The Purpose of the Parthian Galvanic Cells: A First-Century A. D. Electric Battery Used for Analgesia

The Parthian galvanic cells (electric batteries) have been known for some fifty years but have never been scientifically published, have heretofore seemed embarassingly out of context, and have received little scientific treatment since their first publication. I have examined the find and some similar, probably parallel, finds and have considered how the device might have been invented and what the electrolyte might have been. The usual hypothesis, which was proposed by Wilhelm König, who first published the find, is that the purpose of the cells was electroplating, but I argue that this is impossible. Rather, I suggest a purpose, more in keeping with the technological and scientific milieu, as a substitute for the use of electric fish as a local analgesic (attested in the first century A.D. and later). Such an application is supported by the modern clinical practice of transcutaneous electrical nerve stimulation.

More than fifty years ago the Austrian painter and director of the Baghdad Museum and Iraq Antiquities Department Wilhelm König reported the discovery of an ancient electric battery.[1] The device consists of an ovoid ceramic jar about 14 cm tall and 8 cm in diameter. In the 3.3 cm opening was seta 9.8 cm long, 2.6 cm diameter tube formed of rolled and soldered copper sheet, affixed by means of asphalt. The tube was closed at the bottom with a copper disk sealed and covered with a 0.3 cm layer of asphalt. Down the axis of the copper tube an iron rod was suspended from the upper asphalt plug; the rod was about 7.5 cm long (see fig. 1). Both copper and iron projected above the upper asphalt plug. The object was found by members of the Iraq Antiquities Department in the course of regular excavations in an undisturbed stratum of a Parthian settlement site (the modern Khujut Rabou'a near Baghdad)[2] in association with magical bowls, and is dated to the first century B.C. or the first century A.D.[3] (below I argue for the latter).

Similar artifacts have been found elsewhere, dating to the Sasanian period, sometimes in association with magical objects. Near the ancient Seleucia-on-Tigris (40 km downstream from Baghdad), such devices, stuffed with papyrus and lacking the internal iron rod, but in close association with iron and bronze needles, were found in what appeared to be a magician's house (see fig. 2).[4] Several pottery jugs have been found at Ctesiphon (fig. 3 shows one), in each of which were cylinders of bronze, sealed at both ends and filled with organic material.[5] König considered the Seleucian and Ctesiphontine finds parallel, but the absence of the internal iron electrode elicits hesitation.

The Parthian device so closely resembles a wet-cell (i.e., galvanic cell with liquid electrolyte) that König assumed it was one. Various features of the construction point in that direction. The asphalt seal indicates the presence of liquid,[6] and almost all available liquids (save vegetable and mineral oils) were acidic. The presence of dissimilar metals in an acid generates a potential difference and is the key feature of a Voltaic pile. The otherwise useless 0.3-cm asphalt layer on the bottom would serve to prevent the possible shorting of the iron rod to the copper bottom (cf. fig. 4). Asphalt is an inert, water-resistant insulator.[7] It is indeed difficult to see what else the device could be.

But the purpose of the device, as well as its origin, has remained an enigma. Konig suggested electroplating, and most commentators have followed him without examining the possibility of such a discovery in the ancient context.[8] Others have doubted the likelihood of such a use or experimented with models and various electrolytes such as vinegar or copper sulfate solution.[9] These points are considered in order: origin, electroplating, and the electrolyte, all of which lead to the consideration of the purpose.

It is difficult, if not impossible, to see how, on any known ancient theory of matter or its effects, the galvanic activity of dissimilar metals in an electrolyte could have been predicted. While that negative statement is difficult to prove, I should prefer to propose an accidental discovery. Had anyone ever used a bronze spoon in an iron bowl (or vice versa) containing vinegar, for example, the tingling produced in the hand or lip touching both bowl and spoon would be noticed. Such an effect might well have long remained a curiosity.

The models built[10] produce about 0.5 V at a few thousandths of an amp (compare a modern flashlight battery which produces 1.5 V at several tenths of an amp, i.e., several hundred times the power). The iron is the negative "pole" or "electrode”. The voltage depends only on the two metals, while the current depends on the electrolyte. The output of the models[11] is insufficient to accomplish much in the way of electroplating.[12] These cells would have a short "shelf life" of a few weeks,[13] as the electrolyte would tend to consume the iron electrode and depolarize the cell (see below). This may explain the absence of the internal iron electrode in the later examples–perhaps they were only emplaced during actual use.

