The Madonna del Latte at the Pushkin Museum in Moscow: an inquiry about a terracotta relief from the fifteenth-century Lombardia

In this research paper we will investigate the fortune that the series of terracotta reliefs depicting the Madonna del Latte and Child surrounded by angels has had throughout the centuries, focusing on all the copies known today and analysing in depth the Madonna kept from 1945 on in the Pushkin Museum collection (Moscow, Russia) and coming from the Bode Museum (Berlin, Germany). Lastly, we will speculate on the possible prototype of the whole series and suggest its creator[1].

         The story of the Madonna begins in the first half of the fifteenth-century in Cremona. Thanks to the ease of reproduction guaranteed by the medium itself and the use of the a calco technique a lot of variations could be found in Lombardia and today we can count sixteen of them, most of these still being in Cremona. There in the Museo Civico Ala Ponzone persist four copies of the relief[2], while other three Madonnas are inserted in the façades of as many buildings[3]. Nowadays Lombardia retains three more replicas: the first being in Milano (Museo d’Arte Antica del Castello Sforzesco)[4], the second in Viadana (Museo Parazzi)[5] and the third in Vigevano (Museo della Confraternita della Morte, Chiesa di Santa Maria della Neve)[6]. The only known copy still in Italy but kept outside the Lombardia region is in Torino and belongs to a private collector[7]. Four replicas are preserved outside Italian borders: Germany (Bamberg, Franke Kunsthandel)[8], Hungary (Budapest, Szépművészeti Múzeum)[9], France (Marseille, Musée des Beaux-Arts)[10] and Russia (Moscow, Pushkin Museum)[11]. During my research I have encountered one more Madonna relief which, due to its importance, will be investigated in a separate chapter.

         As we have just seen the number of different reliefs known today is fairly large and, even though they share a similar composition, their features can give us a great amount of information. Size is the most important clue in order to date any terracotta piece made with the a calco technique: every copy is obtained by impressing the prototype in a piece of clay, which, after being exposed to fire, will be inevitably smaller than the matrix. Thanks to this procedural detail we can assume that some copies were made before others, such as the Madonna (Inv. G, 1, n.5) preserved in the Museo Civico Ala Ponzone (Cremona), which measures 34 x 28 cm, or the ones from Torino, Bamberg and Milano[12]. Dating is possible also by looking at the artistic quality of the piece itself, as the more the copy is chronologically far from the prototype the more details are lost, resulting in a poor composition: the most evident proof of this are three pieces (Inv. G, n.6; Inv. G, n. 288; Inv. G, n. 367) from the Museo Civico Ala Ponzone (Cremona), which, even though artistically not relevant, testify the great fortune that this relief has had throughout the centuries. One last, but not less interesting, element that some works share is the lack of the drapery in the lower left corner. This detail, in my opinion, could point towards a common matrix that can be identified with the Madonna (Inv. G, 1, n.5) in the Museo Civico Ala Ponzone (Cremona) collection: that piece is indeed one of the most antique within the series and it lacks the lower left corner, therefore any copy obtained from it would share the same feature. Three reliefs have this in common and are the ones from Bamberg, Marseille and Viadana. The first and the third share also another feature, which involves the Madonnas from the Museo Civico Ala Ponzone (Inv. G, n. 6) and Torino too: color. Even though for the Cremonese piece we must rely on Alfredo Puerari’s words only[13] – since color photographs of the relief are unavailable and the piece itself is inaccessible – we can admire the rest with our own eyes. Torino and Bamberg’s Madonnas are defined by a naturalistic range of colors, very beautifully preserved, while the Viadana relief in its way is unique because of the invetriatura process which, during the baking, fixed the pigments to the support[14].

