Donald B. Wagner
Passing through Hong Kong in January 1996 I had a chance to talk with Bill Meacham, Chairman of the Hong Kong Archaeological Society, about some fascinating features they have excavated.
In salvage excavations in 1990 at Ha Law Wan, on the island of Chek Lap Kok in Hong Kong, thirteen kilns have been excavated. These are dome-shaped, typically 75 cm high and 2 m in diameter, with several small-diameter flues and and an opening for charging at one side with a height of 40 cm. Each kiln was constructed by digging a cavity into the side of a ravine and plastering inside with clay. The clay is now baked hard, and small amounts of charcoal have been found in their vicinity. Two radiocarbon dates for the charcoal calibrate to periods of AD 1255-1340 and 1265-1405. (In the illustration, D.G. means 'decomposed granite'.)
What the kilns were used for is a mystery. There is none of the sort of debris
which would be expected from glass or pottery production. The only clue is some
'grey stuff', gravel- to pebble-sized aggregates, sintered or melted at a
considerable temperature (perhaps 1000deg. C?). In the excavation report this
was called slag, but as Bill told me, this was never a very probable
identification. Not enough has been found (only a few buckets-full), and
chemically it does not resemble any industrial slag that either of us has heard
of. Analyses of two samples of grey stuff are as follows:
% SiO2 9.61 22.23 Al2O3 1.75 5.08 TiO2 0.10 0.22 Fe2O3 91.88 72.07 MgO 0.28 0.45 CaO 0.48 0.96 Na2O 0.00 0.00 K2O 0.38 1.28 MnO 0.60 0.61 P2O5 0.15 0.18 L.O.I. -5.55 -3.26 Total 99.51 99.71
The negative L.O.I. (loss on ignition) is accounted for by the oxidation of Fe3O4 to Fe2O3.
Bill has suggested that the grey stuff was the intended product of the kiln rather than a by-product, and that, since this would obviously make an excellent iron ore, the kilns may have been used for processing iron ore in some way preparatory to smelting. I think this is likely to be correct, and I have further suggested that the kilns were specifically used for pelletising ironsand.
Grains of iron oxide (mostly magnetite, Fe3O4) are found in small quantities in most sand produced by the erosion of granite. Since magnetite is much heavier than quartz (the principal component of the sand), the action of flowing water tends to concentrate the magnetite grains in certain places as 'ironsand' with 10% or more iron oxide. Various sluicing or panning methods can easily increase this concentration to as high as 95%. (These methods are discussed in my book Dabieshan, ch. 1 and 3.) Where ironsand is available it can be an excellent raw material for iron smelting, but its fine grain size means that certain technical problems must be overcome: it may clog the furnace burden, or it may blow out of the furnace mouth. Various 19th- and 20th-century traditional Chinese blast furnaces (such as those described in Dabieshan) were specially designed to avoid these problems. In Japan the peculiar tatara furnace (described e.g. in Archeomaterials, 1989, 3.1: 11-25) is believed by many to have been developed originally to deal with the problems of smelting ironsand in a bloomery. And in 1773 the American metallurgist Henry Horne (in his Essays concerning iron and steel) described the difficulties of smelting ironsand in a bloomery.
A modern blast furnace operator who wished to smelt ironsand would pelletise it: that is, sinter the grains together into larger chunks by heating to a high temperature (800-1000deg. C?). I think something like this may have been done in the kilns at Ha Law Wan. A spring-fed stream ran out of the mountains here, and local people told the archaeologists that this was one of the few dependable streams of the island, flowing at all times of the year, every year without fail. It is easy to imagine a major industry here whose activities included sluicing ironsand, cutting firewood in the mountains, making charcoal, and pelletising the ironsand in these kilns. The 'grey stuff', pelletised ironsand, would then have been transported somewhere else, to be smelted in a blast furnace. A reason why smelting did not take place on the spot might be the additional pressure which this would have put on the fuel supply.
It is not possible to test this hypothesis on the ground, for most of the island of Chek Lap Kok has been flattened for the construction of the new Hong Kong airport: the granite mountains have simply been blasted away, except for a small hill left 'for landscaping purposes'. The kilns survive at this place, but almost their entire context is gone. (The kilns are backfilled, so I was unable to see them, but they have been carefully preserved.)
The analysis of the grey stuff does resemble that of concentrated ironsand. There are no samples extant of ironsand from the island, but that of nearby islands may be expected to be similar. On my last day in Hong Kong I went out to a beach on Lamma Island and could see that the sand does contain a few black grains which are probably magnetite. It will be necessary to concentrate these by sluicing and then do an analysis which can be compared with that of the grey stuff. Trace elements, particularly titanium, will be of special importance.
A more popular brochure, Archaeological discovery at Chek Lap Kok (H.K.A.S., 1994), is also available at US$5.
The other books mentioned here are my Dabieshan: Traditional Chinese iron-production techniques practised in southern Henan in the twentieth century (Scandinavian Institute of Asian Studies monograph series, no. 52, London & Malmö: Curzon Press, 1985) and Henry Horne's Essays concerning iron and steel (London: T. Cadell, 1773).