Mr. J. Napier on Copper Smelting. 199 



provides himself with three bottles or glass tubes of equal di- 

 mensions, and a measure graduated into 10 or ]00 divisions; a 

 common alkalimeter answers the purpose : 2 grains of pure 

 copper are weighed off and dissolved in nitromuriatic acid, then 

 excess of ammonia added. This solution is made up to two 

 measures exactly of the graduated glass or alkalimeter. Two of 

 the bottles or tubes are tilled with the solution, and well-stopped 

 and placed upon a frame. To test an ore by this method, 20 

 or more grains are digested in nitromuriatic acid, as previously 

 described, ammonia added and then filtered; the blue solution 

 is now tried by filling with it the remaining third tube, or bottle, 

 and placing it between the other two in the frame and comparing 

 the colour ; it is diluted carefully until the tint is exactly the 

 same as the tubes outside ; the whole is then measured in the 

 alkalimeter, every measure of which being equal to one grain of 

 copper, the per-centage is soon ascertained. A little practice in 

 this method gives tolerably accurate results with expedition ; 

 however, they are generally too high to serve as a guide in pur- 

 chasing the ore. By making up test-bottles of different strengths, 

 such as ^, \, \, Jg- of a grain, the process is well adapted for test- 

 ing slag and refuse of furnaces or other products, and should 

 form a part of the operations in every assaying-room or smelting- 

 works, a subject we may have occasion to refer to again. 



Similar methods have been recommended, but in which, 

 instead of judgmg by the colour, a solution of another salt is used 

 that destroys the colour, such as the sulphurets of sodium and 

 potassium ; solutions of these salts of such strength are employed, 

 that a given measure is equivalent to a given weight of copper. Mr. 

 H. Parkes has recommended a solution of cyanide of potassium as 

 follows: — "Take a given quantity of pure copper (say, for instance, 

 10 grains), place it in a tlask, and dissolve in nitric acid; add 

 ammonia in excess, and then make it into a bulk of about 2500 

 grains by measure by the addition of water, although this is not 

 absolutely necessary. Dissolve 1 oz. (avoirdupois) pm'e cyanide 

 of potassium, free from ferrocyanide or sulphuret of potassium, 

 in 5 oz. by measure of water ; filter, if necessary, and place the 

 solution in a well-stoppered bottle till required for use. I then 

 ascertain the quantity of this solution of cyanide of potassium 

 required to decolorize the solution of copper, by taking a given 

 quantity in any graduated vessel, as a burette, and pour it by 

 degrees into the solution of copper, adding the last quantity drop 

 by drop till decolorized. This is very easily perceived, as there is 

 no precipitate to interfere ; and the operation is conducted at the 

 ordinary atmospheric temperature. Mark down the quantity 

 required (say 500 grains) by volume. After having established 

 these data, it is very easy to estimate the quantity of copper con- 



