ASSAYING. 121 



variable composition, and no mechanical method of mixing is sufficient to 

 bring about the degree of homogeneity required. The smaller the relation 

 of the silver in the rock under assay to the silver in the litharge the greater 

 is the uncertainty arising from the argentiferous character of the flux, and 

 unless the litharge is nearly pure it is impossible to discriminate between 

 errors arising from this cause and those due to insufficiency in the time of 

 melting, imperfect fluidity of the slag, and the like. 



Quantity of reducing material necessary. — The -weight of the lead button is depend- 

 ent upon the amount of reducing material used in the flux and the amount 

 of sesquioxides present in the rock, which it is necessary to reduce to pro- 

 toxides, provided reducing gases be excluded. As there were no other 

 sesquioxides than that of iron present, and this only in very small quanti- 

 ties, the weight of the lead button was not materially altered in that way. 

 For reasons which will be given hereafter, it was found advisable not to 

 use sufficient reducing matter to exclude all the oxide of lead from the slag. 

 Relation of the silver to the amount of lead reduced. — Experiments made with different 

 quantities of reducing material upon the same flux showed that part of the 

 silver in the litharge went into the button of metallic lead, while a part of 

 it remained in the unreduced litharge in the slag. As might naturally be 

 supposed, the proportion of silver to lead in the reduced button was always 

 greater than in the litharge employed. The proportion of silver to lead 

 also increased with the time during which the flux was kept melted and 

 varied with the temperature and perhaps with other circumstances in such 

 a manner that no law governing the proportions in which the two metals 

 were reduced could be detected. When rich litharge, or litharge contain- 

 ing very appreciable amounts of silver, was used, it was therefore impossi- 

 ble to estimate with any sufficient degree of accuracy the amount of silver 

 from the litharge which is united with that of the rock in the lead button. 

 Even when the whole of the litharge is reduced, or as nearly as possible 

 reduced, it is not likely that all the silver it contained is concentrated in the 

 lead button, and it is only by using litharge (or any suitable form of lead) 

 which contains little or no silver that it is possible to render the resulting 

 error small enough to permit of estimating the probable amount of silver 

 which the litharge gives up to the lead button. 



