562 BOOK XII. 



vats. If when tested it tastes very salty, and at the same time acrid, it is 

 good ; but, if not, then it is condemned, and it must be made to percolate 

 again through the same material or through a fresh lot. Even two or three 

 waters may be made to percolate through the same earth and become full 

 of saltpetre, but the solutions thus obtained must not be mixed together 

 unless all have the same taste, which rarely or never happens. The first of 

 these solutions is poured into the first vat, the next into the second, the third 

 into the third vat ; the second and third solutions are used instead of plain 

 water to percolate through fresh material ; the first solution is made in 

 this manner from both the second and third. As soon as there is an abun- 

 dance of this solution it is poured into the rectangular copper caldron and 

 evaporated to one half by boiling ; then it is transferred into a vat covered 

 with a lid, in which the earthy matter settles to the bottom. When the 

 solution is clear it is poured back into the same pan, or into another, and 

 re-boiled. When it bubbles and forms a scum, in order that it should 

 not run over and that it may be greatly purified, there is poured into it three 

 or four pounds of lye, made from three parts of oak or similar ash and one of 

 unslaked lime. But in the water, prior to its being poured in, is dissolved rock- 

 alum, in the proportion of one hundred and twenty librae of the former to five 



to understand. In this case the saltpetre is melted and the sulphur added and set alight. 

 Such an addition to saltpetre would no doubt burn brilliantly. The potassium sulphate 

 formed would possibly settle to the bottom, and if the " greasy matter " were simply 

 organic impurities, they might be burned off. This method of refining appears to have 

 been copied from Biringuccio (x., i), who states it in almost identical terms. 



HISTORICAL NOTE. As mentioned in Note 6 above, it is quite possible that 

 the Ancients did include efflorescence of walls under nitrum ; but, so far as we are aware, 

 no specific mention of such an occurrence of nitrum is given, and, as stated before, there 

 is every reason to believe that all the substances under that term were soda and potash. 

 Especially the frequent mention of the preparation of nitrum by way of burning, argues 

 strongly against saltpetre being included, as they would hardly have failed to notice the 

 decrepitation. Argument has been put forward that Greek fire contained saltpetre, 'but it 

 amounts to nothing more than argument, for in those receipts preserved, no salt of any kind 

 is mentioned. It is most likely that the leprosy of house-walls of the Mosaic code (Leviticus 

 xiv., 34 to 53) was saltpetre efflorescence. The drastic treatment by way of destruction of 

 such " unclean " walls and houses, however, is sufficient evidence that this salt was not used. 

 The first certain mention of saltpetre (sal petrae) is in Geber. As stated before, the date 

 of this work is uncertain ; in any event it was probably as early as the I3th Century. He 

 describes the making of " solvative water " with alum and saltpetre, so there can be no 

 doubt as to the substance (see Note on p. 460, on nitric acid). There is also a work by a 

 nebulous Marcus Graecus, where the word sal petrosum is used. And it appears that Roger 

 Bacon (died 1294) and Albertus Magnus (died 1280) both had access to that work. Bacon 

 uses the term sal petrae frequently enough, and was the first to describe gunpowder (De 

 Mirabili Potestate Artis et Naturae 1242). He gives no mention of the method of making his 

 sal petrae. Agricola uses throughout the Latin text the term halinitrum, a word he appears 

 to have coined himself. However, he gives its German equivalent in the Interpretatio as 

 salpeter. The only previous description of the method of making saltpetre, of which we are 

 aware, is that of Biringuccio (1540), who mentions the boiling of the excrescences from walls, 

 and also says a good deal about boiling solutions from " nitrous " earth, which may or may not 

 be of " plantation " origin. He also gives this same method of refining with sulphur. In 

 any event, this statement by Agricola is the first clear and complete description of the salt- 

 petre " plantations." Saltpetre was in great demand in the Middle Ages for the manufacture 

 of gunpowder, and the first record of that substance and of explosive weapons necessarily 

 involves the knowledge of saltpetre. However, authentic mention of such weapons only 

 begins early in the I4th Century. Among the earliest is an authority to the Council of Twelve 

 at Florence to appoint persons to make cannon, etc., (1326), references to cannon in the 

 stores of the Tower of London, 1388, &c. 



