Coulomb's Data on Harpsichord Wire

Abstract

Charles-Augustin Coulomb published data on the size, strength, and stiffness of harpsichord wire available in Paris in the late 18th century which has not been cited in the literature on musical instruments despite this century's revival of baroque music and the instruments upon which it was played. During an investigation of the internal friction of several antique music wires by means of a torsion pendulum similar to the one invented by Coulomb and first described in the 1784 mémoire (Figure 1), it occurred to us to ask what Coulomb had used as the suspensions in his pendulum. His biographer, C.S. Gillmor, mentioned “harpsichord strings”, and in Coulomb's 1784 Torsion mémoire we found: J'ai pris trois fils de clavcin, tels qu'on les trouve répandus dans le commerce, roulés sur des bobines, et numerotés (“I used three harpsicord wires, such as one finds distributed in commerce, wound up on spools, and numbered”).The spools of wire recently discovered in the stand drawer of a harpsichord in France fit Coulomb's description. The instrument was built in Paris in 1732 by Antoine Vater, and the wire found with it was wound up on small wooden spools similar to spools for thread. The spool of red brass wire had a mark punched in one end and the spool of iron wire had the gauge size (No 5) written on it in ink, while the spool of yellow brass wire had neither mark nor number. Samples of the wires from these three spools, generally agreed to be of 18th century manufacture, were collected by J. Scott Odell of the Smithsonian Institution as part of a research project which he initiated into the physical properties of antique wire. With these samples in hand it was possible to measure certain of their properties and to compare these properties with those reported in Coulomb's Torsion mémoire in 1784. Since Coulomb's wire was undoubtedly of 18th century manufacture (a degree of confidence that we do not always have with collected samples), and the data were taken not long after the wire was drawn, comparison with the Vater wire allows us to detect if, during two centuries, significant changes occur in highly drawn wire.