S5003154

166

1. la che first place there ara smali, so-called random slag pj_t# .

are scattered among Che concemporary hablcacion sltes ln the settle-mencs of the tlme. These are the remains of smali ahaft furnaces with an interna 1 dlameter of about 20 cm.

2. The second kind of metallurgical site is represented by large smelting establishments. These were deliberately sited away from the llving areas. The remains consist of slag pita arranged ln parallel rows with a gap ln the middle of each row (Fig. 1). The numbers of furnaces on each side of the gap are usually three, four or five and the largest site so far surveyed ln the Holy Cross Mountains contalned 123 bloomery furnaces.

The bloomery furnace used ln the area was of the immersed type, that is to say its lower part was dug into the soli to a depth of 40-50 cm (Fig. 2). This part formed a so-called slag pit which served as a recepta-cle for the slag which formed durlng the smelting process. Above the slag pit there was a shaft, probably about 1 m high, with blast holes ir> its lower part. The furnace was charged through the opening at the top of the shaft. The reduction sonę was on a level with the blast holes and the reduced nodules of iron collected here, forming iron sponge. In the process the metal did not pass through a fluid State but formed a piece of malleable iron by a solid-state reaction, while the gangue and some of the unreduced ore formed a slag which dripped down through the charge into the pit at the bottom of the furnace. This pit gradually filled with slag and when its upper level reached the reduction zonę, the smelting process was over.

Afterwards the upper part of the shaft was dismantled so as to reoove the metallic sponge, and the lump of solidified slag ln the pit was left in situ. For the next smelting another furnace was built. As a result of this technology in which the furnace site was only used once, the lumps of slag remained to be found today, where they bear witness to the number of smelting operations carried out in the past.

As the excavations raised a number of technological problems which requlred further study, a series of experimental iron smeltings has been carried out sińce 1960. At the very outset these were rather typical laboratory experiments. At a later stage it was necessary to transfer this research work to the region of the Holy Cross Mountains and in a place called Nova Słupia, which was an ancient iron smelting site, the natural conditions for our experiments could be provided.

The ore for iron smelting was the hematite ore from the nearby Rudki minę. It should be mentioned here that the minę had been operating at the same time as the iron smelting centre. There are remains of ancient raining such as drlfts and passages to be found in the modern minę. Wood samples taken for carbon-14 analysis from the remains of the timber lining of the passages put their age between the second and the fourth century of our era. Gossan, that is to say the iron ore in the oxidised zonę, was mined here in antiquity down to a depth of twenty flve metres. These deposits of hematite have the consistency of creas and are mixed with dolomite, siderite and pyrlte. For our experiments we have managed to obtain a certain amount of hematite with the consistency of cream. The ore analysis is givem in the table.

The cheml cal composition of the ore used for experlaental smelting

concentration X ore    - ■ _

	F«tot	^Fe2°3	S102	CaO MgO	MnO P205	S	AljOj
hematite raw	52.33	72.77	15.50	0.55 tr	0.85 0.06	1.06	7.77
hematite roasted						0.22
hematite floated						0.05

Table^;

Sulphur content was high and amounted to 1.06Z. Before smelting the ore was subjected to a roasting process Just as it had been in ancient tlmes.

Next to the slag pits the archaeologlsts excavated traces of orlglnal ancient roasters which had been detected by the magnetic surveying nethod. The ore found in the roasters was maghemite. As regards the charcoal nsed for these snelting experiments, sampling of the hearths under the slag blocks has established that a mixed charcoal was used consisting of some charcoal obtained from hard woods like beech and oak, and some from conlferous trees like pine, fir and larch. Charcoal has also been identifled from poplar and linę trees.

Experimental reductions showed that the optimum ratlo of beech to pine was 1:2, and these proportions were maintained throughout the experiments so as not to introduce another variable factor into the sneltings. The charcoal for the experiments was obtained from our own kllns and the size of the pieces used was regulated by sieving to give lumps of about six cubic centimetres in volume.

The first experiments were carried out using an induced draught, without the aid of bellows, in shafts which were forty-flve centimetres high. The charge solidified into a błock, but did not produce any iron (Fig. 2).    ; o

In the next phase of the experiments the height of the shaft was increased to 100-110 cm and the result of this was that the smelting could be carried out with an induced natural blast from four holes in the side of the furnace (Fig. 3).

In the third phase of the experiments an artificial draught was used with a pair of bellows working continuously. This produced a sponge of iron and the slag ran down into the slag pit underneath the furnace.

The furnace for the experimental smeltings was nade of clay. mixed with chaff and grass. The thickness of the wali was about ten centimetres. Before the experiments can start, the furnaces have to be air~dried for several days and then heated. For the first few hours the furnace is heated by burning wood which is then followed by charcoal. The duration of the heating had a direct impact on the whole course of our experimeats. Our observations have led us to the conclusion that a twelve hour pre-heating results in proper functioning of the furnace. The heated furnace was then