The traditional Chinese iron industry and its modern fate


Figure captions and Text boxes

Figure 1. Flow diagram of traditional Chinese blast-furnace iron production.

Figure 2. Blast furnace in Xinyang, Henan, photographed ca. 1916 by E. T. Nyström. Reproduced from Tegengren (1923-24, vol. 2, plate 34).

Figure 3. Blast furnace in Xinyang, Henan, photographed ca. 1916 by E. T. Nyström. Reproduced from Tegengren (1923-24, vol. 2, plate 34).

Figure 4. Blast furnace in Jinzhai, Anhui, ca. 1958. Reproduced from Anon. (1959).

Figure 5. Plan of a `smelting house' at East Wind People's Commune in Muzidian, Macheng County, Hubei, redrawn from Anon. (1958c, p. 22). Dimensions are given in `market feet' (shichi, 33 cm).

Figure 6. Sketch of the Huang Jiguang blast furnace in Macheng, Hubei, as used in the Great Leap Forward period, and associated requisites, reproduced from Anon. (1958c, p. 22). Dimensions are given in `market inches' (shicun, 3.3 cm).

Figure 7. Vertical section of the Huang Jiguang blast furnace in Macheng, Hubei, as used in the Great Leap Forward period, redrawn from Anon. (1958c, p. 22). Dimensions are given in `market inches' (shicun, 3.3 cm). Cf. Figures 5 and 6.


Box 1. Technical details of the operation of an iron blast furnace.

Figure 8. Diagram of a modern large-scale coal-fuelled iron blast furnace, after Peacey & Davenport (1979, p. 17, fig. 2.1) and Rosenqvist (1974, p. 274, fig. 9-4). The height is typically 20-30 metres. Temperatures in deg.C are indicated for isotherms inside the furnace. The reactions which take place are indicated at the right; underlined elements are in solution in iron.

Empirical research provides considerable detail on what happens inside the modern blast furnace, and that is what will be described here. Of large charcoal blast furnaces like those traditionally used in Sichuan it is possible to say that the basic principles are the same but with the major difference that all temperatures are lower. Of small charcoal blast furnaces like those of Dabieshan it is reasonable to assume that the basic principles are approximately the same, but with the possibility of fundamental differences, of which we know, at the moment, nothing.

. . .

A modern blast furnace operates continuously for months or years at a time, with coke, ore, and flux being charged in the top, air being blown through numerous tuyères near the bottom, and molten iron and slag being tapped out of tapholes at the bottom. Operation continues until the furnace lining has been so damaged by the high temperatures that it is necessary to repair it.

The ore has been calcined (roasted) before charging, so that the iron in it is entirely in the form of Fe2O3 (ferric oxide, hematite). The fundamental reactions are the reduction of this by CO (carbon monoxide), first to Fe3O4 (ferrosoferric oxide, magnetite), then to FeO (ferrous oxide, wustite), and finally to metallic iron. Some carbon, typically ca. 4 per cent by weight, dissolves in the iron near the bottom; with this carbon content the melting point of the iron is under 1200deg.C.

The combustion of coal at the tuyères produces CO2 (carbon dioxide), and this reacts with carbon to produce the necessary CO. The CO reacts with the iron oxides to produce CO2 again, this reacts with carbon to produce CO, and so forth in a cycle. The necessary conditions for each reaction are diagrammed in Wagner (1993, p. 388, fig. 7.53) or in any textbook of engineering thermodynamics. What is important is that high temperatures and very high concentrations of CO in the furnace atmosphere are required for the reduction of FeO. The necessary concentration of CO is more readily obtained with charcoal as the fuel than with coke (because charcoal is more reactive), and therefore charcoal blast furnaces operate at lower temperatures.

. . .

Iron smelting would be easy if the ore were composed of nothing but iron oxide. In fact all ores contain significant amounts of SiO2 (silica) as well as other minerals: this unwanted material is collectively called the gangue of the ore. If the furnace is to operate continuously the gangue must be removed in molten form, but it will normally have a melting point which is much higher than the temperatures required for the reduction of iron oxides. Therefore a flux, typically CaCO3 (limestone), is charged along with the ore and the fuel. The flux is chosen to form with the gangue a slag with a practicably low melting point (less than 1400deg.C in modern practice). CaCO3 decomposes in the furnace to CaO (lime) and CO2, and the mixture of CaO and SiO2 has a much lower melting point than either mineral alone. Other minerals in the charge, either incidentally present in the gangue or intentionally added in the flux, may further depress the melting point of the slag. Especially important here is Al2O3 (alumina).

