EXCHANGE EXCHANGE AUG 30 of Science anfc &rt of tf)e itfommittte of Council on REPORT ON THE ANALYSIS OF VABIOUS EXAMPLES OF ORIENTAL METAL-WORK, &c., IN THE SOUTH KENSINGTON MUSEUM AND OTHER COLLECTIONS. MADE UNDER THE DIRECTION OF PROFESSOR W. C. ROBERTS -AUSTEN, C.B., F.R.S., BY ARTHUR WINGHAM, F.I.G. LONDON: PRINTED FOR HER MAJESTY'S STATIONERY OFFICE, BY EYRE AND SPOTTISWOODE, PRINTERS TO THE QUEEN'S MOST EXCELLENT MAJESTY. 1892. Price Sixpence. of Science antr Hrt of ttje Committee of OTouneil on REPORT ON THE ANALYSIS OF VARIOUS EXAMPLES OF ORIENTAL METAL-WORK, &c., IN THE SOUTH KENSINGTON MUSEUM AND OTHER COLLECTIONS. MADE UNDER THE DIRECTION OF PROFESSOR W. C. ROBERTS-AUSTEN, C.B, F.R.S., BY ARTHUR WINGHAM, F.I.C. LONDON: PRINTED FOR HER MAJESTY'S STATIONERY OFFICE, BY EYRE AND SPOTTISWOODE, PRINTERS TO THE QUEEN'S MOST EXCELLENT MAJESTY. 1892. Price Sixpence. INTRODUCTION. THIS report is the result of an investigation undertaken at the suggestion of Prof. Valentine Ball, C.B., F.R.S., with the view to ascertain whether any marked characteristic existed in the composition of Oriental brass from different localities, or from the same locality but of different ages, and it was hoped that it would be possible by the aid of an analysis to name the locality or to assign the time of production to a brass of un- certain origin, or at least to support any existing idea as to the date of a doubtful specimen. It was necessary that the analyses should be as complete as possible, as a comparatively rough estimation of the principal metals only would be of little or no use. It was considered that a few exhaustive analyses would be more likely to yield definite results than a larger number of incomplete ones. Every speci- men was therefore subjected to a rigid examination, and the greatest care was bestowed, at times under considerable dis- advantages, on each analysis, so that every particle of metal could be accounted for and the quantity determined wherever there was sufficient present to admit of this being effected. The chief of the disadvantages referred to was the small quantity of the material at the analyst's disposal, which rendered the esti- mation of the metals present in small quantity a task of great difficulty, and one requiring special knowledge and skill. In commencing the work, and keeping the main purpose well in view, it was considered useless to spend time in analysing any but authentic specimens. In the selection of these, the aid of Mr. C. Purdon Clarke, C.I.E., was sought and obtained in respect to the specimens from the Indian Section of the South Kensington Museum ; while the choice of those from the Per- sian Court was entrusted to Mr. A. B. Skinner. Unfortunately very few brass or bronze objects in the Indian Section could be found whose authenticity as to date was beyond question, and these were nearly all comparatively speaking modern. A few brass specimens of doubtful age but of known locality were therefore chosen, as well as some objects of individual interest. Better fortune was met with in the Persian Court, where there are many authentic objects of various dates during the last ten centuries. A careful selection was made, with the result that the 15 analyses, Nos. 7 to 21 inclusive comprise a very interesting series of Persian objects, arranged in chronological order, extending over a period of from the 10th or o 66223. 500. 12/91. Wt. 237C4. E. & S. A 2 t 339366 % 2/: - 4 llth to the 19th century. These analyses show that the early Persian brasses and bronzes were indefinite mixtures, that the constituent metals were often impure, and further, that no definite composition was persistently aimed at at any one time, except perhaps in the specimens of the 19th century, which appear to be purer than the older ones, and which were probably worked up from imported English sheet brass. No regularity in composition could be traced. These results lead to the view that it would be impossible to judge with certainty the age of an alloy from its composition, which would, however, render it possible to distinguish between ancient and quite modern examples. The question as to the possibility of determining the locality of specimens is not entirely disposed of. It would probably be difficult to name the province in which a brass was made, but a complete analysis might often help to decide between one of two districts. This would apply more particularly to the ancient specimens which would probably have been prepared from the ores found in the locality, before communication with other countries was as easy as it has been of late years, but a much more extended series of complete analyses, including all traces of metals, would have to be made before definite conclusions could be arrived at. It is in connexion with this part of the subject that an instance is afforded of the value of carefully estimating the amounts of the metals present in small quantity. Thus it will be observed that in five out of the six Arabian specimens, Nos. 1 to 6, the percentage of lead is less than 7, while out of the 1 5 Persian specimens, Nos. 7 to 21, only three, Nos. 15, 18, and 21, contain less than TO per cent, of lead, and of these only one (No. 15) contains no zinc, by the use of which in Persia probably would be introduced a considerable amount of lead. Another specimen (No. 21) is very likely made from imported English-made brass, so that it might be safe to conclude that if an old oriental brass contained less than 1 *0 per cent, of lead, its brass was not made in Persia, and that in a question of choice between Arabia and Persia the probability would rest with it being Arabian. No. 18, (an Astrolabe) was, therefore, probably not of Persian, though it may possibly be of Arabian origin. Besides the results to which reference has been made many objects to which especial interest was attached have been analysed, and the results in most instances fully repaid the work bestowed upon them. Among these may be mentioned the Thysius stoneware bottle (No. 58) ; " Hanuman," the monkey ally of Rama (No. 54); the specimens of Bells (Nos. 50, 51 and 52); the long Indian gun belonging to H.K.H. the Prince of Wales (No. 53) ; and the iron dowel (No. 57), which must be at least 700 years old. There are also analyses of some specimens of Indian gold and a few objects of Indian Bidri work. The principal metals have been determined in some very ancient bronze objects in the British Museum, which were selected by Mr. A. S. Murray. The portions of metal analysed consisted of very small quantities of the filings which had been kept from the time the objects were being mounted on plinths for exhibition. Attention may also be directed to an examination of Japanese art metal-work, the treatment of which was until recently enveloped in obscurity in European countries. The results show that the composition of the metal is less important in the production of the beautifully coloured patina so characteristic of Japanese work than was at one time supposed. The tint of the patina apparently depends mainly upon the composition and the manipulation of the various pickling solutions employed, a fact which by no means detracts from the difficulties of the art, or from the skill of the Japanese worker. The descriptions at the head of the analyses are in most cases copies of the tickets exhibited with the objects in the Museum. It will be noticed that in a few instances the description of the metal does not agree with the results of the analysis. These analyses do not support the opinion that the Eastern metal-workers prepared and employed exceptionally pure metals. The following is the order in which the analyses, &c. have been arranged : Nos. Specimens from the Arabian and Persian Courts of the South Kensington Museum Brass and bronze - - 1-21 Persian steel - - 22, 23 Specimens from the India collections Gold and silver - - - 24-35 Bidri - - 36-39 Brass and bronze - 40-56 Iron dowel 57 Thysius bottle - 58 Japanese metal-work from various collections - 59-64 Japanese metal- work, coloured bronze plaques in South Kensington Museum - 65-75 Experiments on Japanese patina, page 49. Pickling solutions used by the Japanese, page 51. Antique bronzes in the British Museum - 76-82 The following analyses have been made by Mr. Arthur Wingham who has admirably conducted the work entrusted to him. W. C. ROBERTS- AUSTEN. December 1891. Analysis of Specimens taken from the Arabian and Persian Courts of the South Kensington Museum. No. 1. CANDLESTICK. For holding a wax caudle in the tomb of a Saint. Bronze, nine-sided, incurved in the middle of the base, chased with inscriptions in Kufic letters, figures in medallions, and other ornament, partly inlaid with silver. Persian. 