﻿Segregation in Ores and Mattes. 181 



From the vertical central lines it will be seen that the 

 curves of copper and nickel are nearly reciprocal. The 

 horizontal central line shows a similar tendency. From 

 these examples, which are given merely as an illustration 

 of what has been proved true by numerous other analyses, 

 the following statements may be inferred : 



1. In a mass of molten copper-nickel iron matte, the 

 sulphides of copper and nickel are mutually antagonistic. 



2. The tendency of nickel and, though in less degree, 

 that of iron is also to concentrate toward the centre, with 

 a slight downward inclination. 



3. The tendency of copper is to disperse toward the 

 outside and to rise toward the top of the casting. 



These statements are verified in a striking manner 

 by furnace practice. The matte as it flows from the blast 

 furnace passes first to a forehearth in which it accu- 

 mulates and the slag rising to the surface is separated. 

 In this forehearth, where the matte is subjected to a 

 prolonged heat, nickel tends to sink to the bottom, and as 

 after every tapping there remains a layer of matte perhaps 

 two or three inches thick in the forehearth below the 

 tapping ring, this matte becomes gradually enriched in 

 nickel and impoverished in copper. On changing the 

 forehearth after several weeks running the bottom is 

 found coated with a tough magnetic matte which averages 

 about 46 per cent, nickel and 12 per cent, copper. The 

 matte made in this forehearth has during the run 

 averaged perhaps 22 per cent, copper and 18 to 20 per 

 cent, nickel. This shows that under prolonged heating 

 the copper nickel sulphides are more perfectly separated 

 the copper going upward and the nickel downward. 



Again, the Orford process^ of separating copper from 

 nickel consists in smelting matte with sodium sulphide 



> Mineral Industry, 1892, vol. i., p. 357. 