Limestone is not only an excellent flux, it also has the property that it can remove sulphur (S) from the liquid iron by the reaction shown. In modern practice, using coke with fairly high sulphur, large amounts of limestone are used, giving a CaO/SiO2 ratio in the slag as high as 1.2; in charcoal-fueled blast furnaces sulphur is rarely a problem, and much smaller amounts of limestone are used.

In some pre-modern Chinese blast furnaces no flux is used. It seems that in these cases one or more of several special conditions must hold: (1) the ore used may be very rich, i.e. contain only a small amount of gangue; (2) the ore may be `self-fluxing', the gangue containing significant proportions of limestone or alumina; (3) the charcoal may be from a wood which has grown on chalky ground and therefore contains a significant proportion of lime; (4) in operation enough of the furnace lining may be consumed to add significant amounts of alumina or other minerals to the slag; (5) the internal form of the furnace may be arranged in such a way that a small amount of the reduced iron is re-oxidised to FeO near the bottom, providing a very effective flux for silica. In pre-modern furnace operation it also appears possible to tolerate a slag which is rather viscous and does not separate well from the iron: the product is a `slaggy' iron which ironfounders, finers, and puddlers nevertheless are able to use without great problems.


Figure 9. Diagram of a fining hearth used in Shangcheng, Henan, in 1958. (Reproduced from Yang Kuan (1982, p. 225); orig. Anon. (1958b, p. 23).) 1. Hearth. 2. Tamped fireclay. 3. Hearth opening. 4. Cover. 5. Blast pipe. 6. Hearth opening. 7. Iron reinforcements. 8. Nest. 9. Ground level.

Figure 10. Drawing by E. T. Nyström of a pair of fining hearths in southern Henan, ca. 1916. In the background a traditional Chinese `windbox', in the foreground a plate of pig iron and a wrought-iron bar. Reproduced by courtesy of Tom Nyström and the Museum of Far Eastern Antiquities, Stockholm.


Box 2.

Figure 11. Sketch and sections of a water-powered blast furnace at Huangnipu in Rongjing County (modern Yingjing), Sichuan, ca. 1877 reproduced from Széchenyi (1893, p. 678, figs. 116-18)). Height 8-9 m, base 5.5-6 m.

Széchenyi describes the operation of this furnace as follows:

Hoani-pu [i.e. Huangnipu] lies in a narrow valley, and bears throughout the stamp of a typical mining district. Everything here is black with coal dust from coal mining and iron industry.

In the neighbourhood chain bridges cross over the streams. Coal and iron occur together in the immediate vicinity. On the opposite bank is a blast furnace, ca. 8-9 m high and 5.5-6 m broad at the base. In form it is quite similar to a European blast furnace; it is built of stone and held together by an external wooden construction.

The blast is provided by a [piston-]bellows 1 m in diameter and 3.5 m long. This bellows, or rather, cylinder, is constructed on the same pattern as the common Chinese kitchen blowing cylinder, except that the piston is driven by a water-wheel.

For the tapping of both ore [i.e. iron] and slag there is only one opening, at least I did not see a special opening for the latter.

The ore smelted here is an ironstone (blackband), with 40-60 per cent iron, which occurs between the coal measures. Next to the shaft of the blast furnace the ore is first mixed with charcoal and roasted. As the works stood empty and out of operation, I was unable to obtain further data on the smelting process. I give the construction of the furnace in [Figure 11].

In the storehouse of the works was a large number of cast iron slabs measuring 1 m long, 0.60 m broad, and 0.02 m thick. The surface of this cast iron is very slaggy because of the lack of a separate outlet for the slag; its fracture is steel-grey and full of blow-holes throughout. Next to the blast furnace was the foundry, which however was also out of operation. . . .

Széchenyi (1893, pp. 678ff); cf. Kreitner (1881, pp. 805-6); Tegengren (1923-24, p. 342). Dr. Katalin T. Biro of the Hungarian National Museum has kindly corrected my translation from German against the Hungarian original, Széchenyi (1890, pp. 606ff)


Figure 12.Vertical section through a blast furnace in Qijiang County, Sichuan, reproduced from Luo Mian (1936, pp. 19-21, figs. 1-3).

Figure 13. Photograph of a blast furnace somewhere in Sichuan, reproduced from Luo Mian (1936, pp. 19-21, figs. 1-3).