9th or 10th century. H. 10 in., diam. 9 in. Bought, 51. 10s. 571.-78. ANALYSIS. Copper . . Arsenic . . Antimony Tin Lead Bismuth . . Iron Nickel .. Zinc, by diff. Silver Gold 79-77 Trace. Nil. 141 341 Trace. 100 Trace. 19-648 Trace. Nil. 100-000 '.. No. 2. Box AND COVER. Brass inlaid with silver, cylindrical, the edge of the cover bevelled and engraved with an Arabic inscrip- tion recording name and titles of El-'Adil Abu-Bekr II. (A.D. 1238-40), grandnephew of Saladin ; the sides covered with aureoled figures, hunting scenes, &rc., in chased silver inlay, enclosed in eightfoil dotted borders, on a ground of arabesques and flowers ; on the cover, diaper of hexagrams 8 enclosing six seated turbaned figures round central sun, within a zone of the signs of the zodiac. On the bottom, an inscrip- tion records that it was made at the royal magazine. Sara- cenic. A.D. 1238-40. H. 4 in., diarn. 4J in. Bought, 60Z. 8508.-'63. ANALYSIS. Copper . . Arsenic. . Antimony Tin Lead . . Bismuth Iron Nickel .. Zinc, by diff. Silver .. Gold 100-000 No. 3. TEAY. Brass, circular; inlaid with gold and silver, ornamented with Arabic inscriptions on a ground of flowers, leaves, and birds. The inscriptions record the name and titles of the Mem hik Sultan En-Nasir ibn Kalaun (A.D. J 293-1 309) and are divided by panels of flowers and medallions containing either the Sultan's name, or an escutcheon, charge, an antelope within a fence. Saracenic (of Cairo). A.D. 1293-1309 Diam. 31 in. Bought, 40Z. 420.-'54. ANALYSIS. Copper Arsenic Antimony Tin Lead Bismuth Iron Nickel .. Zinc, by diff. Silver .. Gold 100*000 No. 4. CANDLESTICK. Brass, inlaid with silver, and engraved with Arabic inscriptions recording Memluk titles of the 14th cen- tury round base and neck and on other parts, divided by medallions of geometrical designs, and by rosettes of flowers. Saracenic. 14th century. H. 10 in., diam. 9 in. Bought, 61. 8s. 4505.- 58. ANALYSIS. Copper Arsenic Antimony Tin Lead Bismuth Iron Nickel .. Zinc, by cliff. Silver .. Gold 100-000 No. 5. BOWL. Brass, inlaid with silver, and engraved with Arabic inscriptions, in a zone divided by four medallions, recording name and titles of the Memluk Sultan Kait-Bey (A.D. 1468- 96), on a ground of delicate arabesque ornament, and beaten out in a larger floral design engraved with scroll work. Saracenic. 15th century. H. 6 in., diam. 16 J in. Bought, 101. 10s. 1325.-'56. ANALYSIS. Copper Arsenic Antimony Tin Lead . . Bismuth Iron Nickel . . Zinc, by diff. Silver .. Gold 100 -(XX) 10 No. 6. VASE, one of a pair. Brass, engraved with geometrical and arabesque ornament and Arabic religious inscriptions, and inlaid with silver dots. Saracenic (of Syria). 18th century. H. 12 in., diam. 6i in. Bought, 12Z. the pair. 379.-80. ANALYSIS. Copper Arsenic Antimony Tin Lead . . Bismuth Iron Nickel .. Zinc, by diff. Silver .. Gold 100-000 No. 7. MORTAR. Bronze, octagonal, chased with arabesques and Kufic inscriptions. Bound the sides are projecting knobs, and on one side is a ring suspended from a bull's head. Found in the ruins of the city of Rhages. Persian. 10th or llth century. H. 5J in., diam. 7 in. Bought, Si. 4G6.-76. ANALYSIS. Copper . . Arsenic Antimony Tin Lead Bismuth Iron Nickel .. Zinc, by diff. Silver . . Gold 100-000 11 No. 8. MIRROR. Bronze, circular, the back ornamented in relief with two rampant sphinxes, addorsed, among floral scrolls ; around these is a Kufic inscription. Various inscriptions and devices are engraved on the back. Persian or Saracenic, llth (?) century. Diam. 4 in. Bought, II 10s. 442.-S7. ANALYSIS. Copper . . Arsenic Antimony Tin Lead .. Bismuth Iron Nickel .. Zinc, by diff. Silver . . Gold 100-000 No. 9. ASTROLABE ? Bronze, engraved. It appears to have been gilt, and is slightly inlaid with silver. It bears the date of the Hegira 598 (A.D. 1202), and was made at Damascus by Abdul Rahman, son of lusuf. Persian or Saracenic. ' 9| in. by 8 in, Bought, 71. 17s. 6d. 504.-' 88. ANALYSIS. Copper . . Arsenic Antimony Tin Lead .. Bismuth Iron Nickel .. Zinc, by diff. Silver .. Gold 100-000 12 No. 10. BOWL. One of a pair. Brass, chased with inscriptions and diaper bands, partly inlaid with silver. Persian. 13th or 14th century. H. 4J in., diam. 9J in. Bought, 41. the pair. S59.-76. ANALYSIS, Copper . . Arsenic Antimony Tin Lead . . Bismuth Iron Nickel .. Zinc, by diff. Silver .. Gold 100-000 No. 11. LAMP STAND. Brass, cylindrical, chased with flowers, lozenges, and zigzag bands, and an inscription, filled in with black inlay. Persian. 13th or 14th century. H. 11 J in., diam. 7J in. Bought, 11 17s. 6d 4S3.-76. ANALYSIS. Copper . . Arsenic Antimony Tin Lead . . Bismuth Iron Nickel .. Zinc, by diff. Silver . : Gold 68-03 Trace. Nil. 5-986 5-950 Trace. 1-239 269 18-526 Nil. 100-000 No. 12. WATER POT. Brass, with scroll handle and curved spout, inlaid with bands of inscription arid arabesque ornament in silver. Persian. Bought, Dated, A.H. 866 (A.D. 1463). H. 6f in., diam. 5 J in. 943.- 86. ANALYSIS. Copper . . Arsenic Antimony Tin Lead Bismuth Iron Nickel .. Zinc, by diff. Silver . . Gold 100-000 No. 13. POT WITH HANDLE. Bronze, chased with flowers and inscrip- tions. Old Persian. 15th century. H. 4f in., diam. 7| in. Bought, 15s. 538.- 76. ANALYSIS. Copper . . Arsenic Antimony Tin Lead Bismuth Iron Nickel .. Zinc, by diff. Silver .. Gold 73-19 Trace. Nil. 6-317 16-980 Trace. 378 Trace. 2-649 436 050 100-000 No. 14. MORTAR. Bronze, with cylindrical body encircled by raised bands, and a broad flange at top and bottom. The top flange is incised with an Arabic inscription and flowers, and the body with circular ornaments. Persian. 15th or 16th century. H. 7 in., diam. lOf in. Bought, 21. 15s. 948.- 86. ANALYSIS. Copper . . Arsenic Antimony Tin Lead .. Bismuth Iron Nickel .. Zinc, by diff. Silver . . Gold 100-000 No. 15. BOWL. Bronze, engraved with an inscription in Persian characters. Persian. 1511 A.D. Bought, 9. 3s. 9c. 1191.-54. ANALYSIS. Copper . . Arsenic Antimony Tin Lead Bismuth Iron Nickel .. Zinc Silver . . Gold 99-766 15 No. 16. BOWL. Light bronze, with incurved mouth, the outside engraved all over with figures in medallions, Kufic inscriptions, and diaper ornament. Persian. 16th century. H. 3^ in., diam. 7 in. Bought, 13s. 7.- 86. ANALYSIS. Copper . . Arsenic Antimony Tin Lead Bismuth Iron Nickel .. Zinc, by diff. Silver .. Gold 75-80 Trace. Nil. 1-211 2-514 Trace. 210 068 20-145 052 100-000 No. 17. BOWL. Bronze, engraved outside with bands of inscription filled in with floral details. Persian. Apparently dated A.H. 1021 (A.D. 1612). H. 5* in., diam. 14* in. Bought, II 16s. 770.-88. ANALYSIS. Copper . . Arsenic Antimony Tin Lead Bismuth Iron Nickel .. Zinc, by diff. Silver . . Gold 70-51 110 Nil. 047 6-900 Trace. 266 076 22-016 075 100-000 16 No. 18. ASTROLABE. Brass, round, chased, with five inner plates. It is dated 1074 A.H. (A.D. 1663). Persian. 17th century. Diam. 7 in. Bought, 61 15s. 530.-76. ANALYSIS. Copper . . Arsenic Antimony Tin Lead Bismuth Iron Nickel .. Zinc, by diff. Silver .. Gold 100-000 No. 19. ASTROLABE. Brass, circular, chased or engraved all over, with revolving perforated disc and five inner plates. Made by the famous Abdul Ahmeh, and dated A.H. 1127 (A.D. 1715). Persian. Entire L. 8 in., diam. 4J in. Bought, \2l. 2s. 5d. 458.-'88. ANALYSIS. Copper . . Arsenic Antimony Tin Lead .. Bismuth Iron Nickel .. Zinc, by diff. Silver . . Gold 100-000 17 No. 20. BASIN. " Lagan." Brass, with wide curved lip, chased with floral ornament. Persian. 18th century. H. 5 in., diam. i n . Bought, II 15s. 952-86. ANALYSIS. Copper . . Arsenic Antimony Tin Lead Bismuth Iron Nickel . . Zinc, by cliff. Silver . . Gold 75-15 789 Trace. 4-727 5-988 Trace. 798 Trace. 12-548 Trace. Nil. 100-000 No. 21. VASE WITH COVER AND HANDLE. " Dakhl-i-pul." Brass, pierced with floral designs, and chased with figures in medallions and a band of inscriptions. These vessels are suspended in shops in the bazaars, principally in the kebab or roast meat shops, for the purpose of holding copper money. Persian. 