         The Madonna del Latte and Child which carries the most troubled history is, by far, the one that belongs to the Pushkin Museum collection. Its traces start in 1894, when an anonymous donor gifted the relief to the Prussian State Collections as stated in the two catalogues written by Frida Schottmüller in 1913[15] and 1933[16], where we can find the terracotta bearing the inventory number 2221. The scholar, while describing the piece, gives us useful information about its original appearance: she defines it as black, with a modern frame and with the lower left corner restored. It is therefore clear that the rupture must have occurred before the year 1913 (img. 1). Very little photographic evidence of how the Kaiser-Friedrich-Museum used to exhibit its collections remains but, among those few surviving shots, I was able to trace the original location of the Madonna as displayed in the Room n. 29 in 1909 (img. 2): this space was reserved to Tuscan masters and created by Wilhelm von Bode itself with the intent of establish a dialogue between sculpture and picture. We must keep in mind that until the year 1913 the relief, due to its Early Renaissance features, was believed to come from Tuscany and this misunderstanding was cleared only thanks to Frida Schottmüller[17]: in fact we can see how her theory affected the exhibition while looking at a photograph taken in a period of time that spans from the year 1909 to the year 1926, where the Madonna has been replaced by a work of the Tuscan painter Benedetto di Bindo.

There is no further documentary evidence related to the terracotta piece after the year 1933 and, because of the great fire that in 1945 hit the deposit in Berlin where the relief was kept during the Second World War, it was presumed to be lost for good[18]: luckily it was not (img. 3). Today we are aware that several damaged works of art were moved from Germany to Moscow in 1946 and stored in the USSR Ministry of Culture's Archive of Art Treasures in Zagorsk before being moved to the Pushkin State Museum in 2003, where the museum’s conservators began to work on them almost immediately. This endeavour eventually led to the creation of the Donatello Project[19], thanks to the joint effort of the Pushkin Museum, The Prussian Cultural Heritage Foundation and The Collection of Sculptures from the State Museums of Berlin. The Madonna del Latte, which survived along with other fine pieces of sculpture, is currently undergoing a strict conservation process that will allow, once completed, to put back in display this terracotta to the public of the Pushkin Museum.

         Amongst the sixteen different versions of the Madonna relief known to this date I have been able, with the help of Aldo Galli, to find the one that - in my opinion – could be considered as the prototype (img. 4). Known only thanks to two photographs, this relief is characterized by an exquisite attention to the detail which stands out in comparison to any of the other copies and suggests the direct work of the master: we can see it clearly by looking at the quality of the draperies or the angels’ wings. The frame itself is peculiar, being a wooden tabernacle with both doors painted, portraying Saint Bernardino of Siena on the left and Saint Jerome on the right. One of the two photographs belongs to the Museo di Palazzo Davanzati (Firenze) and, during my research, I was able to discover that it was shot in the year 1910 for the drafting of Elia Volpi collection’s auction catalogue. The sale took place on 25th April 1910 at the Villino Volpi and the works of art came from the Palazzo Davanzati, an historic building located in Firenze that Volpi himself had bought six years earlier[20]. During the restoration of the palace several pieces were discarded and sold to the best bidder: thanks to a noted copy of the catalogue that I got from the Fondazione Zeri (Bologna) we know that it was sold for 700 lire, but the buyer is still unknown, as is the Madonna’s fate.

         Uncertainty seems to be the leitmotiv of our dissertation and proposing a possible attribution of the prototype makes no exception, even though some things can be said with a certain degree of confidence. The style of the paintings clearly takes us to Cremona since it’s very close to the Bembo maniera, a family of painters originally from there, while the presence of Saint Bernardino of Siena suggests that the work should be collocated shortly after the year of his death, occurred in 1444.