Box 3. Description by Luo Mian of a blast furnace in Sichuan in the 1930's. Cf. Figures 12-13.

The smelting furnace. The dimensions of the furnace are different in different places. Usually the height is 2 zhang 4 chi [8 m], mouth [diameter] 2 chi 8 cun [93 cm], [inner diameter of the] belly at the widest point [the boshes], 5 chi 8 cun [193 cm], bottom [diameter] 9 cun [30 cm].[1] The interior has the shape of a vase [tanzi] of the above dimensions, widest in the middle and diminishing toward the top and bottom, as in [Figure 12], which shows a section of the smelting furnace at the most recently built ironworks in Qijiang [County]. The mouth is a round hole into which wood, charcoal, and ore are charged and from which smoke escapes. The taphole is ca. 6 cun [20 cm] higher than the bottom; it has diameter 8 cun [27 cm, surely a typographical error for 8 fen, 2.7 cm], and is used for tapping both iron and slag.

The exterior of the furnace is square. It is narrowest at the top, ca. 1 zhang 4 chi [4.7 m], and widest at the bottom, 1 zhang 8 chi [6 m]. [Figure 13] gives an overall view. The work area [the tapping arch] looks like a city gate. . . The place where blast is blown in [the blowing arch] is at the side or the back, and has the same form as the work area [tapping arch]. The mouth [tuyère] of the pipe from the windbox is inserted into the furnace at a height of ca. 1 chi [33 cm]; it has an adjustable prop, so that the height can be varied as desired. The blast is operated either by human power or by water power. The blast is cold, though some ironworks in Qijiang have recently changed over to hot blast.

The furnaces are built of refractory sandstone, which is acidic. In the interior, above and below the tuyère, it is plastered with a mixture of salt, clay, and sand to prevent the corrosion of the sandstone by basic materials. The plaster is usually composed of ca. 1 part salt, 4 parts clay, and 5 parts sand; some use danba (a magnesium salt) mixed with salt and clay.

The sandstone is supplied seasonally by stoneworkers. The furnace as a whole can last more than 200 days, but it is never worked for more than 200 days in a year. The next year it is rebuilt. The expenditure [per year] for the furnace and tuyères is somewhat over 1,000 yuan. Assuming the worst case, a production of only 250 tonnes of pig iron per year, then the cost of the furnace is about 4-5 yuan per ton, which is not expensive; the disadvantage of this type of furnace lies rather in the quality of the pig iron produced.

Production of pig iron. After the bottom of the smelting furnace has been dried a layer of 400 jin [240 kg] of semi-charred charcoal is laid out on the bottom, followed by 600 jin [360 kg] of calcined ore. Further layers are added, alternating in the same way, until the top of the furnace is reached, giving a total of about 14,000 jin [8.4 tonnes] of ore. Then the bottom layer is ignited and the blast is started to maintain combustion. After about a day [morning to evening] the molten iron begins to flow. Outside the taphole the workers prepare a bed of fine sand which they rake to form a mould for a flat plate. The taphole is opened by striking with an iron rod, and iron and slag flow out together. As they flow into the mould the rake is used to remove the slag; the iron left behind in the sand mould, when it has cooled, is the cast-iron plate [the pig]; it is ca. 2 chi 5-6 cun [ca. 85 cm] long, 1 chi 4-5 cun [ca. 50 cm] wide, and weighs ca. 100 jin [60 kg].[2] After the first day tapping can be done from time to time as convenient. The quantity of iron tapped is highly dependent on whether the `heat' [reli, i.e. temperature] is sufficient; therefore it is important not to neglect the `heat'. After the first tapping, fuel and calcined ore continue to be charged alternately as described above. As a general rule more than 4,000 jin [2.4 tonnes] of cast iron can be produced from 12,000-13,000 jin [7.2-7.8 tonnes] of calcined ore. By custom more than 5,000 jin is considered excellent, 4,000 jin normal, and 3,000 jin inferior. [These quantities are 3, 2.4, and 1.8 tonnes respectively, presumably per day.]

Luo Mian (1936, pp. 18-21)


Figure 14. Internal form of two traditional blast furnaces in Sichuan in 1940, redrawn from Wang Ziyou (1940, foldout fig. 1). Dimensions are given in mm. On the left is the `older' form, on the right the `newer' form. Cf. Figures 11-12.