19th century. H. 11 in., diam. 6 in. Bought, 21. the pair. 510a.-76. ANALYSIS. Copper . . Arsenic Antimony Tin Lead . . Bismuth Iron Nickel Zinc, by diff. Silver . . Gold 100-000 e (6223. B 18 No. 22, BOWL. Watered steel, with garlands, trellis ornament, &c. in gold inlay outside, and an inlaid inscription at the bottom inside. Persian. H. 3f in., diam. 8J in. Bought (with 1304 to 1317), 36Z. 1312.-74. ANALYSIS. Carbon -94 No. 23. BOTTLE WITH COVER, one of a pair. Steel, chased with flowers, birds, animals, and inscriptions, inlaid with gold. Persian. 19th century. H. 18| in., diam. 5J in. Bought, 24Z. the pair. 383.-'80, ANALYSIS. Carbon .. -92 EEMAEKS ON ANALYSES. (1 to 23.) The foregoing results possess a collective value, and a close examina- tion of them will bring to light many points of interest. The first six specimens are Arabian, then follow 15 Persian brasses and bronzes chronologically arranged, and finally two Persian steel objects. It will be noticed that the Arabian specimens, and especially the earlier ones, appear to be richer in copper than the Persian, and are of the nature of brass as we now understand it, i.e., an alloy of copper and zinc. As previously pointed out in the introduction, five of the six Arabian specimens contain less than 0*7 per cent, of lead while only three of the 15 Persian analyses give low results. This would indicate that the Arabian zinc ores were purer, or, at any rate, freer from lead, than the Persian. Zinc as an isolated metal was not known in early times, and brass was made by melting copper with zinc ores or reducing copper ores in conjunction with those of zinc. Consequently any metallic impurities in the ores themselves would almost with certainty- show themselves in the resulting brass. Lead as Galena is generally present in considerable quantities in the zinc ores known as Blende, which also often contain small amounts of iron and other metals. There is also a marked difference in favour of the Arabian samples in respect to the per-centage of iron. In good modern brass, the iron is nearly always below 0-17 per cent, and the lead should be less than 5 per cent. ; more than these quantities would indicate that a bad variety of zinc had been used in the preparation of the alloy. A considerable amount of iron might indicate one or both of two things : either that impure zinc had been employed or that the brass had been frequently remelted and stirred or brought into contact with iron materials. The large quantity of iron, indicated by most of the analyses of the Persian specimens could not be accounted for by the latter 19 explanation. Consequently its presence may with certainty be attributed to the fact that it existed as an impurity in the zinc ore from which the brass was made. This would be supported by the irregularity in the quantity present, and by the fact that in the bronze object (No. 15) where zinc is absent the per-centage of iron is very low. A curious point about these analyses is the large quantity of silver occasionally present, especially in some of the older specimens, Nos. 8, 12, and 13. Gold is rarely present, and in only two cases, Nos. 12 and 13, is there a sufficient quantity to render its estimation possible. Nickel is generally present, but in small amounts, Nos. 5, 11, and 2 being the highest records. Arsenic as a rule is low, as it should be in brass, but it is very high in No. 20 and in Nos. 14 and 2, whilst there is over 0-1 per cent, in Nos. 9 and 10. Antimony is generally absent, though sometimes a "trace " is present. It is a great enemy to brass, rendering it crystalline and brittle. The amount, therefore, in No. 3 is excessive, and the existence of only a trace of arsenic in conjunction with it is remarkable. This analysis is unique as showing a large predominance of antimony over arsenic, an unusual occurrence the amounts being generally reversed. Another instance of this a still more marked one will be noticed amongst the Indian specimens (No. 40). Bismuth was only just detected in some cases by a very delicate test, and has been recorded as a trace in all the specimens (Nos. 1 to 21) except No. 5, where it was absent. In respect to the principal metals of these alloys (copper, zinc, lead, and tin) there seems to be no regularity whatever in the relative amounts present. Of course the copper largely predominates, and is always over 60 per cent. ; if less were present the metal would be rather unworkable. The zinc next predominates over the other two, except in one or two instances when the lead takes the second place ; sometimes the tin predominates over the lead, and sometimes over the zinc, but it never ranks second to the copper except in the case of No. 15, which contains no zinc and very little lead, and is not a brass but a typical bronze, in which of course tin is the secondary metal. The per-centage of copper is irregular, but on the whole high, omitting the very low examples. The amount generally present in modern brass for industrial purposes is between 66 and 75 percent., the remainder being zinc with about 0-5 per cent, of lead which exists as an impurity in the zinc. " Yellow " or " Muntz metal " for ship sheathing, etc. contains 60 per cent, of copper. Ancient brasses possess> better colour than modern varieties, the majority of which would probably contain less than 70 per cent, of copper. The superiority in colour of the older examples may be due entirely to the composition of the metal. In this report are a number of analyses which it would be of great interest to scrutinise in conjunction with the tints of the metals, and which might prove of much value to the modern art metal-worker in the selection of his metal. The 19th century specimen (No. 21) looks very much like English ^ yellow metal made with a bad or " dross " zinc containing much lead and iron an inferior brass not sufficiently good for home manufacturers and consequently exported to the East. The last two specimens in this list, Nos. 22 and 23, are what are technically called high-carbon steels. They are as rich in carbon as die-steel and as some tool-steels. They are very hard, and a great amount of time must have been spent in working them. B 2 20 Analysis of Specimens taken from the Indian Court of the South Kensington Museum. No. 24. BURMESE GOLD. Taken from various objects in the Royal Treasure from Mandalay, now in the India Museum. ANALYSIS. Gold 90-70 Silver 6'92 Copper . . -92 Iron -33 Platinum and other rare metals . . 1 12 99-99 No. 25. THRONE, octagonal, on eight feet ; gold plates chased and repousse' with bold floral and foliated ornament, mounted on wood ; with three red and yellow velvet cushions. Made for Runjeet Singh after his accession to Lahore in 1799 A.D. H. 3 ft. 1 in., diam. 2 ft. 5 in. 2518. ANALYSIS. Gold Silver Copper Iron ... . . . . :: v 99-87 No. 26. GOLD COIN. Moghul. 16th century. Transferred from the India Office. Mus. No. -1882. ANALYSIS showed that it is of pure gold, as not a trace of any other metal could be detected. Specially tested for silver, platinum, copper, iron, and lead. Specific gravity of coin .. .. 19*265 21 No. 27. GOLD COIN (Indo-Scythiari). Specific gravity - - - 18*698 ANALYSIS. Gold Silver Copper 100-01 No. 28. RELIC SHRINE. " Stupa," in four parts (incomplete). Gold repousse and chased with conventional ornament in bands. Burmese. Discovered in levelling a Buddhist Temple at Rangoon in April, 1855. H. 15 in., diam. 12^ in. 02755. ANALYSIS. Gold 72-02 Silver 22' 96 Copper . . . . . . . . 5 24 100-22 No. 29. RELIC CASKET, cylindrical ; gold, repousse*, set with two rows of rubies, and surrounded by an arcade containing figures of saints, the spandrils filled by birds with out- stretched wings ; on the base a conventional lotus. Found in Afghanistan. Buddhist. 1st century B.C. H. 2f in., diam. 2f in. Lent by the Secretary of State for India in Council. ANALYSIS. Gold Silver Copper 100-13 22 No. 30. 'RELIC SHRINE. " Stupa." Gold repousse and chased with conventional ornament in bands, surmounted by a finial terminating in a bud ; set with 38 rubies and one emerald. Burmese. Discovered in levelling a Buddhist Temple at Rangoon in April, 1855. H. 14J in., diam. 7| in. 02756. ANALYSIS. Gold Silver Copper 100-19 No. 31. TASSEL. 16 stems, gold plates held by spirals of gold wire. Burmese. Discovered in levelling a Buddhist Temple at Rangoon in April 1855. 