History of Technology in India, Vol. I: From Antiquity to c. 1200 A.D. v1.c14.s1: II
History of Technology in India, Vol. I: From Antiquity to c. 1200 A.D. v1.c14.s1
- Language
- English
II
v1.c14.s1.1
A great number of coins carry devices, which are not as sharply cut or clear as those on a great number of archaic coins including the ‘punch-marked’ series. These devices seem to have been impressed on coins not by hard striking, but by casting metal in moulds bearing the required impression in negative. The process of manufacturing these coins involves preparation of mould(s), melting of metal and casting of the latter in the mould(s). The negative of the intended coin-type could have been engraved or etched on a blank mould made of heat-resisting plastic clay or on a mould in cast metal. It was perhaps technically also possible to have the metal mould with the intended impression by casting the required amount of molten metal in a clay mould bearing the positive of the impression in relief. Similarly, according to an alternative and possibly more convenient method of preparing a clay mould for producing coins, the intended coin-device could have been carved in relief on a lump of clay or on a piece of metal. It was not difficult to rectify the positive impression, if it was found necessary. The latter, made either in clay or metal, served as a model. By pressing this positive impression on a lump of heat-resisting plastic clay, specially prepared according to the shape (round, rectangular, square, etc) and size of the intended coin, a mould was made, consisting of the intagliated negative impression. The mould with raised rim around the coin-socket was then baked.
v1.c14.s1.2
Molten metal, of pre-determined quality was now poured into the mould. Gradually hardened, it would receive the positive impression of the type found in negative on the mould. Some copper coins bearing device only on one side may have been minted following this process.
v1.c14.s1.3
If a coin was destined to carry devices on both faces, a second mould had to be prepared and an opening made through the rim of each of the moulds. The openings as well as the moulds were then jointed face to face. Molten metal could now be poured through the channel and into the cavity thus formed.
v1.c14.s1.4
Some cast coins found still joined together by a bar of metal indicate that these were cast in pairs. Each pair of moulds (bearing obverse and reverse devices), facing each other, were jointed with another pair of moulds by a channel. Molten metal poured probably through a hole in the channel, could flow into each socket. But in the process some portion of the metal could have sometimes dried up in the channel resulting in a bar joining both the coins. Ordinarily the bar should have been broken and detached from coins. But sometimes this was (mistakenly) left attached to the pair of coins.
v1.c14.s1.5
Some of the Yaudheya coin moulds, found at Sunet, consist of pairs of discs bound together back to back (i.e. by their blank faces) with the help of mortar. This evidence suggests that sometimes pairs of discs, each forming a set of moulds for obverse and reverse devices, were placed one above the other so as to make a cylindrical pile.
v1.c14.s1.6
The engraved faces of each pair of discs faced each other, while the blank side of each of them was attached to the outer side of the immediately upper or lower pair (as the case might be). The opening on the rim of each pair (as indicated in the last but one paragraph) was usable for channelling molten metal into the inner side. Available data indicate that the series of discs was then plastered over on the outer side, leaving only the channels leading to the sockets uncovered. Molten metal could now be poured into different moulds separately, or, which was more likely, it could be made to flow down through a vertical hollow column communicating with the channels leading to the coin sockets of the moulds placed vertically. The practicability of the latter method is indicated by the ‘U’- or ‘V’-shaped cut leading right into the socket of some Sanchi and Sunet moulds. The deep cutting through the side of each of these moulds seems to have been designed to allow metal run directly into the socket from some central channel (like the above mentioned vertical hollow column).
v1.c14.s1.7
The method of letting the molten metal flow down the hollow column was less time-consuming than that of pouring through the opening in each pair of discs. In the former case several coins could be made in one single operation. However, it also had greater risk of loosing part of the metal before reaching the socket. Commenting on the above noted moulds from Sunet, B. Sahni has observed that two or more cylindrical piles of moulds have been joined together “round a vertical axis so as to bring all the individual channels into communication with the single axial canal fed through a crater at the top” (Sahni, 1945, p.36). Similar Roman coin moulds have been discovered at Lingwell Gate in England (Sahni, 1945, pp.36 and 52; Pl. VII, Figs. 143-45). It is significant to note that the Sunet moulds of the later Yaudheya coins are to be dated long after the inception of Indo- Roman trade.
