as Martín González notes Palomino stat 1925, p. 313). The discrepancy of 1 cm. in height and width must be due ' !.7n gONZ^ zález gives the same measurements (30 x 39 cm.) for the ene in Val a o i ( , -¿qo relief (30.5 op. cu. . 1974. p. 379). These measurements are sufficiently cióse to those of the X 40 cm. according to the present author's measurements) to suggest e 23). However in size, and apparent differences are due to manual errors of measuring (see ^ ^ jj^is was the measurements of the versión in the Camón Aznar collection are. t. cm., ' . ^ kmdly communicated by Belén Díaz de Rábago Cabeza. « to are virtually the same: Ht. 25 cm.; W. 34 cm. (¡bid.. p. 378). T nnhlished in manual error in measuring a slightly irregular piece. The Camón Aznar relie was ^ P Museo Camón Aznar, Obra Social de la Caja de Ahorros de Zaragozxi. ragon y loja, unnumbered píate. „ , . r»f these " The difficulties arising from trying to distinguish autograph rom wor s op reliefs can be seen from the fact that Martín González ascribed the one in e n o n-Joiifi Museo in Juan José MARTIN GONZALEZ, Juan de Juni y su Epoca (exhibition catalogue), Valladolid Nacional de Escultura (Iglesia de la Pasión) and Madrid, Salas de la Artístico y Cultural, 1977, p. 51, cat. no. 2. and p. 96. pl. 2. He cited in note 22), p. 313. There Gómez-Moreno described the piece as: «bajorrelieve de barr po su tamaño, 31 por 40 centímetros; roto. Obra de Juni, conocida por otro ejemp ar '8" González San Martín de Valladolid; representa la Piedad, con paisaje y fi gurillas por on o». . •. also attributed the Valladolid relief to Juan de Juni in the text of his monograph, ^ ® the León and Camón Aznar pieces, which were said to be «after» the Valladolid ZALEZ, op. cit., 1974, pp. 115-7). However in the catalogue at the end of the „ González described the León versión as a product of Juni s workshop (MARTIN G cii. . 1974, p. 378), but by implication both the Valladolid and the Camón Aznar versions hste as autograph; they are included in the catalogue of works without the word (workshop) (ma TIN GONZALEZ, op. cit., 1974, pp. 378 and 379). In correspondence. Sr. Martín González has co - fi rmed that he believes the Camón Aznar relief to be workshop. This example is discussed fi irther below^ Martín González has pointed this out. See MARTIN GONZALEZ, op. cit.. 1974, pp. and fi g. 101. " ¡bid. , pp. 118-120, fi g. 80. ■ I K H I The wording of the inventory is «una echura de una ymagen de barro de medio rreliebe dei descendimiento de la cruz». This is quoted in J. MARTI Y MONSO, Estudios histéricos-artísticos principalmente a Valladolid, Valladolid, 1898-1901, p. 369, which is cited in MARTIN GONZALEZ. op. cit., 1974, p. 115. An earlier inventory dating from December 1596 to May 1597 lists «122 piezas de barro todas modelos de su mano ansi ystorias come fi guras pequeñas de las por acabar». See M. 326 MARJORIE TRUSTED sion now in Valladolid (then in the church of S. Martín in Valladolid) was cast by «algunos escultores» which was why various other versions existed^^. However, with one probable exception, the known surviving pieces are probably contemporary with one another, whereas the theory that copies were made by other artists implies that they were made at a slightly later date, perhaps after Juan de Juni's death in 1577. There is no perceptible difference in the defmition of the forms in the examples in Valladolid, León and London, although the polychromy varies in quality. The Valladolid, León and London pieces are all of the same size; the Camón Aznar versión is approximately 5 cm smaller in both height and width. This piece will be discussed separately below^®. It also seems likely that the other three known versions are cast, and not modelled; there are no physical signs to suggest any one was made differently from the others^^. In addition, these works seem to have been made for display, rather than for study reasons. (Palomino's comment could be read as implying that other artists made copies for their own use, in adulation of Juan de Juni). This can be inferred from the fact that the surviving versions are polychromed, and that in all cases polychromy contemporary with the reliefs survives at least in part, and that wood frames are extant on all but the Valladolid versión^®. For these reasons, it seems possible, and even probable, that the composition was always