02753. ANALYSIS. Gold Silver Copper 99-83 No. 32. RELIC SHRINE. " Stupa/' in three parts (incomplete). Gold repousse' and chased with conventional ornament in bands. Burmese. Discovered in levelling a Buddhist Temple at Rangoon in April 1855. H. 2 in., diam. 3 in. 02754. ANALYSIS. Gold Silver Copper 99'96 23 No. 33. iBowL. Gold, unornamented, containing calcined bones and ashes. Burmese. Discovered in levelling a Buddhist Temple at Rangoon in April, 1855. H. 5 in., diam. 9J in. 02752. ANALYSIS. Gold 83-07 Silver 10-36 Copper 6-43 99-86 No. 34. HELMET, MODEL OF ; probably a votive offering. Gold, with a double pin, and 62 jewels set in a border of repousse orna- ment. Discovered in levelling a Buddhist Temple at Rangoon in April, 1855. 02758, ANALYSIS. Gold Silver Copper 100-21 No. 35. ANCIENT GREEK SILVER PATERA. Found by Dr. Lord in Badakshan. 4th century, A.D. Lent by the India Office. ANALYSIS. Silver Gold Copper Iron 99-88 24 No. 36. EWER with COVER. Oxidised metal, damascened with silver. Indian, modern. H. 2 ft. 11 in,, diam. 2 ft. 4 in. Given by Her Majesty the Queen. 587.-'54. ANALYSIS. Lead Copper . . Iron Silver . . Zinc, by diff. 100-000 No. 37. BUDDHA, seated, praying ; zinc. H. 4 in. 471. ANALYSIS. Copper Tin Lead Iron Nickel .. Silver .. MM 3 '758 Impurity. Zinc, by diff 96 '242 100-000 No. 38. BOTTLE with COVER. Metal, with raised chasing of flower and leaf pattern, covered with gold and silver foil. Hungarian or Indian. Latter half of 17th century. H. 14| in., diam. 5J in. Bought Wl. 356-70. ANALYSIS. Lead .. '956 Tin .. .. .. .. -346 Iron -084 Copper . . . . . . . . Nil. 1.386 Zinc, by diff. 98 '614 25 No. 38A. VASE. Oxidised metal, damascened with silver. Indian, modern. H. 23 in., diam. 22 in. Given by Her Majesty the Queen. 585 -'54. ANALYSIS. Lead.. .'." Tin Iron Copper 8-381 Zinc, bydiff. .. .. .. 91 '619 No. 39. WATER BOTTLE. Zinc, or a composite metal resembling it, of elongated form, with flowers and animals in silver inlay, and a foot and two bands of chased silver originally filled in with cloisonne' enamel. Resembling the Bidri work of India. Persian ? 17th or 18th century. H. 13 J in., diam. 5 in. Bought, 17. 15s. 950.-86. From the Persian Court. ANALYSIS. Lead Tin Iron Copper 5-794 Zinc, bydiff. 94 '206 No. 40. VASE with HANDLE and COVER, "PHULDAN." Perforated brass ; the cover surmounted by a knob ; the handle formed 26 "by two dragons. Nepal. H. 11 J in., diam. 7^ in. (Colonial and Indian Exhibition, 1886.) Bought, 21. 10s. 89. (IS.) 1886. ANALYSIS. Copper . . Arsenic . Antimony Tin Lead .. Bismuth . , Iron. Nickel and Cobalt Zinc 99-701 No. 41. LOTAH. Cast brass, with chased ornamentation. The body is ornamented with deeply indented intersecting rings. Borneo. Modern. H. 4 in., diam. 4J in. Bought (Nos. 54 to 67), 101. 54. (I.S.) 1886. ANALYSIS. Copper . . Arsenic . . Antimony Tin Lead Bismuth . . Iron Nickel and Cobalt Zinc 59-80 Trace. Nil. 298 3'226 Nil. 2-184 121 34-53 100-159 No. 42. PAEESNATH, the twenty-third Tirthankar, or Saint of the Jains ; impure brass. The figure has conventional features and hair, is nude except a girdle, and seated cross-legged on a trian- gular base, the front of which has conventional ornament in 27 compartments, and is faced with copper and inlaid with silver. Madras. H. 2 ft. 2 j in. W. I ft. 9 in. D. 1 ft. 1 in. 467. ANALYSIS. Copper . . Tin Lead Bismuth . . Iron Nickel . . Zinc, by diff. Silver . 100-000 No. 43. SHRINE FOR A JAIN SAINT, in three portions ; brass with bronzed surface. A semi-circular arch supported by two subordinate ornamental shrines, each with a figure of an attendant bearing a chowrie. The bases and crown of the arch have each a small Jain temple enclosing the representation of a Saint. The arch bears a row of seated musicians, and within it is a star-shaped halo, and two elephants bearing musicians and worshippers. Madras. H. 3 ft. 8J in. W. 3 ft. 2 in. 713. ANALYSIS. Copper . . Tin Lead Bismuth . . Iron . . Nickel .. Zinc, by diff. Silver 75-83 Trace. 3-169 Trace. 196 Trace. 20-805 Cons, trace. 100-000 No. 42 was considered to be copper or bronze, but on scraping was found to be brass. In scraping, a fine red dust came off which consisted largely of copper oxide, then the surface became more metallic with copper, on the removal of which the brass was exposed. The brass contained veins and blotches, apparently of copper. No. 43. Brass, with bronzed surface, somewhat similar to above, but the metal was cleaner, softer, and better made. These figures appear to have been originally all brass, from the surface of which the zinc has by some means been removed (by burning or by acid), and during which process the copper has become partly oxidised. 28 No. 44. LOTAH, with RIBBED BOWL. Brass. Calcutta. 04,629. ANALYSIS. Copper Arsenic Antimony Tin Lead Bismuth Iron Nickel and Cobalt Zinc, by diff. No. 45. GOGLET. Brass. Travancore. ANALYSIS. Copper Arsenic Antimony Tin Lead Bismuth Iron Nickel and Cobalt Zinc 100-000 04,644. 99-818 No. 46. TANJORE BRASS STAND for offering Betel in Temples. 2269. (I.S.) ANALYSIS. Copper Arsenic Antimony Tin Lead Bismuth Iron Nickel and Cobalt Zinc 1883. 99-721 29 % No. 47. BULL, brass, on a base with incised ornament ; a representation of the Nandi, the sacred bull of Shiva. Benares. 18th century. H. 6 in., L. 5 in., W. 3 in. 2162. (IS.) 1883. ANALYSIS. Copper Arsenic Antimony Tin Lead Bismuth Iron Nickel and Cobalt Zinc, by diff. 100-000 No. 48. BELL, brass, handle surmounted by a group of figures. Tanjore, Madras. 02 ; 930. ANALYSIS. Copper . . . . . 77-14 Arsenic . . . . . . . 136 Antimony . . Trace. Tin . . . . . . . 20-83 Lead . . . . . . 410 Bismuth . . 048 Iron . . . . 093 Nickel . Trace;. Zinc . . . . . . 398 Gold and Silver Nil. 99-055 No. 49. BOWL, with figures and emblems in relief. Brass. Java. Copper Arsenic Antimony Tin Lead Bismuth Iron Nickel .. Zinc ANALYSIS. 04,831. 98-941 30 The above figures were all confirmed by repented analysis. The sample was probably not clean metal and contained a quantity of oxide, which would account for the 1 per cent, discrepancy. No. 50. BELL, without clapper. Burmah. 9,291. ANALYSIS. Copper . . Arsenic Antimony Tin .. Lead Bismuth Iron Nickel Zinc Gold .. Silver 100-041 INSCRIPTION (translation). " In the month of Tabohdwe (February) on the fifth of the waning moon, in the year 1204 (1842, A.D.) on a Sunday at about 4 p.m. this bell was cast and moulded of pure copper. Its weight is 594,049 kyats (an obvious mistake). There are four lions on the hanging apparatus. Its height is nine fingers' breadths, the diameter five inches, the circum- ference fifteen, the thickness twenty-four. It is called the " Mahahtee Thadda Granda" (the great sweet sound). " The man who had this royal bell moulded was the Burman King, Tharrawahdy, Kohn Boung Min." No. 51. BELL, with dragons in relief and incised inscription. Burmah. 05,219. ANALYSIS. Copper Arsenic Antimony Tin Lead Bismuth Iron Nickel Zinc . . Silver Gold 99-840 31 The bell contained on its outer surface a long inscription, of which the following is part (translation) : " We two, brother and sister, have given this bell as an offering to the seven Precious Things. The exact weight of the bell in current reckon- ing is 2,500 kyats. In this attempt to merit Neh'ban our arrangement was as follows : We took our own weight in gold, and in silver and bright copper and other metal (Laukad, the Pali word used, implies five metals, gold, silver, copper, iron, lead), and mixed them well together. In the year 1209 (1847, A.D.) in the hot season, at a fortunate hour I had it moulded, setting my heart on giving it in alms. " As I wrote the inscription I offered up abundant prayer, that no enemies or troubles might come nigh me, and that I might attain Neh' ban. " Then I dedicated it." No 52. BELL, bronze. Swinging shackle ornamented with two grotesque lions. With inscription, dated 1828 A.D. Burmese. H 3 ft. 4 in., diam. 2 ft. 2 in. Presented by Commander C. McLaughlin, RN. 76. (I.S.) 1884. ANALYSES of Two SAMPLES, the first taken from the lip, the other from the top. Copper Arsenic Antimony Tin Lead Bismuth Iron . . . . Nickel \ Zinc / Silver Gold 99-702 Copper . . Arsenic Antimony Tin Lead . . Bismuth Iron Nickel 1 Zinc J Silver .. Gold 99-778 32 These two samples of the same object were taken to ascertain whether