v1.c14.s1.8
More interesting are the discoveries of several coin sockets in single coin moulds. The shape of such moulds is either rectangular (as that of an Atranji Khera or Taxila piece) or circular (like that of several Taxila finds) in shape (Sahni, 1945, pl. VI, no. 126, pl. V, no. 123, nos. 119-122). The rectangular coin mould from Atranji Khera displays the negative impressions of Huviṣka’s coin devices in sockets arranged in rows. Only the sockets of the row nearest to the opening on the rim or the mould received metal through channels connecting them with opening. Each of the sockets in succeeding rows received metal through the channels connecting it with the nearest socket of the earlier row and also of the same row. Pairs of such moulds were often used to be fixed face to face with the help of tenons and were probably plastered over. Several Taxila moulds, however, do not betray traces of tenon or mortise. Here “the funnel-like inlet on one side of a mould narrows into a final canal which feeds the nearest coin socket and it is from this socket that all others have to be fed by narrow connecting channels” (Sahni, 1945, pp. 41 and 48).
v1.c14.s1.9
A similar arrangement was noticed by J. Dattari in some Roman moulds discovered in Egypt (Sahni, 1945, pp. 42 and 54). Several copper coins yeilded by a hoard at Masubazar (32 miles south of Purulia in West Bengal), which are joined with one another by metal bars, might have been produced through an analogous method.
v1.c14.s1.10
A complete set of three discs, found at Mathura, have the top and bottom ones blank on their exposed sides. Their inner faces bear coin sockets, while the central discs have coin-impressions on both faces. Traces of luting on the outer side indicate that these were plastered over. A part of the edge of the middle disc was cut away into a ‘V’-shaped notch, with a straight channel on each face leading from the notch to a point just beyond the centre. From this main channel four short branch channels came off at straight angles, two from either side of it. The ends of these four channels and that of the main channel itself communicated with the five coin sockets.
v1.c14.s1.11
To complete the opening at the edge of the triple mould, the two outer discs were excavated on their inner sides into shallow depressions which were either fitted opposite each other (on either side of the notch in the middle disc), or lay obliquely on two sides of the middle disc. Molten metal could be poured through the channels made of the above noted ‘V’ shaped notches (on two sides of the middle discs) and depressions (on the inner sides of the outer discs).
v1.c14.s1.12
The most elaborate process of coin-casting has been revealed by the fragments of early Yaudheya coin moulds found at Khokrakot, near Rohtak. Each disc seems to have a hole in the centre, from which radiating channels run into eight coin sockets arranged in a circle.
v1.c14.s1.13
B. Sahni thinks that at first a clay disc was prepared around a peg on the middle of a base plate enclosed by a hollow cylinder. Then an eight-rayed metal with a hollow centre was slipped down the peg and pressed on the clay. At the end of each of the radiating channel, thus excavated, “a coin model” was pressed to create a coin socket bearing the negative impression of the obverse or reverse concerned.
v1.c14.s1.14
Two discs, prepared in this way and coupled face to face, made one pair. Two such discs, one carrying the obverse and the other reverse impressions, were coupled face to face and fixed to each other with the help of tenons. Several pairs of discs, placed one above the other and marked with grooves cutting the outer face of the rims, formed one set. A number of such sets, dried up in sun, were placed one above the other. These were then plastered over with a paste of clay mixed with unhusked grain and then baked in a furnace. However, the central hole of the top disc and so of all discs were left open.
v1.c14.s1.15
These holes formed a hollow shaft. Molten metal poured down through it, with the help of a funnel-like opening at the top, would run by means of the radiating canals of each pair of discs into the cavities of coin sockets. B. Sahni is of the opinion that casting was done when the pile of moulds was still hot. The pile was probably enclosed in a furnace into which wind was passed with the help of a bellow, and the molten metal was probably poured in the manner described here. When the pile cooled down after the operation, it was broken up and coins were then taken out and cleared.
v1.c14.s1.16
The Rohtak moulds may be compared with ancient Chinese system of casting coins in a circle (T de Lacouperie, 1892, p. xxv), or with a stone mould for casting ten Pao luh kwa coins “arranged like a tree whose stem and branches are the rivulets for the metal and fruits or leaves the coins” (T de Lacouperie, 1892, p. xxviii; Macdonald, 1916, p.61; Sahni, 1945, p.55).
v1.c14.s1.17
A great advantage of the Rohtak system of casting coin was its capability of producing a very large number of coins in one single operation. However, an equally great disadvantage was that the moulds were broken and lost in that process. This drawback might not have been always faced in casting coins in moulds having one socket. At least we have evidence of producing coins in modern times from such ancient Roman moulds (Sahni, 1945, p.52). As part of the metal was likely to be lost in channels and before reaching sockets, it was difficult to control weight of each and every cast coin. This fact explains the divergence in weight of cast coins of apparently one and the same denomination. A part of the metal hardened in the channel leading to a coin socket in a mould which was sometimes left unseparated from the metal hardened in the socket. This distorted the shape of the coin produced from that socket.
v1.c14.s1.18
Again faulty casting could allow formation of air bubbles in the metal, resulting in the fragility of the coins concerned. Moreover, mould-impressions were not always properly transferred (with all their details and sharpness) to coins since such a transfer was not mechanically or manually controlled.⁷