intended to be reproduced in múltiples, and that the versions in Valladolid, León and London should be accorded equal status, no one of them necessarily being closer to (or fi irther from) the work of the Sculptor than any of the others. Having modell ed an «original» in clay, which was probably then fired in order to harden it, Juan de Juni could have had made in his workshop a mould of this, from which were cast numerous copies^'. These were then painted, framed and sold to customers^^. A. FERNANDEZ DEL HOYO, «Datos para la Biografía de Juan de Juni», Boletín del Seminario de Estudios de Arte y Arqueología, LVII, Valladolid, 1991, p. 339. " Palomino states: «En la iglesia de San Martín de dicha ciudad [Valladolid] hay una historieta de barro cocido del Descendimiento de la Cruz, que le han vaciado algunos escultores por ser cosa tan peregrina». Antonio PALOMINO, El Parnaso Español, [ 1714] in F. J. SANCHEZ CANTON, Fuentes Literarias para la historia del Arte Español, IV, Madrid, 1923-41, p. 76, cited in MARTIN GONZALEZ, op. cit., 1974, p. 115. See above, notes 16 and 22. " Unfortunately the author has not been able to examine the backs of any of the reliefs. (The frame on the London versión could only have been removed at great risk to the terracotta). The front of the versión in the Victoria and Albert Museum has been closely studied, but the author has only seen briefly those in Valladolid and León, and has not seen (except in reproduction) the versión in the Camón Aznar Collection. This comment is therefore based in part on the fact that MARTIN GONZALEZ, (op. cit., pp. 115-7) does not discuss any physical dissimilarities that would suggest one of the reliefs was modelled rather than cast. In correspondence Sr. Martín González has kindly communicated that because the Camón Aznar versión is likely to be workshop, it is not reproduced in his monograph. The wood frames are not decorated in the same way: the León versión is inscribed «O mater Dei, memento mei...», while the Zaragoza piece is painted to imitate marble (perhaps a later repainting). See MARTIN GONZALEZ, op. cit., 1974, pp. 116-117. " The piece from which the mould was made is likely to have been fired in order to harden it. J. W. Milis notes, «Although it is possible to make a piece-mould from clay it is not as precise as that taken from a hard surface». MILLS, op. cit. , p. 55. The painting is likely to have been carried out in Juan de Juni's own workshop, rather than to have been sub-contracted. See MARTIN GONZALEZ, 1974, p. 58. 327 THREE SPANISH TERRACOTTAS The relief mentioned in the 1613 inventory (if it does have the present pieces) may be a fifth versión which has been su ' may be identical with ene of the surviving examples in V a o i , don. Although Juana Martínez lived in Valladolid, insufficient evi ene suggest it is the piece now in Valladolid. ^ thrftpThe relief in Zaragoza is a reduced versión of the composition o cmaller it does not seem to have been cut down, despite the fact that d is a ou ® _ than them. This could be accounted for by jwstulating w/jnt nressed mould had been made from one of the surviving (larger) re le s, ^ ^ for the into this, and then fired, the shrinkage caused by the fi rmg difference in size^^. This would also explain why the foms ^e es ' it was made at one remove. It may be contemporary wi ® ° for the is more likely to date from a later period, after the original mou other pieces) was lost or destroyed. . j • The London versión should be assigned to the workshop 'í® can be tentatively dated to the 1540s, approximately contempoi^ ^nHed to be hung ca relief mentioned above. It was probably a devotional o jec , m reinforces in a chapel or church, or in a private house. The inscription on e • » work its devotional purpose. While this attribution detracts from the idea o q of art, and fromThe conception of the artist as personally it conversely illuminates the way in which workshops opera , —e-es were knowledge of the kinds of objects which were being produced. Suc p ready-made items, rather than commissions from individu Virein The third work to be discussed is the polychromed terracotta group ^ Appearing to S. Diego ofAlcalá^^ The Virgin and Child ^ f cross to a kneeling Franciscan saint, thought to be S. Diego of Alcala a P by two angels, one kneeling, one standing. Six cupids ^ s Virgin's feet. All the fi gures except the