the metal was the same at one end of such a heavy and large Oriental casting as at the other, this being of especial interest as the amount of metal in the bell is very much more than could be melted in one pot, such as the Indian metal-workers could obtain. Weight, about 9 or 10 cwts. It is well known that the Indians, when casting a large object, divide their metal into lots, which are placed and melted in small pots heated by separate fires, situated all round the foundry. When everything is ready each pot is taken out of its fire in turn and its contents poured into the mould. Consequently a large casting may have a very different composition at its opposite ends, especially if the alloy be a complex one. From the above analyses the metal in this instance appears to be more homogeneous than would have been expected. A curious point about the analyses also is the tendency shown by the heavier metals* to predominate in the sample from the top, and vice versa in that from the lip, distinctly pointing to separation. This would seem to suggest that the metal was cast in a hot mould and an attempt made to mix the various additions in the mould itself, with the conse- quent slow cooling, and, as a result, partial separation of the metals. It would also tend to show that the bell was cast upside down, lip uppermost. This bell contains a most interesting inscription full of pious sen- timents, after which come the following extracts : In the city of Yathanahpoorah there lives a King, the Tshat-tan- seng -Meng-Shung -Bwah- Sheng-Meng-darah-gyu, who loves and honours God; .... and this King's devoted servant who can supply all the wants of his Sovereign instantaneously, who is gifted with all intelligence and bravery, and who also pursues the practices of a good man, first collecting all the property which he himself and his family have honestly acquired and then raising a sub- scription from all pious and good citizens, have with such means caused this bell .... of brass to be cast at the foot of the Tshoolay pagodah, and consecrated it to that pagodah. For this good act in thus consecrating this bell, which sounds well, I pray that at a future period I may attain the same knowledge as God himself possesses, .... and thus become an Aruyah. I also pray that for this good act of mine the same blessing may attend my father, mother, teacher, the King, and the inhabitants of the three states 1 call the Nuts (devils) of this earth to bear witness to this my act. The Angel has placed his Seal. I pray for knowledge of the past, of what exists, and of what is to come, or to see what is before me. Accomplished. No. 53. BURMESE GUN, in form of a dragon. Lent by H.R.H. Prince of Wales. Length 9 ft. 3 in., bore 3} in. 1887. * Lead, bismuth, antimony, and tin. ANALYSES of two samples, the first from the mouth end, the other from the tail or breech end. Copper Arsenic Antimony . Tin Lead Bismuth Iron . . Nickel Zinc Gold Silver . 99-788 Copper 85-93 Arsenic Trace. Antimony. . Trace. Tin 7-654 Lead 4-748 Bismuth . . Trace. Iron 210 Nickel . . Trace. Zinc 1-501 Gold Nil. Silver Nil. 100-043 No. 54. HANUMAN ; the monkey ally of Rama ; copper and brass, cut through to show process of casting. Madras. Modern. H. 4 in. 726. This specimen of art metal work is of peculiar interest, insomuch as it was apparently a compound casting showing on its surface two dis- tinct metals, viz., bronze and brass, very much interspersed. The figure was not of one metal altered in colour in parts by superficial treatment, as inspection showed the colour to be due to the metal itself. Owing to the intermixture of the metals the way in which they protruded one beyond the other at different points, and the small quantity at parts of one metal over the other such as the bracelets, armlets for example it was difficult .to understand from outside exa- mination how these figures were made. With the object of solving the question it was considered that some light might be thrown on the s'ubject if the casting were cut in halves from top to bottom. This was done, and immediately the whole process was explained, as it is quite clear that a core of copper was originally cast, of a shape showing due regard to the result desired, and that the brass was cast round the copper. o 66223. C 34 From what is known in connexion with small Oriental castings, it is probable that the following process has been employed in the production of these double castings. First, a model has been carved in wax, of a shape and size necessary to bring out the copper where that metal is required at the surface, and leaving space where the yellow surface is desired for the future casting of brass. This wax model has then been moulded, the mould heated and the wax melted out. after which the copper has been cast in the mould. Then round this half figure, with its prominent parts, where necessary, more wax has been cast, and carved into the shape of the figure ultimately required. The whole has then been moulded, the wax removed as before, and the brass run into the mould filling up the spaces existing between it and the copper core in the centre. The double casting has then been removed, the brass filed down, wherever it might accidentally and unnecessarily have- covered the copper until the red metal was exposed, the whole being then chased and completed. This casting is probably about 40 years old, and was made in Madras. The art of double casting, as represented by this figure, is very old (a thousand years or more), and has been practised at, and almost entirely confined to, the east coast of the Madras Presidency. Mr. Havell, of the School of Art of Madras, who has been engaged on a survey of Art manufactures for the Government, reported in 1887, that these castings were no longer made. Mr. C. Purdon Clarke thinks that if this be true it is probably due to the high cost of pro- duction. No. 55. ONE of many coins found in the delta of the Nile. Ancient Roman. Lent by C. Purdon Clarke, Esq. ANALYSIS. Copper . . Silver .. Gold . . Arsenic . . Antimony Tin Lead 99-931 No. 56. TEAY or KHORASSAN METAL. Persian. Lent by C. Purdon Clarke, Esq. ANALYSIS. Copper . . . . . . . . 50-37 Nickel Cobalt Iron Lead Zinc, by diff. 100-00 35 No. 57. IRON DOWEL FOR STONEWORK. From the stonework of a temple destroyed by Altomash. - A.D. 1211-1236. Actual size of dowel. Section through AB was polished and etched with perchloride of mercury. The etched surface showed the presence of a con- siderable amount of slag. The section between the lines CD was taken for analysis. ANALYSIS. Silica -20 =-44 Slag. Carbon . . . . . . . . Trace. Manganese . . . . . . Nil. Phosphorus . . . . . . Nil. The dowel is evidently part of a bloom hammered up into shape. The iron is practically pure, but the dowel contains about a half per cent, of slag, which had not been eliminated in working. No. 58. THYSIUS STONEWARE BOTTLE. Egyptian. c 2 36 ANALYSIS. Insoluble in hydrochloric acid. Silica 65-32 Alumina 19' 22 Ferric Oxide 5 '60 Lime T24 Magnesia '68 92-06 Soluble in hydrochloric acid. Alumina Ferric Oxide Lime Magnesia Potash Amount insoluble in hydrochloric acid . . 92 * 64 Silica alumina, &c., as above . . . . 