242 BOOK VII. 



wdth ore in the scoriiier and afterward again separated from it in the cupel ; 

 the other, by which it is first melted in the triangular earthen crucible and 

 afterward mixed with lead in the scorifier, and later separated from it in the 

 cupel. Now let us consider which is more suitable for each ore, or, if neither 

 is suitable, by what other method in one way or another we can assay it. 



We justly begin with a gold ore, which we assay by both methods, for 

 if it is rich and seems not to be strongly resistant to fire, but to liquefy easily, 

 one centumpondium of it (known to us as the lesser weights),^' together with 

 one and a half, or two unciae of lead of the larger weights, are mixed together 

 and placed in the scorifier, and the two are heated in the fire until they are 

 well mixed. But since such an ore sometimes resists melting, add a Httle 

 salt to it, either sal torref actus or sal artificiosus, for this will subdue it, and 

 prevent the alloy from collecting much dross ; stir it frequently with an iron 

 rod, in order that the lead may flow around the gold on every side, and absorb 

 it and cast out the waste. When this has been done, take out the aUoy and 

 cleanse it of slag ; then place it in the cupel and heat it until it exhales all 

 the lead, and a bead of gold settles in the bottom. 



If the gold ore is seen not to be easily melted in the fire, roast it and 

 extinguish it with brine. Do this again and again, for the more often you 

 roast it and extinguish it, the more easily the ore can be crushed fine, and the 

 more quickly does it melt in the fire and give up whatever dross it possesses. 



2'This method of proportionate weights for assay charges is simpler than the 

 modern English " assay ton," both because of the use of loo units in the standard of 

 weight (the centumpondium), and because of the lack of complication between the 

 Avoirdupois and Troy scales. For instance, an ore containing a libra of silver to the 

 cenhmipondiutn would contain i/iooth part, and the same ratio would obtain, no matter 

 what the actual weight of a ceniumpojidium of the " lesser weight " might be. To follow 

 the matter still further, an uncia being i /i,200 of a centumpondium, if the ore ran one 

 " uncia of the lesser weight " to the " centumpondium of the lesser weight," it would also run 

 one actual uncia to the actual centumpondium ; it being a matter of indifference what 

 might be the actual weight of the centumpondium upon which the scale of lesser weights 

 is based. In fact Agricola's statement (p. 261) indicates that it weighed an actual drachma. 

 We have, in some places, interpolated the expressions " lesser " and " greater " weights 

 for clarity. 



This is not the first mention of this scheme of lesser weights, as it appears in the 

 Probierbiichlein (1500 ? see Appendix B) and Biringuccio {1540). For a more complete dis- 

 cussion of weights and measures see Appendix C. For convenience, we repeat here the Roman 

 scale, although, as will be seen in the Appendix, Agricola used the Latin terms in many 

 places merely as nomenclature equivalents of the old German scale. 



Troy Ozs. dwts. gr. 



Grains. per short ton. 



I Siliqua . . . . . . 2.87 Per Centumpondium . . 03 9 



6 Siliquae = I Scripulum . . 17.2 ,, ,, ..106 



4 Scripula = i Sextula . . 68.7 ,, ,, ..410 



6 Sexiulae = i Uncia . . 412.2 ,, „ . . 24 6 2 



12 Unciae = i Libra . . 4946.4 „ „ . . 291 13 8 



100 Librae = i Centumpondium 494640.0 



However Agricola may occasionally use 

 16 Unciae = 1 Libra .. 6592.0 (?) 



100 Librae = i Centumpondium 659200.0 (?) 



Also Oz. dwts. gr. 



per short ton. 



1 Scripulum . . . . . . 17.2 Per Centumpondium . . 1 06 



3 Scripula = i Drachma . . 51.5 ,, ,, . . 3 19 



2 Drachmae = i Sicilicus .. 103.0 ,, ,, . . 6 i 15 



4 Sicilici = I Uncia . . 412.2 „ ,, . . 24 6 12 

 8 Unciae = i Bes . . .. 3297.6 „ „ .. 194 12 



