"The constructor, having dug out in the side of the hill an opening circular, or rather semi-circular, at the base, with a diameter nearly three times as wide as the future furnace, arranged in the centre of this hole a kind of furnace-bed made of plastic clay at bottom, and covered with a layer of fire-proof clay on the top of it. The bed of the furnace, which lies on the natural and hardly levelled earth, is, generally speaking, not so thick as the side walls, which are formed of sandy or siliceous clay, always fire-proof on the inside, but sometimes of a more plastic nature on the exterior; the empty space left between the walls of the furnace and the solid ground round it was filled up with earth and other material. In front the furnace was enclosed by a rough wall, sometimes straight and sometimes curving, built, without mortar, of rough limestone, and dressed with earth to fill up the gaps. In front of the furnace an opening was made in this wall, taking its rise a few inches above the bottom of the furnace, and increasing in size in an outward direction, so as to enable the workmen to see into, and work in, the furnace.
"The work thus commenced was carried up to the requisite height; and when the excavation in the side of the hill was not lofty enough, the dome of the furnace was raised by placing buttresses against the fire-clay, so as to prevent the earth falling in. When these furnaces were established on almost level ground, as is sometimes the case, they form a truncated cone, with a base varying in size according to the height of the apparatus.
"The furnace was not always built upright; it often deviated from the perpendicular, leaning to one side or the other to an extent as considerable as its own diameter, but no constant rule as to this can be recognised. The internal shape was just as irregular, changing from circular to oval, without any apparent motive beyond want of care in the workman. The crucibles or furnaces are sometimes larger at the top than at the bottom, and sometimes these proportions are reversed, but always with extreme irregularity. We have noticed some which at a point 10 or 12 inches above the crucible were perceptibly contracted on three sides, thus representing the first rudiments of the appearance of our modern furnaces. But this, perhaps, was nothing but a caprice on the part of the builder.
"The furnace thus being established, the wood was withdrawn which had formed the cone, if, indeed, any had been used, and at the hole made at the base of the crucible a clod of fire-clay some inches in height was placed, so as to form a dam, and to confine in the crucible the molten or soft metal; the scoriæ, being of a lighter nature and floating at the top, made their escape over the top of the dam. As the latter were not very liquid, their issue was promoted by means of pokers or wooden poles, perhaps damped, with which also the metal was stirred in the crucible.
"In neither of these two kinds of furnaces do we find any trace of bellows, and a more or less strong draught must have been procured through the opening made for the escape of the scoriæ, according to the elevation of the dome of the furnace. The limestones which have been found in certain furnaces were probably employed with a view of increasing the draught; they doubtless belonged to the upper part of the furnace, where they had been fixed so as to add height to the orifice. This rudimentary plan must have been likewise used in the earliest crucibles. The mode of obtaining a draught which we have just pointed out is indicated most plainly by the scorification of the walls of the furnace on the side opposite to the air-passage; this side has evidently experienced a more intense heat, whilst on the other the walls are much less affected by the fire, and in some cases pieces of the mineral still remain in a pasty or semi-molten state, just as they were when the work of the furnace ceased....
"The absence of any machine in the shape of bellows in the ancient metal works of the Jura appears all the more remarkable as these implements were known both to the Greeks and Romans; hence we may at least infer, not only that these nations did not introduce the art of iron-working into the Jura, but that it must have existed at a much earlier period. It must also be remarked that the openings in the furnaces are not placed in the direction of the winds prevailing in the country—a plan which might have increased the draught—but are made quite at hazard, just as the nature of the spot rendered the construction of the furnace more easy.
" ... In respect to fuel it must be remarked that in all the siderurgical establishments which we have discovered, certain features indicate that wood carbonised in a stack was exclusively used as fuel. The furnaces are too small for the employment of rough wood; added to this, charcoal stores are placed near the furnaces; and charcoal burnt in a stack is constantly met with all round the sites, in the scoriæ, and all the débris. We must, besides, mention the discovery, at Bellelay, of a charcoal store 8 feet in diameter, situated under a compact bed of peat 20 feet in thickness. It was established on the solid earth, anterior to the formation of the peat. Now from this very peat a parcel of coins belonging to the fifteenth century was recovered, over which only 2 feet of peat had grown in a period of 400 years. There, too, at a depth of 9 feet, were found the scattered bones of a horse, with the foot still shod with those undulating edged shoes with elongated and strongly punched holes, in which were fitted the ends of nails of the shape of a T, the heads of which were conical. This kind of shoe is found in the Celtic settlements, the villages, habitations, and ironworks, also in the pasturages and forests of the country, but rarely in the Roman camps; in the latter they are always in less number than the wider metallic shoes, which are larger, and furnished with a groove indicating the line in which the nail-holes were punched. The calculations which have been made from the discovery of the coins of the fifteenth century (A.D. 1478) would give an antiquity of at least twenty to twenty-four centuries to the horse-shoe we have just mentioned, for the animal must have died and been devoured on the then existing surface of the ground, and could not have been buried in the peat, as the bones, instead of lying grouped together, were dispersed in every direction. These same calculations would carry back the date of the charcoal-store to an era 4000 years ago.
"Owing to the imperfection of the furnaces, the quantity of charcoal used must have been quadruple the present consumption for the same results. The metal, as it was extracted from the ore, fell down into the bottom of the crucible. In proportion as the mass of metal increased, a workman, with a poker made of damp green wood, brought out the scoriæ which floated on the top, and stirred the metal so as to fine it. It is proved that these wooden pokers or poles were made use of in all the furnace-works. A quantity of morsels of scoriæ is found which, having been in a soft state when extracted, have retained the imprint of the piece of wood, the end of which was evidently charred. M. Morlot, in his article on the Roman ironworks at Wocheim, in Upper Carniola, has also noticed the existence, in the scoriæ, of frequent traces of pokers, sometimes round and sometimes three-cornered in shape, but all of them must have been made of iron, whilst throughout the whole of the Jura we have never recognised the traces of any but wooden implements of this kind.
"Owing to the imperfection of the furnaces, and especially, the deficiency in the draught caused by the want of bellows, the metal contained in the ore could be but very imperfectly extracted; the scoriæ are therefore still so very rich in iron that, about twenty years ago, the manager of the ironworks at Untervelier tried to use them over again as ore. Accumulations of this dross, measuring from 100 to 200 yards square, may be seen near certain furnaces-a fact which would infer a somewhat considerable production of iron. The examination of these scoriæ proves that iron was then made by one single operation, and not liquid pigs fit for casting, or to be converted into iron by a second series of operations.
"The iron produced was introduced into commerce in large blocks, shaped like two quadrangular pyramids joined at the base, weighing from 12 to 16 lbs. One of these pieces was found near a furnace which had been demolished in order to establish a charcoal furnace, in the commune of Untervelier, and another in one of the furnaces of Boécourt.