92*06 Amount of potash insoluble in hydrochloric acid .... -58 The inside of the bottle contained an incrustation. This was exa- mined. Mercury could not be found chemically or microscopically, showing that the bottle is not a mercury bottle. The deposit consisted of chalk, with a small trace of nitrates and chlorine, which shows that this bottle has probably been used for carrying water. The deposit is not from sea water, as the chlorine is present in such small traces. The bottle, from its composition, has evidently been made from a granitic alluvial deposit, and on examining microscopically crystals of black mica and felspar were seen. HISTORY. This bottle was one of a kind found in many parts of different coun- tries, sometimes in large heaps, sometimes only two or three together. They are generally about four or five inches high, three or four inches broad, and have a very small orifice at one end. They are somewhat of this shape, and are not made to stand. The actual origin of these bottles has been very much discussed, and various uses have been assigned to them. 37 The bottle examined was found to be incrusted with a thin deposit of chalk,^ which indicates that this bottle was used for carrying water. The deposit was not, as might be supposed, the result of submergence under tn sea and subsequent evaporation of the water, as the absence of anything but the merest trace of chlorine shows the chalk to have been derived from fresh water. It has been suggested that these bottles could not be used as water bottles, as they have such a very small orifice, and they are supposed to have been used as mercury bottles. Both the incrustation and the bottle itself were examined most carefully for mercury, both chemically and microscopically, and not a trace could be found. It has also been suggested that they might have been used for carry- ing the commercial articles of the day, such as saltpetre, or even used as bombs, but the absence of anything but a mere trace of nitrate renders both these suppositions improbable. The trace of nitrate, on the other hand, tends to support the view that they were used for water, as the quantity found was about equal in proportion to the chalk in the ratio- often found in some fresh waters. The bottle is very hard and cuts glass readily. A fracture showed a black or dark grey interior with yellow portions near both the inside and outside edges. The material had not been fused, but had been heated to a sufficiently high temperature to frit the mass together. From the regularity of the marks on the inside and outside of the bottle it is evident that some kind of potter's wheel has been used in turning the bottle into shape. From the analysis the material used was apparently some alluvial deposit resulting from the decomposition of granite. Examining a powdered portion under the microscope, pieces of Felspar were plainly discernible, as also some rounded pieces of black mica and horn- blende. Finally, the suppositions arrived at are as follows : That the material from which the bottles were made was a river deposit from a granitic source ; in this instance, probably the bed of the Nile. That the deposit was used as it was found, having been ground fine enough by the water's action. That a kind of potter's wheel was used in turning the material into shape. That the firing was sufficient to just frit the bottle material. Finally, that the bottles were used for carrying water. Prof. Judd has had a microscopical section made in his department r which, on examination, may throw further light on the geological ques- tion. He has also promised to obtain some Nile deposits, and it will be of interest to know whether or not they give the same peculiar dark grey appearance on being " fritted." The following is the result of Prof. Judd's examination : " Felspar very distinctly present, and also a few crystals of horn- blende. Plenty of quartz. All imbedded in a kind of glass." " The material from which the bottle was made might possibly have been a deposit of the Nile. The section had a great resemblance to such deposits, and could very well be of that origin, but is not proof of such, although with other evidence it would tend to prove so." " Nile deposits vary very much in their constituents, some being muddy, others less muddy with more sand." 38 A sample of Nile deposit, apparently muddy, was baked at a low temperature and gave a red cake due to excessive iron present. In a reducing atmosphere at a white heat the mass fused into a dark-coloured semi-glass. Thus it may be that with a deposit containing more sand a temperature might be obtained sufficient to fuse the mud, and thus, with the sand and other crystals of material present unfused but incased in the fused matrix, a material might be obtained very easily having the appearance of the bottle examined. From the fact that this bottle was found in Egypt also it is highly probable that the material from which it was made was a Nile deposit. REMARKS ON ANALYSES. (24 to 58.) The analyses Nos. 24 to 58 need but little comment. The first 11 are specimens of gold, most of which appear to be native gold which little attempt has been made to purify. Native gold invariably contains silver, and very frequently copper. Some of the analyses show that means were adopted to purify the metal, especially No. 26, and Nos. 25, 30, and 32. All the rest are very impure. No. 26 consisted of gold, the purity of which is very remarkable con- sidering it is of the 16th century, an age when the methods of refining were somewhat crude. The specific gravity of pure gold varies accord- ing to its mechanical treatment between 19 25 and 19 36 ; this specimen was 19*265. The analysis No. 24 is an exceptional one as it indicates the presence of over I'O per cent, of platinum and other rare metals, and a small quantity of iron. It would be of interest to trace the source of this gold if it is possible to do so. The specimens Nos. 28 to 34 are all exhibited in the same glass case in the Museum, and the difference in colour between the rich and the poor specimens is readily discernible. The Bidri objects, Nos. 36 to 39, are a peculiar series of alloys. They are chiefly zinc, containing a quantity of copper and lead with smaller amounts of tin and iron. The amount of lead in No. 37 is remarkably low, and, as it is very difficult to purify zinc from lead, it would appear that a very pure ore of zinc must have been used in the preparation of this metal. Apparently copper is the predominating metal associated with zinc in this Bidri alloy. From this it is probable that No. 38, which contains none of it, is not Indian but Hungarian, as the Museum description suggests. In the other specimen, to which a doubtful origin is assigned (No. 39), copper is present but in small Eroportion, tin being the secondary metal. This indicates that it is not ndian. The remaining specimens are more of individual than collective interest. A few of the brass ones were selected as being derived from different localities, but their ages are uncertain, and they may or may Bot be typical alloys of the different districts. A greater number of specimens from each province would have to be examined before a definite decision could be given with certainty. The first thing notice- able in these analyses (Nos. 40 to 53) is the large quantity of bismuth present in some of them, and in one or two cases the extraordinary amount of antimony. The large amount of bismuth is in marked contrast to the Persian and Arabian analyses ; with one exception, none of these alloys rich in bismuth are reported as coming from any district in the Madras Presidency, most of them coming from the north-east Provinces 39 of India or Burrnah. The alloys from these latter parts are altogether much more impure than those from the neighbourhood of Madras, indicating that the ores from the Himalaya districts must be rather complex. The extraordinary quantity of antimony in the Nepal specimen (No. 40) renders it a matter of surprise that a metal with such a com- position could be worked up into the shape of the object. Iron is abundant in some of these metals, but it will be observed that it is generally low when the zinc is low or altogether absent ; so that its presence is probably due, as in the case of the Persian samples, to its presence as an impurity in the zinc ores used in making the brass. The per-centages of copper, zinc, lead, and tin, are very variable, and no deductions can be nor should be made from so few results. The analyses of the Burmese Gun (No. 53), which is nine feet long, illustrate the difficulty of the Oriental native in obtaining a homogeneous alloy for a large casting. Samples were taken from each end of this long casting and analysed separately. The analyses differ widely one from the other, and show a want of uniformity in composition. Comparing them with those of other Burmese objects, it will be noticed that whereas Burmese alloys generally are very impure this metal is remarkably free from impurity. This clearly indicates that in producing this casting great pains were taken to purify the metals, or to select pure ores, so that a metal free from prejudicial impurities might be obtained. Neither of the alloys used is a particularly suitable one for a gun, and its want of uniformity would not improve the metal for the purpose for which it is required. Modern gun-metal generally consists of 90 per cent, of copper and 10 per cent, of tin. 40 Japanese Metal-work from various Collections. No. 59. CHOPSTICK CASE. Modern Japanese. ANALYSIS. Copper A rsenic Lead Bismuth Iron Nickel Cobalt Zinc, by diif Gold Silver . 