The process of electroplating is sufficiently complex that it was unlikely to be discovered accidentally. A galvanic cell is composed of two dissimilar metals in an electrolyte, and no such cell can generate sufficient voltage to reduce the more electronegative metal (here the iron of the cathode) at the surface of the more electropositive (here the copper of the anode). True electroplating occurs when a voltage is connected across two pieces of metal, i.e., the electrodes are immersed in an electrolyte capable of dissolving the metal of the anode (positive electrode) or which already contains ions of some metal in solution; in either case the dissolved metal ions plate out on the cathode (negative electrode).[14] Thus two cells would be required-the voltage-generating cell and the plating cell.

Electrodeless plating,[15] which can occur when a less noble metal is immersed in a solution of a more noble metal–iron in a copper-ion solution or copper in a gold-ion solution, for example–might have been discovered accidentally, had the ancients had any way of dissolving noble metals. Now silver will dissolve in nitric acid, and gold only in the nitric-hydrochloric acid mixture aqua regia, but the "mineral" acids such as these were not isolated before ca. A.D. 1300.[16] The complex salt-saltpeter-alum solution, which is equivalent to aqua regia, apparently used by the Moche Indians in Peru (first-sixth centuries A.D.) to dissolve gold and silver seems remote.[17] The dissolution of copper in acetic acid is possible and was practiced: the product is ίός or aerugo/aeraca.[18] Moreover, the alchemists report the (electrodeless) deposition of copper on iron[19] from copper sulfate solutions.[20]

But not only is the discovery of electroplating before the development of the modern electrical and chemical theories (by Faraday, A.D. 1833)[21] extremely unlikely on theoretical grounds, there is an insurmountable practical difficulty. As just noted in connection with electrodeless plating, it was not possible to make an aquaeous solution of gold or silver in antiquity. But a solution of the metal ions to be plated is required for electroplating (as noted above). This difficulty seems not to have been appreciated by any of the commentators adopting or questioning the electroplating hypothesis. Electroplating is no more likely than electric telegraphy, for example:[22] the technological context is absent.

Furthermore, there were simple and effective methods of plating gold on silver or copper, or silver on copper, available from Sumerian times (as is shown by various plated finds),[23] One method practiced in antiquity was the etching of baser gold alloys to remove those portions of the surface composed of base metal and so leave a gold-enriched surface-"surface leaching," "surface enrichment," or "pickling"; the cementation process was applied to leach silver from a silver-gold alloy to gild the object.[24] Numismatic evidence suggests that other methods were used: silver foil eutectically fused or soldered to a copper base or dipping the base-metal coin in molten silver chloride (cerargyrite), for example.[25] The method of choice attested in the ancient world (at least from the first century B.C. to the first century A.D.) was mercury-amalgamation plating,[26] attested also at P. Leyden X 26 and 55. A similar process is alloying the noble metal with lead at P. Leyden X 37:[27] in both cases the noble-metal/base-metal alloy was applied to the copper and the object subsequently heated to evaporate the base metal.

Various commentators have suggested various possibilities. Konig imagined gold cyanide solution,[28] which is impossible, or salt water. W. Ley and H. M. Schwalb report a model containing copper sulfate solution-possible (see above),[29] although Schwalb notes that acetic or citric acids would also work and were known. The same pair of possibilities, vinegar and lemon juice, is suggested by E. K. Hornauer, followed by H. Winkler.[30] A. Al-Haik suggests wine or vinegar.[31]

There is another possibility. Already in Sumerian times distillation was practiced in Mesopotamia.[32] Experiments have been conducted with models of ancient stills,[33] which show that it was possible to concentrate acetic acid from vinegar (although it was easier to separate ethanol [grain alcohol] from water). The parallel case of the distillation of alcohol is suggestive: the evidence indicates that it was accomplished by the first century A.D.[34] Note that during this period (first century B.C. – first century A.D.) at least two instances are reported of inflamed wine, presumably brandy (distilled wine), as ordinary wine will not burn.[35] Moreover, both wine and soured wine (vinegar) were considered to be liquids containing the "Hot" principle[36] (in the widely accepted Four-Element theory);[37] if adding heat to one "Hot" liquid made it more efficacious, why not another? R. J. Forbes notes that vinegar was the only strong acid in antiquity,[38] and there is some Assyrian evidence of an especially strong form, i.e., possibly distilled vinegar.[39] Thus I would suggest that besides citric acid and vinegar, distilled vinegar was also possible and even likely.

