Materials & Techniques
A molten metal or alloy of a lower melting point than the metal used for a piece of jewelry was run between sections to solder them. Hard soldering, in which the melting point of the solder is just below that of the material being joined, took place much as it does today. (Soft soldering, in which the melting point is much lower, produces a far weaker join.) Small pieces of a solder, having perhaps six parts gold and one part silver, with a melting temperature of 970 degrees centigrade, were placed between sections and heated to the desired temperature. The heat was intensified by a bellow or blowpipe. Because all the metals besides pure gold oxidized when heated, a flux was needed to prevent oxidation and to enable the solder to flow evenly. It is thought that perhaps borax or even sea salt was used as a flux in ancient times. The flux residue was removed after the soldering was completed.
Ancient goldsmiths also used colloid hard soldering, which utilizes the phenomenon of copper and gold together having a lower melting temperature than either metal has separately. Copper salts were ground and mixed with a glue to form a paste. The paste was used to join sections of jewelry and to adhere wire for filigree work and small beads for granulation. The copper salts melted and joined at about 900 degrees centigrade, but the pure gold elements would not melt below 1063 degrees. This difference in melting points provided a margin of safety before the goldwork would melt and fuse.
Barbara Deppert-Lippitz, with contributions from Anne R. Bromberg and John Dennis, Ancient Gold Jewelry at the Dallas Museum of Art (Dallas: Dallas Museum of Art in association with the University of Washington Press, 1996), 23-24.