65-88 Cons, trace. 33 . Trace. 37 . 6-14 50 . 26-78 . Trace. Trace. 100-00 This sample was an electro-plated one, and was covered with a coating of silver, which was removed prior to analysis. The surface scrapings contained much silver. The colour of the above metal was nearly white. No. 60. BRONZE TORTOISE. Modern Japanese. ANALYSIS, Copper Tin Lead Iron Zinc 99-08 This sample was a very good and well-defined casting. Moderately hard and brittle. 41 No. 61. SWORD-HILT. Modern Japanese. ANALYSIS. Copper Arsenic . . . :; - Antimony Lead Bismuth Iron Nickel and Cobalt Zinc Tin 99-182 This sample was a chocolate brown colour, but was the ordinary copper colour on removal of the lacquered surface. Nos. 62 and 63. SwoRD-HiLTS. Iron, one round and one square, inlaid with Shaku-do. Modern Japanese Both from Prof. Roberts- Austen. ANALYSIS, No. 62 (round). Carbon Trace. Sulphur Silicon Phosphorus . . Manganese . . 038 130 Nil. Nil. ANALYSIS, No. 63 (square). Carbon . . Trace. Sulphur -027 Silicon 140 Phosphorus . . . . . . . . Nil. Manganese . . . . . . . . Nil. These analyses show that it is impossible that these sword-hilts could have been cast, as they are practically pure wrought iron with only traces of carbon, and therefore could not be melted at any temperature likely to have been used by the Japanese. This result was confirmed by the appearance of the etched polished surface of each, where the " lines of flow of the slag " proved that the metal had been hammered up. 42 No. 64. MONKEY'S HEAD. Specimen of modern Japanese cast metal- work. From Prof. Roberts- Austen. ANALYSIS. Lead 92'39 Antimony . . . . . . . . 7-61 by diff. The above metal is coated with a very thin layer of copper, on the surface of which the various effects are produced. The brown colour is a patina of copper. The eyes and teeth, which are of a golden colour, are not gold, but Dutch metal or copper. The back of the object is a white silver leaf, and consists of a thin coating of silver deposited on the layer of copper. 43 Metal-work from Japanese Court, South Kensington Museum. SPECIMENS of BRONZE. A collection of fifty-seven oblong plaques illustrating the different sorts of bronze, some of them with artificial patina, produced in Japan. Bought, 72. 1099 to ^.-75. Certain samples selected from the above were experimented on. The backs of the samples were carefully cleaned and scraped, and the scrapings were taken for analysis. The whole series were not analysed, but a few specimens were selected as being typical and characteristic, while two or three were chosen on account of a similarity in colour. These, upon analysis, gave results which showed that very varied colours could be produced from alloys, of which copper is the main con- stituent. No. 65. JIDAI-MEKKI-UTSUSHI. Imitation of old gilt bronze. Gold- coloured hammered surface plate. Museum No. 1099^.-75. The back was copper colour, rather red. Metal moderately soft and free from cracks. ANALYSIS. Copper Arsenic Antimony . . Tin . . Lead Bismuth Iron Nickel 1 Cobalt J Zinc Gold Silver Phosphorus Sulphur 99-63 063 Trace. Nil. 281 Trace. Oil 021 Nil. Trace. 042 No. 66. HlDO, Red Bronze. Bright, red-coloured, smooth surface plate. Museum No. 1099^. -75. The back was light copper colour. Metal rather hard and contained many cracks. ANALYSIS. Copper Arsenic Antimony . , Tin . . Lead Bismuth Iron Nickel"! Cobalt / ' ' Zinc Gold Silver Phosphorus Sulphur 97-87 Trace. Nil. 941 328 Nil. 077 043 546 Nil. Nil. 99-805 No. 67. NiGURUMlL Black Bronze. Black-coloured smooth surface plate. Museum No. 1099 2 -g.-75. The back was ordinary copper colour. Metal moderately hard, and contained a few small cracks. ANALYSIS. Copper Arsenic Antimony . . Tin.. Lead Bismuth Iron Nickel 1 Cobalt/ Zinc Gold Silver Phosphorus Sulphur 99-25 443 016 Nil. 197 Trace. Oil 016 Nil. Trace. 039 Nil. 99-972 45 No. 68. NIGURUM-NI-MATSUKAIVA. Light brown-coloured, coarsely hammered to resemble pine bark on a grey black bronze. Museum No. 1099^-75. The back was ordinary copper colour. ANALYSIS. Copper 99-37 No. 69. NIGURUM TSHUN& Stone face on a black bronze. Dark brown-coloured, rough surface plate. Museum No. 1099^.-75. The back was ordinary copper colour. ANALYSIS. Copper 99-26 No. 70. KARAKAN& Ordinary Bronze. Dark brown-coloured, smooth surface plate. Museum No. ,1099^.-75. The back was ap- parently a light-coloured copper, but on scraping appeared quite white. Metal was very soft and easily cut. Scrapings were white, but tarnished to a light brown colour. ANALYSIS. Copper Tin.. Lead Bismuth Zinc Nickel Iron 73-27 865 20-344 Nil. 5-460 060 098 100-097 No. 71. SENTOKUDO. Yellow Bronze of Shinta. Brass, yellow-coloured smooth surface plate. Museum No. 1099 -.-75. The back 8t) was a light brass colour, somewhat similar to the front surface. Metal moderately soft. Scrapings yellow. 46 ANALYSIS. Copper .. 72'32 Tin 8-126 Lead 6-217 Bismuth . . . . . . . . Trace. Zinc 13-102 (bydiff.) Nickel -065 Iron -170 100-000 From the above it will be seen that the appearance of the surface of the plates depends in a great measure on the subsequent treatment and not upon the composition of the metal. No. 65 is apparently gilt, as the existence of not more than a trace of gold in the metal prevents the possibility of the gold surface being due to the removal by solution of the other metals present. The same might be said to be the case with No. 67, where the black surface, which in some cases is known to be produced by chemical means, is here undoubtedly a surface coating only. No. 66 sample might have been coloured by oxidation, and after- wards lacquered, as the red colour is probably due to an oxidised com- pound of copper. As regards the composition of the metal in this sample the tin and zinc might have been added to harden the metal, as it is not likely that they were obtained simultaneously with the copper from the copper ore. The quantity of iron present suggests that the refined metal has been remelted at some time or other, and stirred with an iron rod probably when the tin and zinc were added. Or it may be that this metal was refined copper, to which some old bronze was added. The amount of lead present in each sample is probably due to the addition of that metal in refining the copper. The amount of arsenit; in No. 67 is large, and may possibly be due to the ore from which the copper was obtained, although the small per-centage of arsenic usually found in Japanese copper would not tend to support this. The amounts of silver present are no more than are often found in copper from other sources. As regards Nos. 68, 69, and 70, these were all of a brown colour, No. 68 being of a lighter shade, but Nos. 69 and 70 evidently the same colour ; the other differences in these plates being the indentations or otherwise of the surfaces. It may be that a particular alloy is used for No. 68, which is indented and coloured to represent the bark of a tree, but the determination of the per-centage of copper showed that the metal was only copper, con- taining the usual impurities. It was not deemed necessary to complete the analysis. The same with No. 69 sample. The sample No. 70, con- taining only 73 per cent, copper, was, however, completed, and showed an alloy of curious composition. It is probably refuse copper, from which the silver has been removed by the lead process. This sample was coloured the same tint of brown as No. 69, so that it is evident in this case that the composition of the metal has nothing to do with the colour of the surface. The Japanese apparently also do not specially select a soft metal for carving or indenting, as No. 70, which, from being soft, is well suited for indentation, is exhibited as a smooth sur- face, while the work of indentation is bestowed upon a harder metal. 