We are in possession of documentary evidence that certifies as in the timespan between 1443 and 1448 the most prominent artist in Cremona, specialized in terracotta figures and until before resident in Parma, was Giovanni da Roma. His most important work – undertaken before 21st October 1444 and after 25th October 1441, left unfinished – was a great tabula made for the altar dedicated to the Saints Crisante and Daria in Sant’Agostino church, inside which was located a large series of terracotta statues in full relief. The altar itself was under the patronage of Francesco Sforza and Bianca Maria Visconti and meant for celebrating the day of their marriage: 25th October 1441. This prestigious task was left undone by Giovanni because of his return to Parma and on 2nd April 1450 Rinaldo de Staulis took his place[21], completing the tabula in 1468. Another important assignment came from the friars of Sant’Antonio di Vienne, who envisioned a multitude of figures to be placed in the façade of Sant’Antonio church: Giovanni da Roma was already working on them on 14th March 1443 and by 9th April 1445 the work was completed, along with some additional statues ordered by the Cremonese nobildonna Giovanna Cavalcabò. The last archival testimony of the artist is a letter that he wrote on 8th February 1448 to the current rulers of Parma in order to ask the permission to return there, but with tax benefits: as they were granted we can assume that short afterwards Giovanni and his family actually shifted back from Cremona to Parma[22]. Unfortunately, none of his works has survived: this absence must be linked to the frailness of the material itself but not only, since several terracotta pieces were destroyed just for the sake of the taste of the time which, both aesthetically and liturgically, refused them[23].

The hybrid style of the alleged prototype, which combines Late Gothic elements with Early Renaissance details, suggests that the author must have possessed an updated figurative culture and, therefore, not being originally from Lombardia: taking for granted a passage in Tuscany[24] while going from Rome to Parma (where originally Giovanni da Roma was located, before moving to Cremona in order to avoid taxes and fees), all the clues points towards the proposal of Giovanni as the creator of the composition[25], which – due to the increase of devotional practices – may subsequently be serialized by Rinaldo de Stauli, the first industrial terracotta maker of the region. This ease of reproduction eventually led to the proliferation of the relief itself during the following decades, of which we have testimony still today.

The next steps, in order to proceed towards further developments in this field, will inevitably point at two directions. The first, closely related to the Madonnas itself, leads to the discovery of the alleged prototype’s current location, since only a first-hand study could confirm the veracity of the theories elaborated in this paper. The second aims to continue the deepening in Giovanni da Roma’s documentary evidence, which - as far as we know - is split between Cremona and Parma but, in my opinion, could be found in another location too: Milano, the capital of the former duchy.

The goal of these insights, in conclusion, will be the confirmation or the confutation of the attribution proposed by Aldo Galli and me, to shed light on the author of such a fortunate composition whom, for now, is still without a name.




Ferrazza Roberta, Palazzo Davanzati e le collezioni di Elia Volpi, Cassa di risparmio di Firenze, Firenze, 1993


Galli Aldo, Michele Da Firenze: i Problemi Dell'attività Giovanile. In “Prospettiva”, no. 68, 1992


Galli Aldo, Pittura e scultura a Parma, 1400-1520. Orientamenti, occasioni, resistenze, In “Quintavalle, A.C. (edited by), Storia di Parma. VIII/1. La storia dell'arte: secoli XI-XV”, MUP, Parma, 2019


  1. Lambacher, et al., Staatliche Museen zu Berlin. Dokumentation der Verluste. Skulpturensammlung, VII. Skulpturen, Möbel, Berlin, 2006


Musée des beaux-arts (Marseille), Catalogue des peintures, sculptures, pastels et dessins: illustré de 161 reproductions photographiques, Barlatier (Marseille), 1908


Palvarini Gobio Casali Maria Rosa, Ceramiche d'arte e devozione popolare in territorio mantovano, Publi Paolini, Mantova, 2000


Peregriny Janos – Pulszky Károly, A Szépművészeti Múzeum részére vásárolt festmények, plastikai művek és graphicai lapok lajstroma a vételárak kitüntetésével, 1896


Puerari Alfredo, Museo civico “Ala Ponzone”, Cremona. Raccolte artistiche, Cremona, 1976


Schottmüller Frida, Königliche Museen zu Berlin, Beschreibung der Bildwerke der christlichen Epochen, Die italienischen und spanischen Bildwerke der Renaissance und des Barocks, In Marmor, Ton, Holz und Stuck, Berlino, 1913


Schottmüller Frida, Bildwerke des Kaiser Friedrich Museums, Die italienischen und spanischen Bildwerke der Renaissance und des Barock, Die Bildwerke in Stein, Holz, Ton und Wachs, Berlino, 1933


Trombini Kendra, Stampi per terrecotte conservati nel Museo “Ala Ponzone” di Cremona, in “Terrecotte nel Ducato di Milano. Artisti e cantieri del primo Rinascimento”, Edizioni ET, Milano, 2013


Venturi Adolfo, Storia dell’arte italiana, Volume VI, Hoepli editore, Milano, 1908







Img. 1: Madonna del Latte and Child. Pre-war picture. (Kaiser-Friedrich-Museum, Berlin).