My own tests showed that a salt solution (i.e., NaCl at ca. 10%) rapidly corroded the iron, depolarizing the cell so that the voltage dropped to ca. 0.4 V within less than one minute; xakxavOov (copper sulfate solution, 10% by volume) produced about 0.45 V for several hours only, until the accumulation of copper on the iron depolarized the cell. Citric acid in the form of freshly squeezed grapefruit juice and acetic acid (glacial acetic acid diluted to 25%, 36%, and 50% by volume) produced 0.49 V, with slow depolarization (all voltages are t 0.03 V).

The suggestion of a medical purpose for this device is not based on a deduction since we lack the evidence, but on an induction, as any attempt to determine the purpose of any archaeological artifact in the absence of clear parallels and literary attestation must be. Such an induction or intuition must remain hypothetical until further confirming evidence is forthcoming. Any such suggestion must take into account the complete context (archaeological, cultural, and technological) of the device and must also account for its electrical parameters.[40]

Parthia, though having its own sources of cultural strength, was surely a buffer state between the Romans and the Indian and Chinese realms.[41] The Roman trade with Parthia passed through Dura-Europus and Palmyra[42] and from there to Seleucia along the Silk Road.[43] Influence passed from Rome to China and back[44] making Seleucia an entrepot for East-West trade in ideas as well as goods.[45] From A.D. 43 the city shows a greater predominance of Oriental influence.[46] Medical influence passed from Rome to China with the drug trade: storax = suho, frankincense = hs0lu, henna = chia-chia, theriaca = tiyehka.[47] That the devices are found in Seleucia and merely 40 km north suggests that their use and purpose might be sought in the marriage of Chinese and Roman ideas on fertile Mesopotamian soil.

Mesopotamian medical practice included a number of elements conducive to the reception of an electrotherapeutic device of this sort. In Akkadian and Babylonian medicine, following the normative Sumerian practice, two "colleges" of physicians were recognized-the asū and the āšipu.[48] The asū was responsible for prescriptions and incantations, which were formulaic and traditional (the wording is unchanged over a millennium),[49] and he was considered a craftsman or technician and was associated with magicians;[50] the devices were found in magical contexts. The āšipu, on the other hand, practiced divination and diagnosis from the patient's symptoms, but not therapy, and gained status over the asū in Late Babylonian times.[51] The Mesopotamian therapy was typically non-invasive,[52] using drugs in preference to surgery: one common drug component was vinegar.[53] Little is known of Parthian medicine, but it likely included most of the traditional elements of Mesopotamian medicine;[54] the Chaldean tradition continued under the Greeks,[55] and the Parthians typically tolerated local rulers and customs.[56]

Little is known of Chinese medicine during the Han period, which is nearly coterminous with the Parthian era,[57] but acupuncture was already a standard practice since at least the Chou Dynasty (i.e., before the Han). Acupuncture is described in the Huang Ti Nei Ching (the Yellow Emperor's Classic of Internal Medicine), variously dated, but no later than the first century A.D.[58] This may explain the fact that bronze and iron "needles" are found with the Seleucian devices.

The critical stimulus was, I believe, provided by the Greco-Roman use of electric fish as an analgesic. In the mid-first century A.D. (ca. A.D. 47-48), Scribonius Largus[59] first records the practice of applying the torpedo ocellata[60] as an analgesic for headache (Compositiones 11) and gout (Compositiones 162). The recipe for gout indicates the method:

Ad utramlibet podagram torpedinem nigram uiuam, cum accesserit dolor, subicere pedibus oportet tantibus in litore non sicco, sed quod alluit mare, donec sentiat torpere pedem totum et tibiam usque ad genua. hoc et in presenti tollit dolorem et in futurum remediat.

For any sort of podagria (foot-gout): when the pain comes on, it is good for one to put a living black torpedo-fish under his feet while standing on a beach (not dry but one on which the sea washes), until he feels that his whole foot and shank are numb just up to the knees. This will both relieve the current pain and alleviate future recurrences.