saint rest on clouds, " is an ecstatic visión experienced by the saint, as has been jximt ou Hall-van den Bísenos. The polychromed and partially gilt terracota g P on a gilt and gessoed wood base, which is almost certainly origin * ® , . original colour survives, although parts of the group have been re pain has suffered some damage^^. I am grateful to John Larson for his comments on this. . , . i-„ Rnhin<u)n from " Victoria and Alber. Musenm, inv. no. 250-1864. Bough. W Soriano of Madrid in 1864 for £ 5 1 Is. 7d. (530 reals). Ht. 51 cm (w.A base); 36 cm » base); W. 66 cm. (with base); 54.5 cm. (without base); Depth 41.5 cm. (with base), 31.5 HALL-VAN DEN ELSEN, C^atherine: La Vida y las Obras de Luisa Roldán. (unpublished M. A. thesis). La Trobe University, Bundoora, Australia, p. 100. . ,j k c ni<»on The group may have been broken in two at the front at one tiine; the cross held by b. u g has been broken and repaired. A photograph taken in 1925 (museum negative no. 55171) shows cross, and areas which were subsequently re-painted, which include parts of the Virgin s robe, . habit, and the floor area in front of the group. Other areas of damage are the miMing wings o e ing ángel, and the broken left wing of the kneeling one. One foot of the Christ Child is missing, other is chipped; the Virgin's head is cracked at the back. 328 MARJORIE TRUSTED Although at one time the iconography was uncertain, the identification of the saint as S. Diego of Alcalá must be correct^'. He was a Franciscan friar (c. 1400-1463), one of whose miracles occurred when he was accused by a fellow friar of stealing bread from the convent to give to the poor. When his robes were searched, the hidden bread was miraculously transformed into roses Rose petáis can be seen in the folds of the Franciscan habit of the saint represented here. Beatrice Gilman Proske was the first to attribute the group to Luisa Roldán (c. 1656-1704), when she catalogued the works by that artist in the collection of the Hispanic Society in New York^^. In later articles she stated that it was likely to be a workshop product, as its quality was ostensibly lower than autograph works"*®. Sánchez-Mesa Martín and other have supported the attribution to Luisa Roldán"*'. Comparisons with signed works by the artist reveal similar characteristics. The composition of the group in the Victoria and Albert Museum is closely comparable with the signed Mystic Marriage in the collection of the Hispanic Society: in particular, the relation of the Virgin and Child to the kneeling saint (S. Diego of Alcalá and S. Catherine respectively), the positions of the ángel holding the cross in the London group and the S. Joseph in the New York group, and the face and pose of the angels on the left of each group. Despite the view of Gilman Proske, the quality of the present piece does not suggest that it is by a different hand from pieces known to be autograph, although arguably a degree of workshop collaboration may have occurred in all Roldán's output, including signed works. In her unpublished MA thesis, Catherine Hall-van den Elsen states that she believes the present work is autograph, and notes that comparisons with signed works (notably the New York Mystic Marriage) confirm this"*-. Many of Luisa Roldán's small terracottas were made during the Madrid " The saint was thought at different times to represent St. Francis (when first acquired by the Museum in 1864), and later St. Anthony of Padua. John Charles Robinson believed the scene depicted St. Francis at Portiuncula. According to this legend, the saint roiled himself in thorns, and the blood shed was transformed into roses (Victoria and Albert Museum records); other versions of this story are iisted in L. RÉAU, Iconographie de l'Ari Chréíien, Paris, 1958, III, pp. 530-531. See RIANO, J. F.: Catalogue of Art Objects of Spanish Production in the South Kensington Museum. London, 1872, p. 1; GILMAN, Beatrice: Catalogue ofSculpture (Sixteenth to Eighteenth Centuries) in the Collection of the Hispanic Society of América, New York, 1930, p. 294, note 2; GOMEZ-MORENO, María Elena: Ars Hispaniae. XVI. Escultura del Siglo XVII. Madrid, 1963, p. 311; SANCHEZ CANTON, F. J.: Ars Hispaniae. XVII, Escultura del Siglo XVIII, Madrid, 1965, p. 46. REAU, L.: Iconographie de L'Art Chrétien, III, Paris, 1958, p. 385. GILMAN, Hispanic Society, op. cit. , loe. cit. ^ GILMAN PROSKE, Beatrice: «Luisa Roldán at Madrid» (parts I-III), The Connoisseur, CLV, nos. 624-626, 1964, pp. 128-132, 199-203, 269-273. The present piece is mentioned in Part III on p. 269, and illustrated in fig. 19. ■*' SANCHEZ-MESA MARTIN, Domingo: «Nuevas Obras de Luisa Roldán y José Risueño en Londres y Granada», Archivo Español de Arte, XL. n.