The decoration of the lower zone with the volutes arranged on a single line at the bottom and not on two registers, brings the base closer not to the specimens believed to have been found at Hadrian's Villa, but rather to those in the Louvre and Venice (inv. 114), which are also united by the presence of a small base line on which the three cupids rest. Just as Orietta Vasori and Hans-Ulrich Cain have done before, the various replicas can thus be distinguished into two groups: group A (Uffizi, British Museum, Vatican, Venice inv. 104) and group B (Louvre, Museo Nazionale Romano, Venice inv. 114, Mantua) (fig. 6). [13]

Having tracked down this specimen as well, having completed the series of currently known bases with these figurations, all coming from historical collections, and thus being able to distinguish on the originals the details of individual exemplars may perhaps help in solving a kind of rebus, which concerns the iconographic tradition of these marbles. While recent images of our base seem to be particularly rare, there have been numerous reproductions of this series of objects in early modern times.

In fact, several drawings are known from the mid-sixteenth century onward depicting our ‘armed’ cupids. However, since these bases are serial products, i.e. objects that are very similar to each other, it is not always possible to tell what object exactly is portrayed in the drawings. This is a constant problem when dealing with this kind of production. Apart from sculptures like as these, similar cases occur for objects of small format­ – such as small bronzes, terracottas etc.–, replicated in large quantities (just as coins, fibulae, oil lamps). Often in such cases one has many replicas of objects, hardly all of which are known today, and a relatively small number of drawings. Our’s is a particular, yet not unique example of a complex and rich iconographic tradition involving a fairly small and controlled corpus of antiquities. Still, there is no shortage of difficulties, as the related records in the Census of Antique Works of Art and Architecture Known in the Renaissance warn.

Nevertheless, there are a few fixed points (fig. 7). Three anonymous eighteenth-century drawings at Eton College reproduce one of the two Venetian 'tripods,' inv. 104. The most 'finished' and detailed drawing reproduces the base in its entirety, revealing that a possibly non-antique bust of Emperor Hadrian was placed on it; the other two are rush-made records of the other sides. [14] The specimen now in the Uffizi, but until 1756 at the Villa Medici in Rome, is explicitly mentioned instead in the drawings by Carlo Calderi (also at Eton, c. 1710–1730). [15] The Medici base is also referred to in the four drawings by Pietro Testa in the Museo Cartaceo of Cassiano dal Pozzo, executed in the first half of the 17th century (again, with varying degrees of quality). [16] In discussing the graphic tradition of the Florentine base, Amanda Claridge and Eloisa Dodero also recognize one side of it in the seventeenth-century drawing by Charles Le Brun and in one, about a century earlier, by Girolamo da Carpi. [17] All these drawings and specimens belong therefore to group A.

An anonymous drawing in the Uffizi, also dated to the mid-sixteenth century and published by Orietta Vasori in 1978, was instead referred to the Venetian base inv. 114, from group B. [18] According to the above criteria, the following drawings known to date can now also be assigned to the same group B (fig. 8):

• Codex Escurialensis. [19]
• a sheet at the Uffizi (2nd quarter of the 16th century, attributed to Baldassarre Peruzzi). [20]
• an anonymous drawing in Berlin (mid-16th century), in which the cupid with helmet is reproduced, without the small base line; the inscription Pellestrina di marmore indicates a provenance not recognized at present for any of the known bases. [21]
• Codex Destailleur B in St. Petersburg, shown in the Census as a parallel copy of the drawing in Berlin, depicting only the profile of the base, with the ram’s head and the sphinx, but with no cupid; an anonymous copy of this drawing is in the Padua library. [22]
• Codex Coburgensis (1550–1555), with the sphinx’s heads missing as in the base of the Kircheriano. [23]
• Codex Pighianus (1550–1555), copied after the Codex Coburgensis. The annotation, only preserved in the Pighianus, locates the piece in the Della Valle collection. But, as Claridge and Dodero have already excluded both drawings as documents of the base at the Uffizi because of the differing decorations, [24] this piece cannot have passed through Villa Medici to Florence, as so many other marbles have.
• a drawing by Etienne Du Perac in Paris (mid-16th century), also already excluded by Claridge and Dodero in relation to the base at the Uffizi. [25]

More difficult, however, to place the drawing by Amico Aspertini (c. 1540): it depicts only the cupid with the sword as an independent figure, while the base is not shown. [26]

In the text referring to his plate no. 1, Bonanni says that Kircher’s candelabrum base had been in the Cenci brothers’ collection for many, many years: perhaps it was already known in the 1500s or 1600s? Or is Bonanni’s engraving really destined to remain the only image of the candelabrum base before the 21st century, along with the one that Bernard de Montfaucon drew from the Musaeum Kircherianum in 1719? [27] New studies will tell. In the meantime, research around these particular cupids, which are lost, found, and mingled among others, has initiated a collaboration between Antiquitatum Thesaurus and the Museo Nazionale Romano. The aim is to reconstruct, using Bonanni’s own catalog as a guide, that collection that the Jesuit father had so painstakingly rescued, and which had so much importance in the process that led to the establishment of the museums of ‘Roma Capitale’.