The numbing effect had long been known.[61] The ability of the electric ray to transmit its discharge was reported by Heron and Pliny in the late first century A.D.[62] Heron even singles out as conductors iron and bronze, just the materials used in the cell and the needles. The transmission of the discharge was probably known earlier.[63]

Electric fish of one species or another are found in the Mediterranean and in the Nile but not in the Persian Gulf or the Tigris-Euphrates system. Is it possible that some Parthian asū began applying the long-known galvanic tingling produced by dissimilar metals in an electrolyte, perhaps with conductive acupuncture needles of bronze and iron, as a substitute for the Greco-Roman ichthyoelectroanalgesia?

Modern medical practice provides an instructive parallel[64] First it must be noted that the current produced in the cell models (ca. 1 milliamp [mA]) is readily detectible on the skin or tongue and especially in cuts or punctures.[65] Since the publication of R. Melzack and P. D. Wall's "Gate Theory of Pain" twenty-five years ago[66] electrically induced analgesia (and anaesthesia) has been subjected to an increasing number of successful clinical tests.[67]

The electrical parameters vary but for (partial) local analgesia are a current of roughly a few milliamps at a voltage of a few volts (and both direct current and alternating current up to 700 Hz are used): such analgesia might well have been produced by a device such as the Parthian cell. Three other effects have been investigated with clinical success in modern times as well: (1) trauma healing and antisepsis,[68] involving currents of 0.2 to 1.0 mA, voltages of 0.8 to 1.4 V, direct current; (2) bone regeneration,[69] involving currents of 0.001 to 1.0 mA, voltages of about 1 V, and direct current (or alternating current up to 1Hz; though the approach is sometimes contraindicated due to concern about induced osteogenic sarcomas);[70] and (3) the still-controversial induced remission of malignant tumors.[71] The Parthian cells may not have been used for any particular one of these purposes: the bone regeneration and tumor regression results cited above involve implanted electrodes-unlikely in first century A.D. Babylon. That the cells generate currents and voltages shown in modern times to be of positive clinical effect tends to confirm the possibility that they could have had a medical purpose in Parthia.[72]

The object is stratigraphically dated to the first century B.C. to the first century A.D.[73] I am unaware of any attempt to apply thermoluminescence to the clay fabric; 1°C dating of the iron may be possible,[74] and X-Ray fluorescence tests of the metals (copper, iron, solder) for characteristic impurities might confirm a dating. Given the suggested medical purposes, it may be possible to narrow the date to the first century A.D.

That the use of papyrus is first attested for Parthia during the same period (Pliny 13.22) might provide a terminus post quem were we certain that the devices always involved, as the Seleucian and Ctesiphontine models seem to involve the use of papyrus. That the distillation of vinegar followed that of alcohol is likely, and that the distillation of alcohol is first attested in late first century B.C. might provide nearly the same terminus were we certain that the electrolyte was distilled vinegar. That Seleucia shows greater Oriental influence after A.D. 43 might provide a similar terminus were we certain that a necessary stimulus came along the Silk Road from the East.

More reliable than these is the fact that the use of living electric fish as an analgesic is first attested during the first century A.D., and any medical application of this device must be later.[75] Hence, the passages cited provide a terminus post quem.

The Parthian galvanic cells seem to have parallels (though crucial details differ) in the Sasanian period and could not have been used for electroplating. This re-opens the question of their application. Various details of the device, the ancient context, and modern medical practice combine to make it possible and even likely that the cells were used as a local electrical analgesic. The inventor was likely a Parthian asū, and the devices were associated with magicians and so never entered Greco-Roman science.

Seeing these electric batteries, we are struck by the foregone opportunities, but in their context, they were merely one not necessarily very effective tool of practical, magic-using physicians on the edges of the Greco-Roman world. In such a context, opportunities such as electroplating did not exist (to say nothing of electric heat, telegraphy, electromagnets, and electric lights). I would prefer to view the devices in light of the ancient outlook and seek an explanation therein (whether as substitute ichthyoelectroanalgesia or otherwise). It is probable that the device later became merely a conjurors' trick and gradually faded from view, just as the magicians of Mesopotamia did.

NOTES:

[1] W. König, "Ein galvanisches Element aus der Partherzeit?," Forschungen and Fortschritte 14 (1938): 8-9, and idem, Im Verlorenen ParadiesNeun Jahre Irak (Munich and Vienna, 1939), pp. 166-68, pl. between pp. 160-61. For biographical information, see Neun Jahre lrak, pp. 3-4. Wilhelm König was a member of the "Warka Expedition" of the Deutsche Orient-Gesellschaft; I have been unable to locate other biographical information.