° 160, pp. 325-331; HALL-VAN DEN ELSEN. op. cit. , pp. 98ff. HALL-VAN DEN ELSEN, op. cit. p. 98: «Entre las obras en barro atribuidas a Luisa Roldán y no fi rmadas, la más conforme al estilo de la escultora es la de la colección del Victoria and Albert Museum...» She later states, «A pesar de la falta de una fi rma... ésta me parece atribuible a Luisa. La comparación con otras obras fi rmadas por Luisa (sobre todo Los desposorios místicos de santa Catalina de c. 1691, Hispanic Society, Nueva York)... lo confirmará». Ibid., p. 100. 329 THREE SPANISH TERRACOTTAS period^-\ The present work is undocumented and undated; however, Roldán had formed a connection with the Franciscan convent dedicated to San Diego of A ca a, at Alcalá de Henares, from 1697 to 1701, when she was working on a coi^ission for Carlos II. This was a figure of Christ carrying the Cross (an ["^8^ vestir) which was intended for the convent. In fact, the death of Carlos II in 17 prevent the sculpture from reaching its destination, and it was eventually e con vent of Clarisas de Sisante at Cuenca, after the death of Roldán . s . i^o of Alcalá is rarely represented in sculpture, it seems reasonable to assume t a Roldán's initial commission for the convent may have led to her eing as e o carry out a further small-scale work, showing their patrón saint. I so, e presen piece may date from about 1697 to 1701. Catherine Hall-van den sen a so sug gested a possible connection of this group with the convent at A ca a e enare in her thesis''^, but has proposed that stylistically the group accor s more c ose y with works probably executed around 1691-2 (the New Yotk Mystic dated) and the Repose on the Flight into Egypt dated 1691 in the collection of Condesa Ruiseñada at San Sebastián)''^. She points out the simi arity o compo tion of the complex groups, and the details of the bent toes of ® the New York group and the angels to the left in the present one an t e uisen groups The present group ought therefore to be dated c.1691 c. • In the publications to date on Luisa Roldán s oeuvre, no iscussion given as to whether she modelled or cast her works. It is assume at a racottas are modelled"*®. Nevertheless, many of her works bear stri ng y si features, in terms of compositions, and individual fi gures; moreover, wi m same works repetitions of certain motifs can be seen. Casting c ear y aci i reproduction, and may have been a technique used in some of pieces. The similarities between the group in the Victoria and Albert Mystic Marriage in the Hispanic Society remarked on above may e ue simp to the homogeneity of Luisa Roldán's style, rather than because she employe moulds. However, in order to test the theory that moulds might have een use the overall dimensions, and certain details of each piece were measure or com See CEAN BERMUDEZ, J. A.: Diccionario Histórico de los más ilustres Profesores de las Bellas Artes en España. Madrid, 1800, IV. pp. 235-239; GILMAN MARTIN GONZALEZ, Juan José: Escultura Barrocca en España, 1600-1700. Madn , 177-184. . ^ PALOMINO, Antonio: El Parnaso Español [I7I4] in SANCHEZ CANTON, op. ^ 362, cited in GILMAN PROSKE: Connoisseur. op. cit.. pp. 621-2 and notes 27-8. See aiso M GONZALEZ, Escultura Barroca, op. cit.. p. 180. HALL-VAN DEN ELSEN, MA thesis, op. cit., p. 99. For the Ruiseñada group, see GILMAN PROSKE: Connoisseur, op. cit. , fi g. Personal communication, 1990. *,iadtit>j See GILMAN PROSKE: Connoisseur. op. cit. above and SANCHEZ-MESA MARTIN, op. cit. above. GILMAN PROSKE {Connoisseur. op. cit.. Part III, note 53) records that the three pieces by Luisa Roldán in the collection of the Hispanic Society «were examined by optical —emission spectrography an X— ray diffraction, in order to find out whether there were differences in the materia used, with the possibility that wide differences might indicate more than one workshop». The results indicated that almost certainly oniy one workshop was active. Gilman Proske does not however comment on an investigation of the possibility of moulds being used.