Figure 15. Diagram of a puddling hearth in Sichuan, reproduced from Luo Mian (1936, p. 24, fig. 4). a. Firebox, height 2 chi 8 cun (93 cm). b. Puddling bed, diameter 2 chi (67 cm). c. Blast pipe. d. Flame channel.

Figure 16. Puddling furnace used in Sichuan in 1958, redrawn from Yang Kuan (1960, p. 187, fig. 69). The dimensions are not given. Mineral coal or charcoal is burned in the large chamber; blast is blown in at the left. The flame travels through the long passage to the puddling bed at the right.

Figure 17. Stall furnace for crucible iron smelting in southern Shanxi, photograph reproduced from Kocher (1921, p. 10, fig. 1).

Figure 18. Lumps of cast iron and puddled wrought iron in southern Shanxi, photograph reproduced from Kocher (1921, p. 10, fig. 3).

Figure 19. Crucible smelting of iron in progress in Gaoping County, Shanxi, photograph reproduced from Shockley (1904, fig. 1, facing p. 854).

Figure 20. Operation of a furnace in Shanxi for converting cast iron to wrought iron, photograph reproduced from Alley (1961a).

Figures 21-28. The first eight of twelve watercolours in an album preserved in the Bibliothèque Nationale, Paris (`Fer', C.E. Oe 119 in-4to, 1-8).

To see Figures 2128 in colour, click here.


21. `Exploring the soil' (tan tu).Colour


22. `Testing the iron [ore]' (yan tie). Colour


23. `Extracting the iron [ore]' (qu tie). Colour


24. `Separating the lead' (fen qian). This picture clearly shows the calcining of iron ore; the idea that lead is somehow involved is curious. Colour


25. `Purifying the dirt' (qing ni). The `dirt' is the calcined ore.Colour


26. `Removing the fire' (chu huo). This probably shows the slag from the blast furnace being disposed of. The water-colour has been reproduced in colour by Hauser (1974, p. 21). Colour


27. `Tipping out the iron' (dao tie). Colour


28. `Making bars' (zuo tiao).Colour


Box 4. Translation of a passage on iron-smelting by Qu Dajun (1630-96) in Guangdong xinyu, `New discourses on the province of Guangdong', (1700 ed., ch. 15, pp. 7b-10a; cf. 1974 ed., pp. 408-10).

There is no better iron than Guangdong iron. In the iron-producing mountains of Guangdong, wherever there is yellow water seeping out, one knows there is iron. Digging there, one will find a large body of iron ore in the form of an ox; this is the `iron ox'. If one follows the path of the underground water and digs deeply, more iron will be obtained.[3]

However, of the mountains which produce iron, it is only on those which are forested that one can operate a furnace. If the mountain is bare, even if there is a great deal of iron it will be of no use. This is why `iron mountains' are not easy to find.

Splitting a body of iron ore layer by layer, one finds that each [layer] has a tree-leaf pattern. This differs on the two sides. If the mountain has a certain type of tree, then that tree's leaf pattern will be found in its iron ore. Even if one digs as deep as several tens of zhang [several times 32 metres], the same phenomenon is found. When it is extremely cold in south China, the leaves do not [normally] fall from the trees; it is only on mountains which produce iron that the leaves fall, and these are absorbed by the essence of the iron. This is an example of the Way of `metal conquering wood'.

The iron ore has a spirit, and to this the furnace-master must sacrifice devoutly before he dares to operate a furnace.

The furnace has the shape of a vase [ping] with its mouth upward. The breadth at the mouth is about a zhang [3.2 m]. The base is 3 zhang 5 chi [11.2 m] thick [sic!], and the height is half of that [5.6 m].[4] The thickness of the body is slightly more than 2 chi [64 cm]. It is built of ashes, sand, salt, and vinegar.[5] It is bound about with thick cane and braced with wood of the tielimu and the zijingmu.[6] It is also built against a mountainside for greater solidity.

At the back of the furnace is an opening, and outside the opening is an earthen wall. The wall has two `doors' [blower fans], 5-6 chi high [1.6-1.9 m] and 4 chi broad [1.3 m]. Four persons operate these `doors', `closing' and `opening' [pushing and pulling] alternately in order to produce the force of the blast.

The two openings [the blast-hole and the tap-hole] are lined with `water stone' [shuishi, perhaps diatomite]. `Water stone' is produced at Dajiang Mountain in Dongan District [modern Yunfu County, Guangdong]. Its substance is not hard, and not being hard it does not `receive the fire'. Not receiving the fire, it can endure long without altering; hence the name `water stone'.