[2] Idem, "Galvanisches Element," and A. Al-Haik, "The Rabbou'a Galvanic Cell," Sumer 20 (1964): 103-4.

[3] Rolf Stucky, in Eva Strommenger, ed., Der Garten in Eden: 7 Jahrtausende Kunst and Kultur an Euphrat and Tigris (Mainz am Rhein, 1978), p. 211 (no. 183). I am indebted to Max Kunze and Ralf B. Wartke of the Staatliche Museen zu Berlin for this reference and a pleasant hour of discussion.

[4] Leroy Waterman, Preliminary Report upon the Excavations at Tel Umar (Ann Arbor, 1931), vol. 1, pp. 61-62, pl. 12.

[5] E. Kuhnel et al., Die Ausgrabung der zweiten Ktesiphon-Expedition 1931/32 (Berlin, 1933), p. 28, pl. 45.

[6] Bitumen was, in fact, the normal Mesopotamian (cf Strabo 16.1.9, 15) and Parthian seal. See N. C. Debevoise, Parthian Pottery from Seleucia on the Tigris, University of Michigan Studies, Humanistic Series 32 (Ann Arbor, 1934), p. 18, and R. J. Forbes, Studies in Ancient Technology (Leiden, 1964), pp. 74-80, 9095; the Greco-Roman seal was resin (cf. Pliny 14.25). See M. Wheeler, Rome beyond the Imperial Frontiers (London, 1954; reprint New York, 1971), p. 149, and A. Lucas, Ancient Egyptian Materials and Industries, 4th ed., rev. (London, 1962), pp. 19-20.

[7] H. Winkler, "Galvani and Volta nut Wiederentdecker-Eine dringend notwendige Berichtigung," Elektrie 14 (1960): 71-72, here p. 72. I thank Werner Krenkel for this reference.

[8] G. Gamow, The Birth and Death of the Sun (New York, 1940), pp. 33-34, fig. 7; 2d ed. (New York, 1952), pp. 29-30, fig. 7; W. Ley, "The Elements of Khujut Rabu'a and Ctesiphon," Galaxy 9/3 (December 1954): 44-51; [Anonymous], "Batteries B.C.," The Laboratory 25/4 (1956-57): 112-13, reprinted (with very few verbal changes) as H. M. Schwalb, "Electric Batteries of 2,000 Years Ago," Science Digest 41/4 (April 1957): 17-19; [Anonymous], "Kannte man schon vor 2000 Jahren galvanischen Elemente?," Elektro-Welt. Ausgabe B. lndustrielle Elektrotechnik 4 (1959): 176; E. K. Hornauer, "Elektrische Batterien-vor 2000 Jahren," Elektro-Nachrichten 11/1 (1959): 15; and L. Sprague de Camp, The Ancient Engineers (New York, 1960; reprint Cambridge, Mass., 1970), p. 234; 2d ed. (New York, 1974), p. 252.

[9] Winkler, "Wiederentdecker," pp. 71-72; W. Winton, "Baghdad Batteries B.C.," Sumer 18 (1962): 8687, pl. 1; Al-Haik, "Galvanic Cell"; Stucky, Garten in Eden; and [Anonymous] in Science Digest 90/2 (February 1982): 24.

[10] Schwalb, "Electric Batteries"; Winkler, "Wiederentdecker"; and Al-Haik, "Galvanic Cell." My own model (see below) is similar.

[11] Contra Al-Haik, "Galvanic Cell," p. 104.

[12] At 1mA it would take about 1.1 days to plate 1gm of silver and thus would not be economically worthwhile. For Faraday's Law, and the constants used to calculate the time, see W. Blum and G. B. Hogaboom, Principles of Electroplating and Electroforming, 3d ed. (New York, 1949), pp. 39-40, or G. Langbein, Electro-deposition of Metals, 8th ed., trans. and ed. W. T. Brannt (New York, 1920), pp. 62-63.

[13] Al-Haik, "Galvanic Cell," reports eighteen days.

[14] I simplify-if the electrolyte contain metal ions, the anode typically is insoluable, carbon, platinum, or titanium being common, as is usually the case in gold plating. See Blum and Hogaboom, Electroplating, pp.288-306 (gold); Langbein, Elecrrodeposition, pp. 454-69 (silver), 505-19 (gold); F. A. (gold), 358-68 (silver); or A. Watt and Arnold Philip, The Electroplating and Elecrrorefining of Metals, 2d ed. (London and New York, 1911), pp. 17484 (fold), 227-38 (silver).

