488 De sulphur at ion of Metals . 
the contact of the coals upon the undecomposed sulphate, is one 
of the principal causes of the desulphuration produced. 
We have had occasion to speak of several kinds of furnaces, 
(the Fahlim and Scotch furnaces among the rest,) in which the 
metallic suiphurets will undergo a real roasting ; but there are 
others where this effect is scarcely perceptible. I consider the 
present as a proper opportunity for introducing some reflections 
upon the differences they exhibit in this respect. They ought to 
excite the more interest, because they are intimately connected 
with the present subject, and explain some phenomena which 
cannot be accounted for from the way in which the operation pi' 
roasting has generally been regarded. 
It is a well-known fact, in foundries, that the highest furnaces 
are those which admit of desulphuration the least, or, in the lan- 
guage of the workmen, they produce the most mattes. If a con- 
vincing proof of this is wanted, it will be sufficient to mention, 
that at Pezey there have been seen roasted ores of lead, contain- 
ing a great deal of sulphate of lead, the flux of which, in the 
Scotch furnace, gave no mattes , and yet they produce a great 
quantity when they are passed to the common furnace. 
If heat alone decomposes easily and completely the metallic 
suiphurets, the upper part of the high furnaces will be very pro- 
per for operating the roasting of ores ; for besides the tempera- 
ture being a little elevated, the air which ascends to that height, 
being deprived of a part of its oxygen, forms very little more of 
these sulphates which are opposed to the separation of the sul- 
phur : but it is quite different, and in my eyes it is a new proof of 
the little effect of the action of caloric alone upon substances. The 
sulphur is separated from the suiphurets, as we have seen, in the 
state of sulphurous acid, and oxygen is indispensable to its forma- 
tion. In furnaces not much raised, the air which touches the ore 
recently thrown in, still contains a great deal of oxygen ; the sul- 
phurous acid formed is soon subjected to the deoxydating action 
of the coals ; if there be a small portion of it decomposed, a new 
sulphuret is formed, which is afterwards roasted like the mineral. 
In the Scotch furnace, for example, when mattes are melted, they 
are thrown successively into the furnace, raid what has escaped one 
operati«ji is decomposed by a second. In high furnaces, on the 
contrary, the ores placed in the upper part undergo but a very im- 
perfect desulphuration, because the air which comes in contact 
with it contains but very little free oxygen, the sulphurous add 