Furnace operation begins in the autumn and ends in the spring. Because the weather is cold the iron contains a great deal of water; metal is the source of water, and water flourishes in the winter, so that molten iron is engendered by cold.

When the iron ore is charged [xia] [into the furnace] it is mixed with `hard charcoal' [jian tan].[7] Usually a mechanical device [ji che] is used to cast down [the furnace charge] from the mountain into the furnace.

The flames [from the furnace] light up the sky, and its dirty black smoke [qi] does not disperse for several tens of li [1 li = ca. 0.6 km].

When the iron ore has `melted' it flows out into a rectangular mould and solidifies into an iron slab. In order to obtain it, the furnace is `stirred' [? jiao, `struck'?] with a wooden pole [removing the clay which blocks the taphole], so that the molten iron flows out and is cast into another slab.

In twelve hours [24 modern hours] one slab should be produced each hour, with a weight of about 10 jun [180 kg]. If two slabs are produced per hour, this is called a `doubled cycle' [shuang gou]; then the furnace is excessively vigorous [wàng =], and in danger of damage. One must anoint the furnace with the blood of a white dog; then no accident will occur.

Among the Five Metals, Iron is the one which corresponds to Water; it is called the Black Metal [hei jin], and is formed by the Essence of Great Yin [tai yin zhi jing]. Its spirit is a woman. According to tradition the wife of a certain Mr. Lin, when her husband was in arrears in his official iron quota, threw herself into the furnace in order to make it produce more iron. Today those who operate the furnaces always sacrifice to her, calling her `Madame Gushing-Iron' [Yong tie furen]. Her story is bizarre in the extreme.

An ironworks generally has 300 families living around it, [providing services as follows:] furnace tenders, more than 200; miners, more than 300; water-carriers[8] and charcoal-producers, more than 200; pack animals, 200 oxen; freighters, 50 vessels. The expense [i.e. capital investment] of an ironworks totals not less than 10,000 jin.[9] If it produces more than 20 slabs per day it is profitable, while eight or nine slabs [per day] is unprofitable; this is a fundamental principle.

Of all the smelters it is the iron of Datangji Furnace in Luoding which is best. All of it is `first grade iron' [kai tie], glossy and soft, which can be drawn into wire. When it is cast into woks [huo][10] it is also good and `hard' [jian]. It is more expensive than the first-quality [iron] of any of the other smelters.

After smelting at any of the smelters, all of the iron is shipped to the port city of Foshan, where the people are excellent founders. Of the woks which they produce, the largest are called `tangwei ' [`sugar pot'?], `deep-seven', `deep-six', `ox-one', and `ox-two'. The smallest are called `ox-three', `ox-four', and `ox-five'. If there are five to a bundle they are called `five-kou ', and if there are three to a bundle they are called `three-kou'. Those without handles are called `oxen'. The best are called `pure'.

In former times those who cast [woks] without handles were not permitted to cast them with handles, and those who cast them without handles were not permitted to cast them with handles. Those who cast both were indicted.

After [a wok] is cast it is coated with yellow clay and pork fat [to prevent rust].

When [a wok] is struck with a light stick, if it [sounds] like wood then it is good. It sounds like wood because its substance is `hard' [jian]. Thus the woks of Foshan are expensive, because they are `hard', and those of Shiwan [in modern Boluo County, Guangdong] are cheap, because they are brittle.[11] When they are sold [as far away as] central China, people can distinguish them by their thinness and smoothness; the foundrywork is refined, and the craftsmanship is accomplished. Foundry products are generally better in Foshan, while ceramic [products are better] in Shiwan.

In iron-fining [chaotie], cast iron is kneaded [tuan] and charged into the furnace [i.e. the fining hearth].[12] It is fired until it glows red,[13] then taken out and put on the anvil. One person holds it with tongs and two or three hammer it. At the side ten or more youths [tongzi] fan it [i.e. work the bellows for the blast in the fining hearth]. The youths always sing a chanty without stopping. After this [the iron] can be converted into wrought iron and made into thin plates.

[In Foshan?] there are several tens of iron-fining plants [employing] several thousand persons. Each plant has several tens of anvils and at each anvil there are more than ten persons.

This [the fining hearth] is the `small furnace'. Furnaces are large or small according to whether the iron [produced] is cast or wrought. The treatment [zhi] of cast iron is a matter for a `large furnace' [a blast furnace], while the treatment [zhi][14] of wrought iron [i.e. the conversion of cast iron to wrought iron] is a matter for a small furnace.

The excellence of the strength [jian][15] of steel lies in its quench-hardening [cui]. Before it has been quench-hardened, its softness remains. In quench-hardening, the steel is first hammered [into the desired shape] at the forge, then removed from the fire and put into water. A great fire is needed to soften it, and pure water is needed to strengthen it and form pure steel. This is the refining of steel [lian gang].

Ganquan[16] has said: Observe the casting of metal in a great furnace, and you will know the end and beginning of Heaven and Earth. When [the metal] is melted in the furnace, this is its birth. When it leaves the furnace and solidifies, this is its accomplishment. Melting is a matter of the watery beginning, and solidification is a matter of the earthy end. Men consider melting to be subjugation, but do not realise that it is the beginning of birth; what could be more enduring. Men consider solidification to be enduring, but do not realise that it is the end of accomplishment; what could be more of a subjugation? `Beginnings and ends alternate, subjugation and endurance interact',[17] but the metal never changes; this is an image of the Way.


Figure 29. One possible reconstruction of an early Qing blast furnace investigated at Luxia Village in Luoding County, Guangdong, redrawn from Cao Tengfei & Li Caiyao (1985, p. 72, fig. 5). Dimensions are given in centimetres. The figure does not correspond well to the description in the text of the article, according to which the entire north wall of the furnace above the taphole is missing, the remaining height of the south wall is 271 cm, and the original height was 6-8 m. The measurements shown here for the mouth of the furnace are those of the remaining upper part of the furnace as excavated.


[1]Clearly the furnace diagrammed in Figure 11 does not have the same proportions as these dimensions.

[2]The thickness of the plate would then be ca. 1.8 cm.

[3]The importance of underground water in the search for ores is also mentioned by the 16th-century writers Biringuccio and Agricola (Biringuccio (1942, p. 15); Agricola (1912, p. 116)).

[4]Di hou san zhang wu chi, chong ban zhi.

[5]Presumably it is built of clay with these ingredients added.

[6]Mesua ferrea and Madhuca subquincuncialis, `common mesua' and `peanut madhuca'.

[7]This does not refer to mineral coal. Charcoal varies greatly in its mechanical properties; in a tall blast furnace it was essential that the charcoal be hard and strong so that it could support the weight of a high column of furnace burden.

[8]Jizhe, perhaps to be understood more generally as `fetchers and carriers'.

[9]In this type of context one jin usually means one liang (ca. 37 grammes) of silver, so that the sum indicated, if it is to be taken literally, amounts to 370 kg of silver.

[10]The Cantonese pronunciation of this word gives us the English word wok for the round-bottomed pan which is a fundamental implement in every Chinese kitchen. The more usual north Chinese word is guo.

[11]The point of this passage is that the best woks will be of grey rather than white cast iron. White cast iron is very hard and can ring like a bell, while grey cast iron is soft and contains microscopic flakes of graphite which deaden vibration. Both are brittle (for very different reasons), but white cast iron is considerably more brittle than grey. The term jian normally means `hard', and is so translated here, but it might be better to take it as meaning `strong' or `robust' in a vague sense. In modern technical terms, the mechanical property which is relevant here is toughness.

[12]I suspect that the author has misunderstood his source here, for in the fining process an action very like `kneading' of the cast iron takes place in the fining hearth rather than before it is charged.

[13]We should expect the iron to be white-hot rather than red-hot when it is taken from the furnace.

[14]Perhaps the two occurrences of zhi, `to order, govern' in this sentence are scribal errors for ye, `to smelt', though in the second case this would be an incorrect use of the word.

[15]Here the relevant mechanical property is in fact hardness.

[16]Possibly the philosopher and educator Zhan Ruoshui (1466-1560), whose literary name was Ganquan. He came from Zengcheng in Guangdong, and wrote a number of books which, judging from their titles, might well contain the passage quoted by Qu Dajun here. See Goodrich & Fang (1976, pp. 36-42). Or conceivably the person referred to might be Qu Dajun's son Qu Minghong, whose alternate name was Ganquan; see Hummel (1944, p. 202). The quotation from Ganquan probably continues to the end of the passage translated here.

[17]Qu shen xiang gan, a quotation from the Xici commentary of the Book of Changes. Shisan jing zhushu, ch. 8, p. 87c; Legge (1882, p. 389).