[15] Again I simplify: in true electrodeless plating the solution contains its own reducing agent and the basis metal acts only as a catalyst, while in "electrochemical replacement" the basis metal acts as both anode and cathode of the cell driving the plating. See Lowenheim, Electroplating, pp. 710-11 (general), 241-42 and 741-42 (gold), and 742 (silver); Blum and Hogaboom, Electroplating, p. 306 (gold).

[16] E. O. von Lippmann, Entstehung and Ausbreitung der Alchemie (Berlin, 1931), vol. 1, s.v. Acetum (p. 3), and s.v. Mineralsäuren (pp. 146-47) and idem in ibid., vol. 3 (Weinheim, 1954), s.v. Säuren (mineralischen) (p. -119); R. J. Forbes, Studies in Ancient Technology (Leiden, 1965), vol. 3, p. 80; J. R. Partington, Origins and Development of Applied Chemistry (London, 1935), p. 486; and R. J. Forbes, "The Evolution of the Still," Proceedings of the Chemical Society (London) (1942): 237-42, here p. 238. On the other hand, though I remain skeptical, Reginald Campbell Thompson, Dictionary of Assyrian Chemistry and Geology (Oxford, 1936), pp. xxxi-vi, argues that the Assyrians perhaps did know of sulfuric acid (H2SO4, oil of vitriol), since they had green vit riol (Fe2[SO4]3, pp. 89-91) and red vitriol (Fe2O3, pp. 97-98), and oil of vitriol can be obtained by distilling green vitriol solution (the residue being red vitriol), but the Assyrian words given by Thompson for red and green vitriol are not related (as they would be if they obtained red vitriol by this process), and Thompson can only offer "with great caution" a hapax legomenon of uncertain interpretation as the Assyrian word for oil of vitriol (pp. 100-104). Even so, neither silver nor gold will dissolve in sulfuric acid.

[17] H. Lechtman, "A Pre-Columbian Technique for Electrochemical Plating of Gold and Silver on Copper Objects," Journal of Metals 31 (1979): 154-60: note that to use the solution to plate, it must be neutralized with baking soda, for example. I am indebted to Michael N. Geselowitz for bringing this reference to my attention.

[18] Theophrastos De lapidibus 57, Vitruvius 7.12.1, Dioskorides Materia medica 5.79, Pliny 34.110-1. K. C. Bailey, The Elder Pliny's Chapters on Chemical Subjects (London, 1932), vol. 2, p. 172. Often used in medicine: cited twenty-four times in Dioskorides Euporista.

[19] M. P. E. Berthelot, Collection des anciens alchimistes grecs (Paris, 1887), vol. 1, pp. 241-42, and Langbein, Electrodeposition, p. 1. (I have been unable to locate the precise reference in Zosimos of Panopolis to which these authors refer.) The effect is rapid: a 10% by weight solution of CUSO4.5H2O in water completely plated the immersed portion of a freshly cleaned iron nail in five seconds. Despite this, the cell will work with a copper sulfate solution, as my tests showed (see below).

[20] χάλκαωθον; cf. Strabo 3.4.15, 16.6.5; Dioskorides Materia medica 5.98-99; Pliny 34.107. Strabo 3.4.15 states that he is following Poseidonios; on the Pliny passage, see Bailey, Elder Pliny, pp. 178-80. Further on χάλκανθον, Dioskorides Materia medica 5.74, 79, and Galen De simplicium medicamentorum temperamentis et facultatibus 1.6 (11.391-92 K⁬uhn) and 1.35 (11.442-43 Kuhn) state that it is hot in nature, while 9.34 (12.238-41 Kiihn) gives the method of production. It was widely used in medicine; see Dioskorides, Euporista 1 and 2 (twenty-three times), Galen, De compositione medicamentorum secundum locos, passim (12.378-13.1020 KUhn); and [Galen] De remediis parabilibus, passim (14.311-581 Kuhn).

[21] Langbein, Electrodeposition, pp. 1-9; Lowenheim, Electroplating, pp. 1-2; and Watt and Philip, Electrorefining, pp. 79-98.

