oftlie Argillaceous Iron- Ore. 253 



hiin in the compounding of his flux, so that he shall select sub- 

 stances of that kind, and in that quantity which are respec- 

 tively the fittest to produce the fusion of the whole mass in the 

 furnace. Nor is it possible to devise any formula for a flux of 

 universal application, oi|e which can be considered as, in all 

 cases, a sure test of the quantity of metal contained in any ore. 

 Our ores are, generally speaking, either siliceous, or calcareous, 

 or argillaceous. Now a siliceous or sandy ore requires a flux 

 in which pure lime is a predominating ingredient. A calca- 

 reous ore, on the contrary, naturally containing a considerable 

 amount of the carbonate of lime, requires a highly siliceous 

 flux. And an argillaceous ore, in order to solve its clay, will 

 need either lime or a mixture of lime and silica. Indeed, some 

 mixtures of silica and lime in an iron-ore may make it neces- 

 sary to add a quantity of pure clay to the flux, before they will 

 liquefy in the furnace. But since all these various kinds of ore 

 often occur in the same coal-field, unless the composition of the 

 flux be made to follow their alterations with corresponding 

 changes on itself, it is plain that the smelting or assaying fur- 

 nace will seldom or never make a fair return of metal. The 

 only certain method, therefore, of trying the real value in metal 

 of any given ore, is to obtain some information respecting its con- 

 stitution, and then to compound a flux expressly adapted to it. 

 It is true, however, that there do exist various powerful sol- 

 vents which will hardly fail to fuse any iron-ore in the furnace, 

 and which, if judiciously compounded, may often be useful, 

 especially when the smelter is left very much in the dark as to 

 the constitution of any particular ore. This indeed is the only 

 situation in which it becomes desirable to possess some uni- 

 form vitrifying material, which may be employed in assaying 

 all kinds of ores, however different from one another in consti- 

 tution. We would recommend for this purpose a mixture of 

 from 15 to 20 parts of silica and 25 parts of quicklime (un- 

 slacked,) to every 100 parts of the (uncalcined) ironstone. 

 This will be found upon trial to be of very general application, 

 and will seldom fail to separate the ore into a metallic button 

 and a scoria : and although the results of the assay will not 

 possess the accuracy of those obtained with a flux selected ac- 

 cording to more systematic principles, it will still afford a to- 

 lerable approximation. When the ore is unusually refractory, 



VOL. VIII. NO. II. APRIL 1828. R 