47 No. 71 sample is of interest inasmuch as the colour of the lacquered surface is apparently due to the colour of the metal, and that the clean surface has simply been lacquered with a colourless lacquer. The above composition of the metal, however, is not necessary for the yellow colour, since it could be produced by copper and zinc alone. It is probably a chance alloy with a yellow colour, which yellow colour has been taken advantage of. This metal also is rather soft and suit- able for easy indentation, but here again the advantage of the properties of the alloy in this respect has not been utilised. The Japanese seemingly do not pay any attention to the composition of their metal in this work, and do not know with what they are working, except when they are using refined copper, and even then they are not particular if the surmise as regards No. 66 sample be cor- rect. From the above few results it is evident that they produce their colours regardless of the impurities in the metal, which impurities it was at one time thought were the cause of the particular colour, and that they use and work with any metal that comes before them. When, during their re-meltings, they come across a metal which in itself is coloured they use it as such, as No. 71 shows, but in all other respects seemingly, for the purposes of colouring and lacquering the composition and properties of the metal are not in the least considered. No. 72. SHIBUICHI. Greyish-white alloy. 1099 TT .-75. ANALYSIS. Silver 37'91 Lead -28 Iron '16 Tin Nil. Nickel Nil. Zinc Nil. 38-35 Analysis not completed. Remainder ... .. .. 61 '65 probably all copper. An analysis of a sword ornament of the finest quality, made by Mr. Gowland , late of the Imperial Japanese Mint at Osaka, gave the following results : Copper 67-31 Silver 32-07 Gold Traces. Iron '52 99'90 Another analysis of this alloy, by Prof. Kalischer, proved it to contain : Copper .. 51*10 Silver 48-93 Gold -12 100-15 No. 73. KODO. Yellow bronze. Copper . . Arsenic. . Antimony Tin Lead Bismuth Iron Nickel . . Zinc, by cliff. Silver .. Gold 1099 Tir .-'75. ANALYSIS. . .. .. 68-99 Small quantity, not determined. Nil. Trace. 888 Trace. 196 098 29-828 Nil. Nil. 100-000 No. 74. KODO-SOKOTSUCHIM^. Lateral hammer- work on yellow bronze. 1099.nr.-75. IT- ANALYSIS. Copper . . Arsenic . . Antimony Tin Lead Bismuth Iron Nickel .. Zinc, by diff. Silver . . Gold 69-72 Small quantity, not determined. Nil. Nil. 915 Trace. 154 Trace. 29-211 Nil. Nil. 100-000 KlODO. Sonorous bronze. Copper . . Arsenic. . Antimony Tin .. Lead . . Bismuth Iron Nickel .. Zinc, by diff. Silver .. Gold No. 75. ANALYSIS. 1099^-75. 74-19 Trace. Nil. 9-732 5-492 Trace. 434 Trace. 10-152 Nil. Nil. 100-000 Experiments on Japanese Patina, ANALYSES OF COPPERS containing small quantities of impurities used in pickling experiments : Copper and gold Copper and silver Copper and arsenic Copper and antimony - Copper and lead Copper and bismuth Copper and iron 465 Au. 402 Ag. 248 As. 254 Sb. 355 Pb. 475 Bi. 213 Fe. Pure copper used for the above contained 99 65 Cu. compared to electrotype copper of tested purity used as a standard. Difference probably due to dissolved oxide. Polished samples of these alloys when heated in the solutions (Nos. 1 and 11) given on p. 53, showed that a remarkable variation of tint is produced by the presence of impurity, but it would require coloured illustrations to demonstrate this. GOLD PATINA, Containing shining pieces of metallic copper, and cakes of a greyish purple dried powder, evidently the true patina. Weight taken - 0442 gram. Wt. after two hours at 100 C. * 0442 ignition - 0490 dissolved in dil. HN0 3 , apparently gold gold after sol. and pption. copper oxide after pption. residue insol. in aqua-regia 0033 0027 0453 0006 Gold Copper . Insol. res. Chlorine Sulphur o 66223. ANALYSIS. 6-11 81-9 1-3 4( Nil. D 50 EXPERIMENT to extract any metallic gold by means of mercury. 0389 grm. of patina was rubbed and pressed through about 1 grm. of mercury, then heated under water to about 60 C. for some time, excess patina washed away and mercury filtered thoroughly, dissolved in acid and solution tested for gold. Not a trace had been dissolved. The Japanese have a remarkable series of alloys, of which Shakudo is the most important, in which the precious metal gold and silver replace the tiu and zinc of ordinary bronze. Analysis of a sword orna- ment of the finest quality by Mr. Gowland gave : Copper 94-50 Silver .. 1 Gold 3 Lead Iron and arsenic 99-89 The amount of gold is, however, sometimes as high as 4 per cent., whilst silver is frequently only present in very small quantity* Shakudo when boiled in suitable pickles, such as No. III. of those given on p. 54, assumes a beautiful purple tint, and the following analyses have been made with a view to ascertain the composition of the purple patina. EXPERIMENT with Chlorine Water. 49 (?) gold extracted by cold dilute Cl. water. 24(?) ., hot 4-92 hot cone. *5'65 This sample of patina is very much richer in gold than the metal from which it was made, and the patina itself is no doubt richer still than shown in above analysis, as the sample there indicated contained scrapings of metallic copper. The figures marked with the query (?) are only approximate on account of the extremely small quantity weighed in each case. The patina may possibly be a sub-oxychloride of copper and gold, or more probably a sub-oxide of copper and gold containing a small proportion of chlorine as sub-oxychloride. From these figures the following deductions are made : That there may be some free gold present in the patina. That it is not proved that free gold is present. That if it be present the quantity is not more than -5 or at the outside per cent. That there is a large quantity of gold present in the combined state. 51 GOLD PATItfA, 2ND SAMPLE. This sample, unlike the previous, consisted entirely of patina, as it had peeled off the copper on which it was formed. No separate pieces of copper could be seen. Under the micro- scope the outer surface of the patina had a grey metallic lustre, with a tinge of purple, but the inner surface was copper- coloured. ANALYSIS. Gold Copper Chlorine Remainder 89-17 10*83 oxygen. No free gold could be detected. It will be seen that in this instance the gold is higher and the copper lower than before, while the chlorine (this time estimated on a larger quantity) confirms the previous result. Working out a formula from the above figures, it is remarkable how closely it comes to a mixture of Cu 2 and Au 2 0. Without concluding that the patina is always of the same compo- sition, or that there is a definite formula for it, these latter results seem to support the previously expressed opinion that the patina is a sub-oxide of copper and gold containing a small proportion of chlorine as sub-oxychloride. PICKLING SOLUTIONS USED BY THE JAPANESE. I. II. III. Verdigris 438 grains 87 grains 220 grains. Sulphate of copper 292 437 540 Nitre - 87 Common salt - 146 Sulphur 233 Water Vinegar J gallon 1 gallon 1 gallon. 5 fluid drachms. The above are taken from a paper printed in the Journal of the Society of Arts, June 13, 1890. The solutions are generally used boiling. 52 Antique Bronzes in the British Museum. No. 76. BRONZE FIGURE, with hand extended, as in Morra. Middle of 5th century B.C. Handle of a mirror. 73. 8-20. 19. ANALYSIS. Copper 89*96 Tin 7'64 Lead T44 99-04 Zinc, Iron, Silver traces. No. 77. BRONZE FIGURE OF DIONYSOS. Giseco-Roman period. 73. 8-20. 26. ANALYSIS. Copper 85-05 Tin 10-35 Lead 4-62 100-02 Iron trace. No. 78. BRONZE FIGURE OF A GLADIATOR. 3rd century A.D. 73. 8-20. 53. ANALYSIS. Copper 79-26 Tin 4-71 Lead 7*05 Zinc 6-80 97-82 Iron trace. 53 No. 79. BRONZE FIGURE OF A LION. Etruscan work. 5th century B.C. 73. 8-20. 251. ANALYSIS. Copper ,,' 82-10 Tin .. " vf f .. .. 12-64 Lead 1-86 Zinc *73 97-33 Iron trace. No. 80. BRONZE FIGURE OF APOLLO, from Orange. 1st century A.D. (?) 77. 8-10. 1. ANALYSIS. Copper 80-70 Tin 6-44 Lead 9'97 97-11 Zinc, Iron, and Silver traces. No. 81. BRONZE GROUP OF M. AURELIUS AND FAUSTINA. 2nd century A.D. 78. 3-9. 1. ANALYSIS. Copper 70-41 Lead 2'44 Zinc 26-70 99-55 Tin, Iron traces. 54 No. 82. FKAGMENT OF DRAPERY found with leg of bronze statue. Greek sculpture. About B.C. 450. '86. 3-24. 7. Copper . . Tin Lead ANA.LYSIS. 99-27 Iron trace. These seven analyses 76 to 82 were all carried out on extremely small quantities of material, only 4 or 5 grains being received in some cases as the sample. Moreover, the samples of drillings were very dirty and partly oxidised. Hence the low totals obtained. The samples, however, were as carefully and thoroughly cleaned as possible before analysis. The estimation of the metals present in small quantity was out of the question. Gaylord Bros. Makers Syracuse, N. V. PAT. JAN. 21, 1908 UNIVERSITY OF CALIFORNIA LIBRARY