[22] Cf. Aischylos Agamemnon 8-10, 20-21, 26-30, for fire signals, and Aeneas Tacticus 31 and Polybios 10.44-45 on secret messages of other kinds (Julius Africanus, Kestoi, fragments 51-54 are similar). For a magisterial survey, see Hermann Diels, Antike Technik, 3d ed. (Leipzig, 1924; reprint Osnabruck, 1965), pp. 71-90; cf. also Forbes, Studies in Ancient Technology, 2d ed. (Leiden, 1966), vol. 6, pp. 171-80. The point is that there was a simpler alternative, and none of the necessary associated technology for the more complex method existed.

[23] See Partington, Applied Chemistry, pp. 25-26 for Egyptian gold on silver and copper; pp. 75-76 for Egyptian arsenic on copper, Metropolitan Museum inv. 26.9.12-13 (of the Fifth-Sixth Dynasty); on which see C. G. Fink and A. H. Kopp, "Ancient Egyptian Antimony [sic] Plating on Copper Objects," Metropolitan Museum Studies 4 (1933): 163-67, and Cyril S. Smith, "An Examination of the Arsenic-Rich Coating on a Bronze Bull from Horoztepe," in W. J. Young, ed., Application of Science in Examination of Works of Art (Boston, 1973), pp. 96-102, esp. p. 102, n. 5; Partington, pp. 230-31 for Sumerian gold on copper; p. 237 for Assyrian silver on copper; pp. 329, 361-62 for Aegean gold or silver on copperlbronze; p. 377 for Hittite silver on copper; and p. 439 for Phoenician gold on silver or bronze; Forbes, Studies in Ancient Technology, 2d ed. (Leiden, 1971), vol. 8, pp. 142-43; H. Blumner, Die Technik and Technologie der Gewerbe and Kunste bei Griechen and Romern, 2d ed. (Leipzig, 1912; reprint New York, 1979), vol. 4, pp. 133-35, 308-15, 318-20; and R. Higgins, Minoan and Mycenaean Art, 2d ed. (New York, 1981), pl. 131 (Mycenaean tin plating).

[24] See [Demokritos] Physika kai mystika 12 and cf. P. Leyden X 14, 24, and 67=74. The text of [Demokritos] is published in Berthellot, Alchimistes grecs, vol. 1, p. 46; for the Leyden papyrus, see R. Halleux, Les Alchimistes grecs (Paris, 1981). On cementation, see Albert Neuburger, Technical Arts and Sciences of the Ancients, trans. Henry L. Brose (London and New York, 1930; reprint 1969), p. 92 (the German original is Die Technik des Altertums [Leipzig, 1919], [2d ed. 1921], [3d ed. 1922]); F. S. Taylor, "A. Survey of Greek Alchemy," Journal of Hellenic Studies 50 (1930): 109-39, here p. 130; Forbes, Studies, vol. 8, p. 171, 174; and H. Lechtman, "Ancient Methods of Gilding Silver-Examples from the Old and the New Worlds;" in Science and Archaeology (Cambridge, 1971), pp. 2-30. Even in the pre-Columbian New World; see Paul Bergsoe, The Gilding Process and the Metallurgy of Copper and Lead among the Precolumbian Indians, trans. C. F. Reynolds, Ingeniorvidenskabelige Skrifter, set. A, 46 (Copenhagen, 1938), pp. 35-37; and H. Lechtman, "The Gilding of Metals in Pre-Columbian Peru," in Young, Science in Examination, pp. 38-52.

[25] W. Campbell, Greek and Roman Plated Coins, American Numismatic Society Numismatic Notes and Monographs 57 (New York, 1933), pp. 144, 174; and Lawrence H. Cope, "Surface-Silvered Ancient Coins," in E. T. Hall and D. M. Metcalf, eds., Methods of Chemical and Metallurgical Investigation of Ancient Coinage, Royal Numismatic Society Special Publication 8 (London, 1972), pp. 261-78, pls. 19-20 respectively. That some or all of these were ancient counterfeits does not alter the fact that they reveal ancient plating methods. On ancient counterfeits at this period, apparently a significant problem, note Sulla's "Lex Cornelia de falsis" of 81 B.c., on which see P. Grierson, "The Roman Law of Counterfeiting," in Robert A. G. Carson and C. H. V. Sutherland, eds., Essays in Roman Coinage Presented to Harold Mattingly (London, 1956; reprint Aalen, 1979), pp. 24061, here p. 242. Eutectic refers to that proportion of two metals being alloyed (here silver and copper) which has the lowest melting point (for silver and copper, the eutectic is about 70% silver, with a melting point near 780° C. See Robert C. Weast, ed., Chemical Rubber Company Handbook of Chemistry and Physics, 69th ed. [Boca Raton, Florida, 198889], p. D 184).