Img. 2: The original exhibit of the relief in 1909, detail. (Kaiser-Friedrich-Museum, Berlin).

Img. 3: Madonna del Latte and Child. Post-war picture (Pushkin Museum, Moscow).

Img. 4: Tabernacle containing the Madonna del Latte and Child with the pictures of Saint Bernardino of Siena and Saint Jerome. (Location unknown).


Photographic references


Img. 1: Donatello Project (Pushkin Museum, Moscow).

Img. 2: Zentralarchiv (Staatliche Museen zu Berlin, Berlin).

Img. 3: Donatello Project (Pushkin Museum, Moscow).

Img. 4: Museo di Palazzo Davanzati (Firenze).


[1] The main source of this paper is my master’s degree thesis in Art history (discussed on 12th February 2020 at the Università degli Studi di Trento), titled: La Madonna del Latte nel Museo Pushkin di Mosca. Indagine su una scultura in terracotta a stampo nella Lombardia del Quattrocento. Rapporteur: Aldo Galli, co-rapporteur: Luca Siracusano.

[2] Inventory numbers: Inv. G, 1, n. 5. Inv. G, 1, n. 6. Inv. G, 1, n. 288. Inv. G, 1, n. 367. See: Puerari 1976, pp. 32-33.

[3] Trombini 2013, p. 355, n. 41.

[4] Venturi 1908, Volume 6, p. 116.

[5] Palvarini 2000, p. 142.

[6] As kindly reported to me by Aldo Galli.

[7] I thank Gabriele Fattorini for the insight.

[8] My appreciation for the suggestion goes to Stefanie Paulmichl.

[9] Peregriny - Pulszky 1896, p. 22.

[10] Musée des beaux-arts (Marseille) 1908, p. 533.

[11] See: Donatello Project, Madonna col Bambino.

[12] The dimensions of Pushkin’s Madonna del Latte are 26,7 x 21 cm.

[13] Puerari 1976, p. 32.

[14] My thanks to the Museo Parazzi (Viadana) and Dr. Daniela Benedetti for letting me study this wonderful piece up close.

[15] Schottmüller 1913, p. 129.

[16] Schottmüller 1933, pp. 123-124.

[17] The first known reference to the relief preserved in Berlin was made by Adolfo Venturi, who also proposed Cremona as the origin of the composition. Schottmüller’s proposal therefore is based on the Italian scholar’s findings. See: Venturi 1908, Volume 6, p. 116.

[18] See: Lambacher, et al. 2006, p. 149. I thank Luca Siracusano for showing me this publication.

[19] See: Donatello Project.

[20] Ferrazza 1993, p. 32.

[21] Within the timeframe that spans from 13th August 1447 (death of Filippo Maria Visconti) and 25th March 1450 (resumption of power by the Sforzas) the power in Parma is held by the so-called Repubblica Ambrosiana. Therefore, it is not by chance that the works on the tabula were resumed shortly after the latter date.

[22] These information are all taken directly from the original documents conserved between the Archivio di Stato di Cremona and the Archivio di Stato di Parma. During my research I was even able to establish that Giovanni’s father was named Stefano and to define his neighbourhood in Cremona: the vicinia di San Vittore, which nowadays roughly corresponds to Largo Paolo Sarpi.

[23] Galli 2019, p. 460.

[24] Firenze was the centre of the Italian terracotta rebirth during the fifteenth-century, thanks to the city baptistery’s Porta Nord (1403-1424) building site managed by Lorenzo Ghiberti. Donatello was formed there, as well as Michele da Firenze (1404-1407), who will later export his knowledge throughout Northern Italy. See: Galli 1992, pp. 13-14.

[25] As in Aldo Galli’s insight.