Samnite Belt with Zoomorphic Hooks

1 of 9

Artist
Name	Unknown
Basic Info
Alternative Title	Celtic Belt with Zoomorphic Clasps
Period	Classical period
Created in	Ancient & Byzantine World, Europe, South Italy
Century	4th century BCE
Culture	Italic
Dimensions	Assembled plates: 10.5 x 81.2 cm (4 1/8 x 31 15/16 in.) Clasps: 10.2 cm (4 in.)
Harvard Museum
Department	Department of Ancient and Byzantine Art & Numismatics
Division	Asian and Mediterranean Art
Contact	am_asianmediterranean@harvard.edu

Context

These attachment hooks and plate fragments belong to a Samnite warrior's belt (1). There are at least fourteen belt plate fragments preserved, although it is difficult to count the fragments due to the poor preservation and the way it is sewn onto its canvas backing. It is unlikely that the full length of the plate is preserved (2). The plate fragments are flat, thin metal. One large fragment preserves six raised circular perforations in two rows for the adjustable attachment of hooks. The top and bottom of this fragment and most of the others also preserve borders of smaller perforations along the intact edges, which were used for sewing the belt plate onto a cloth or leather backing. 
 
The two attachment hooks are in the shape of cicadas; the large eyes are concentric circles, and the wings are incised with fine lines and folded down. An iron rivet between the eyes would have been used to attach them to the belt plate; the perforation in the round tail of one of the hooks suggests that another rivet would have been present at the tail of each (3). Out of the head of each cicada springs a decorative band that becomes the stylized head of a dog. The dog's ears are molded, and the eyes are indicated; the long tongue protrudes from the open mouth to form the hook that would fasten to the belt. The bodies of the cicadas are thin metal and hollow on the underside; the neck and head of the dog are solid. The reconstructed placement of the two hooks may be correct: the plate under the bodies of the hooks bears faint outlines of their bodies, and there are some rust discolorations on the plate that correspond to the locations of the extant rivet holes in the hooks.
 
All four hooks in the Harvard collection are the same type and may have been made in the same workshop (4). The hooks may have had additional uses, evidenced by loose examples that have been found in graves and sanctuaries (5). The belts are typically found in male burials, sometimes in pairs (6), and they are depicted on contemporary vase and wall paintings (7).
 
NOTES:
 
1. Complete examples with belt hooks of different forms can be seen in the British Museum, London, inv. nos. 1824,0499.4 and 1867,0508.201; and the Metropolitan Museum of Art, New York, inv. nos. 08.3a  and 1991.171.50. For a general discussion of the object type, see M. Suano, Sabellian-Samnite Bronze Belts in the British Museum, British Museum Occasional Paper 57 (London, 1986).
 
2. See A.-M. Adam, Bronzes étrusques et italiques (Paris, 1984) 120-21, no. 154, for an intact belt plate 95 cm long. Another in the Staatliche Museen zu Berlin, inv. no. Fr. 1028, is 106.2 cm long.
 
3. Of the four Samnite belt hooks in Harvard’s collection, only one has a preserved tail.
 
4. Type 4.a; see Suano 1986 (supra 1) 2. 
 
5. Ibid., 22.
 
6. Ibid., 34. See also G. Heres, “Samnitische Bronzegurtel der Berliner Antikensammlung,” Eirene 17 (1980): 77-88, esp. 78.
 
7. See Heres 1980 (supra 6) pl. 1 (tomb painting from Capua); G. Schneider-Herrmann, The Samnites of the Fourth Century BC as Depicted on Campanian Vases and in Other Sources (London, 1996) 18-20, pls. 17-26 and 46-47; and R. Benassai, La pittura dei Campani e dei Sanniti (Rome, 2001) 182, 188, and 200-205, figs. 197 and 212-21.
 
Lisa M. Anderson

TechnicalDetails

Chemical Composition

Technical Observations

The patina on the components of both Samnite belts is green with areas of red. Iron corrosion products are present. The clasps of  2012.1.107.A-C may have been cleaned by scraping. There is some white glue on the inside of one clasp. 1995.1138 is assembled from many fragments and currently sewn onto a cloth backing; one of the clasps is loose.
 
The belt clasps of 1995.1138 and 2012.1.107.A-C have similar dimensions and styles. It seems likely that they were made in the same workshop. Neither set exactly matches, and none of the four clasps seems to have come from the same mold as any of the others. For example, there are small differences in the length of the ears and the cast groove above the ears. The clasps may have been made as single pieces and matched later; the two clasps of 1995.1138 match each other more closely than they do those of 2012.1.107.A-C, while the clasps of 2012.1.107.A-C match each other more closely than they do those of 1995.1138. The design on the clasp heads of 2012.1.107.A-C is slightly different. 
 
The plates also have slight differences. 2012.1.107.A-C has fewer preserved perforations on the edge, but those that are preserved have a similar frequency to the other plate. The small perforations have flanges or burs indicating the use of a punch. The holes were possibly started by hand drilling using a bow drill and then punched in order not to deform the surface. The plates have different widths and slightly different thicknesses (1995.1138 is 1 mm thick, while 2012.1.107.A-C is c. 0.8 mm thick). There are no visible hammer marks on either plate under raking light, but the plates must have been formed by hammering using a broad hammer. The repoussé crescents around the holes in the plates would probably not have added any strength. They may have served as guides to help align the clasp hooks with the holes. They may also have been decorative, perhaps mimicking an earlier material (like leather).  They could have been made to accommodate the shape of the fabric or leather lining.
 
The clasps were cast as a single piece from the hooks to the eyes; the back of the eyes shows sinkholes that would have followed the shape of a mold. Cold work in the metal was done on the head of each clasp, including some punch work and very fine dotted circles on the head. The same small pointed punch was used to create lines of dots at the cicada’s head. The four circular punches at the base of the dog or dragon’s neck were made with a single circular punch. The body was perhaps cast as a metal rod, flattened into a sheet, hammered into shape using a matrix, then the sides were trimmed and flattened, and finally the surface details were added. The body is thicker near the eyes and clasps than at the bottom, which is another indication of hammering. 
 
The lines of each cicada’s wings were incised with a very sharp point directly on the metal using very short strokes that are visible under a microscope. Viewed from the top, the lines are perfectly straight, and it is likely that a straight edge was used to guide the incising tool. The very bottom of each line curves slightly to meet the bottom edge. This section of each line is cruder in shape and was made using an abrasive or cutting tool, similar perhaps to a file. 
 
The mount holes at the middle of the cicada head and tail have flange deformations on both faces from the bow-drilling process. Iron corrosion products are present from the iron pin used to secure them to the belt.
 
Henry Lie (submitted 2011)

Harvard Art Museums/Arthur M. Sackler Museum, Gift of Nelson Goodman