[26] Vitruvius 7.8.4; Pliny 33.64-5, 100, 125, and 34.162-3. Blumner, Technik and Technologie, vol. 4, pp. 133 and 308-9, says not earlier. Pliny 34.162-63 refers to tin-plating as well; cf. P. Leyden X 41 and Halleux, "De stagnum 'étang' a stagnum 'étain'; ' Antiquite classique 46 (1977): 557-70. For the use of arsenic on copper to imitate silver, see P. Leyden X 22, 83; Taylor, "Alchemy," p. 125, and now Ian A. Carradice and S. La Niece, "The Libyan War and Coinage: A New Hoard and the Evidence of Metal Analysis," Numismatic Chronicle set. 7, 148 (1988): 33-52, pls. 7-12. Mercury-gilding was also known in the pre-Columbian New World; see Bergs0e, "Gilding Process," pp. 27-28.

[27] E. R. Caley, "The Leyden Papyrus X," Journal of Chemical Education 3 (1926): 1149-66 and Halleux, Alchimistes.

[28] Konig, "Galvanisches Element," see fig. 3, label A.

[29] Ley, "Elements," p. 50; Schwalb, "Electric Batteries," p. 18; [Anonymous], "Batteries B.C.," pp. 112-13.

[30] Hornauer, "Elektrische Batterien," p. 15; Winkler, "Wiederentdecker," p. 72.

[31] Al-Haik, "Galvanic Cell," p. 104.

[32] M. Levey, "Evidences of Ancient Distillation, Sublimation and Extraction in Mesopotamia," Centaurus 4 (1955): 23-33. See also Taylor, "The Evolution of the Still," Annals of Science 5 (1945): 185-202; R. J. Forbes, Short History of the Art of Distillation (Leiden, 1948); and Levey, "The Earliest Stages in the Evolution of the Still," Isis 51 (1960): 31-34.

[33] A. R. Butler and J. Needham, "An Experimental Comparison of the East Asian, Hellenistic, and Indian (Gandharan) Stills in Relation to the Distillation of Ethanol and Acetic Acid," Ambix 27 (1980): 69-76.

[34] H, Diels, Die Entdeckung des Alkohols, Abhandlungen der Koniglichen Akademie der Wissenschaften, Phil. hist. Kl. 5 (Berlin, 1913), reprint in his Kleine Schriften, ed. W. Burkert (Hildesheim, 1969), pp. 409-41; and C. Anne Wilson, Philosophers, losis and the Waters of Life, Proceedings of the Leeds Philosophical and Literary Society, Literary-Historical Section 19 (Leeds, 1984), with the review of J. Dillon, Classical Review n.s. 36 (1986): 35-38.

[35] See Pliny 14.62 and Suetonius Augustus 34.5. 1 would like to thank Werner Krenkel for pointing out these passages to me (personal communication, 13 November 1989).

[36] Aristotle, Meteorologika 4.7 (38412-14) with I. During, Aristotle's Chemical Treatise: Meteorlogica, Book IV (Gbteborg, 1944), p. 45 (wine and vinegar are mostly water); Meteor. 4.10 (388x29-8b9) with During, pp. 54-55 (wine and vinegar = water + earth); Meteor. 4.10 (389x7-I1) with Dilring, pp. 56 (wine and vinegar are mostly water and congeal in the cold); Meteor. 4.11 (389x24-29) with During, pp. 56 (wine contains foreign heat; and Meteor. 4.11 (389b7-18) with During, pp. 57 (earth + water substances are hot).

[37] See my article "Horace Odes 1.13.3-8, 14-16: Humoural and Aetherial Love," Philologus 133 (1989): 75-81, here p. 76.

[38] See n. 16 above.

[39] Thompson, Assyrian Chemistry, p. xx: A.GEŠTIN. NA = "water of wine" (vinegar) and A.GE9T1N.NA. KALAG.GA = "strong water of wine" (possibly distilled vinegar). Cf. also Galen Pro puero epileptico 6 (11.374-78 Kiihn), where vinegar is strengthened by heating (though he does not refer to an άβμιξ΄