According to Vitruvius, the works for manufacturing vermilion from Spanish ore in Rome were situated between the temple of Flora and Quirino. The ore was dried and treated in furnaces, to remove the native mercury it contained, and was then ground in iron mortars and washed. In addition, small quantities of quicksilver and vermilion were made at Almaden. The ancients describe other methods, among which Theophrastus speaks of using vinegar, which, however, appears from modern investigations to have been an erroneous account. Nothing definite is known concerning the methods of the Moors; we possess only as a proof that they produced mercury, an account of a quicksilver fountain in the marvelous palace of Abderrahman III., at Medina-Zahara, and the works of Rasis, an Arab. The Moors probably extracted mercury at Almaden, from the eighth to the twelfth century, by the use of furnaces called "xabecas," which latter, in the fourteenth century, were still employed by the Christians, who continued them till the seventeenth century, when German workmen replaced them by "reverberatory" furnaces, which in turn were superseded in 1646 by aludel or Bustamente furnaces. There is an anonymous description of the working with xabecas as practiced at Almaden in 1543, and later accounts in 1557 and 1565. The ore was put into egg-shaped vessels with a lid, the mineral being covered over with ashes. The vessels were packed in a furnace heated with wood, about 60 pounds being used per pound of quicksilver made. This system was also applied at the Guancavelica mines, discovered in Peru in 1566, where the xabecas were abandoned in 1633, being replaced by the furnaces invented by Lope Saavedra Barba, which there were called "busconiles," while in Spain they were named Bustamente furnaces, and elsewhere aludel furnaces. They were introduced at Almaden thirteen years after their first use in Peru by Juan Alfonso de Bustamente, Barba and his son having been lost at sea on their way to the Peninsula. In 1876, there were at Almaden, at the works at Buitrones, twenty such aludel furnaces and two Idria furnaces. D. Luis de la Escosura y Morrogh, from whose work we take the above notes, has followed the historical details of the growth of Almaden closely, and from his account of the method of working in 1878 we take some data: