248 ANNUAL OF SCIENTIFIC DISCOVERY. 



metals. This product is now roasted, by being put into a reverberatory fur- 

 nace furnished with air holes, and kept at a semi-fluid state, with a free current 

 of air passing over the surface. The reaction may be thus explained : a por- 

 tion of sulphur is carried off by the oxygen of the air, and the copper is 

 oxidated, and this oxide of copper instantly reacts upon, or is decomposed by, 

 another portion of sub-sulphuret, the copper of both being reduced to the 

 metallic state without any carbonaceous matter. Copper in the fused state 

 has a stronger attraction for sulphur than any of the other metals, so that 

 when copper begins to be reduced, it will first reduce all the other sulphurets 

 present, except iron. Therefore, by carrying on this roasting until about the 

 half of the copper is reduced, and then tapping the furnace, this reduced por- 

 tion will contain all, or mostly all, the less pure metals which had existed in 

 the regulus. The sub-sulphuret remaining is selected and reduced by itself 

 in a separate furnace, to make pure or select copper for yellow metal. Thus, 

 the process of selecting affected the whole copper trade, and particularly the 

 sheathing, for the yellow metal was not only a competing article with ordinary 

 sheathing, but its production almost necessitated the deterioration of that 

 against which it was to compete. The reduced copper with the impurities 

 was taken and subjected to long roasting and refining ; if the quality after that 

 would bear rolling it was used up for sheets, if not, it was sold as tile. 



The copper trade is now almost entirely relieved from these circumstances, 

 by the abundant supply of Australian ores, which are mostly all pure, giving 

 copper of the best quality ; however, as far as regards the past, and the ques- 

 tion under discussion, these circumstances all tend to show that the cause of 

 the deterioration of sheathing is impurity in the copper. 



However, the Australian ores having relieved the trade from the great pres- 

 sure of impurities and the drain of the purest metal, there should now be a 

 great improvement in the quality of the sheathing, Australian ores having 

 been in use in great quantity these last ten years ; still we have not heard of 

 any marked improvement upon the copper sheathing since then. 



Looking back upon the whole discussion, it will be seen that the subject is 

 involved in difficulties. In the first place, we have seen it demonstrated by 

 every experimenter and by experience, that pure copper dissolves more rapidly 

 in sea water than copper with a little alloy in it. From our experiments it is 

 also shown that pure copper, in contact with copper containing impurity, is 

 acted upon more rapidly by sea water than when alone, so that, under any 

 circumstances, this would go to prove that pure copper is not the best for 

 sheathing, as it would be subject to the greatest amount of wear, from chemical 

 action ; while, in the early history of sheathing, the probabilities are in favor 

 of the copper having been very pure, and the sheathing then lasted much 

 longer than it is said to do now ; and yet, when the circumstances came that 

 necessitated an impure copper being more general, the rapid waste of sheath- 

 ing instantly became apparent fact and deductions seeming contradictory. 



We had ourselves thought that purity was all that was required for a good 

 sheathing metal, and all the sheets being of the same quality ; but observations 

 and experiments have tended to modify these views. Nevertheless, we are 

 still convinced that a promiscuous impurity of copper will make a bad sheath- 



