[Illustration: SLAG-POUNDING HOLLOWS, GOBBETTS.] The book of Agricola, published in 1556, shows that this primitive method was still in practice so late as the middle of the sixteenth century. But this clumsy method could not be long practised on Dartmoor, where fuel--except peat--was scarce; and it gave way to a furnace of better construction, where the receiver was circular, and a draught-hole was at the bottom. One of these has been dug out and carefully examined at Deep Swincombe. [Illustration: SMELTING ORE. (_After Agricola._)] [Illustration: PLAN OF BLOWING-HOUSE, DEEP SWINCOMBE.] It consists of a single chamber, 18 feet by 11 feet, rudely constructed of masses of granite resting on one another by their own weight and unset in mortar or in clay. The entrance was narrow and low. On one side was the furnace, constructed of granite, one slab set upright to form a side, and the back and other side built up rudely. A fragment of the receptacle for the molten tin was found, with a receiver and channel cut in it. Pottery was also found, which was of a very early description. It was submitted to the late Sir Wollaston Franks, of the British Museum, who said that he would have attributed it to the Celtic period but for the bold scores made at the starting-point of a handle, which are characteristic of Anglo-Saxon pottery. At the extremity furthest from the door was a _cache_ in the thickness of the wall, formed something like a kistvaen, as a place in which to store the metal and tools. The whole structure was banked up with rubble and turf. Outside to the south still lies a mould-stone, a slab of elvan, in which the mould had been cut, measuring 26 inches long by 12 inches at one end and 15 at the other, and 5 inches deep. That this is the earliest tin-furnace yet discovered on Dartmoor admits of no doubt. The curious mould-stone is quite different in shape from any others found on the moor. No mortar-stones were discovered, and this also is a token of antiquity. The earliest smelting arrangements must have been very crude, and much tin was left in the slag. Until recently the Malays threw away their slags, which contained as much as 40 per cent. of tin. As there have been no mortar-stones found at Deep Swincombe, it is to be presumed that the tinners disregarded their slags. These have not, moreover, been found. The reason was this--the sets had been reworked at a later time by the tinners at Gobbetts, further down the river. These later men had stone mortars and a crazing mill, and finding these rich slags, removed them, pounded them up in the hollowed mortar-stones, that may be seen _in situ_ at Gobbetts, and re-smelted them. Deep Swincombe has all the appearance of having been much pulled about by tinners since the first furnace was erected. [Illustration: TIN-MOULD, DEEP SWINCOMBE.]