COP

COP

( 32 8 )

high heat in contaCt with charcoal or carbon-aceous matter of any kind, and fecluded fromthe contact of air. In reducing the metals vo-latile in heat, fuch as zinc or antimony, this re-duction requires fomewhat of a different appa-ratus from the fixed metals, to which chieflythe pvefent obfervations apply. It has been aconftant cuftom with chemifts in almoft allcafes (till of late years) to add to the ore in re-duction, not only a due quantity of carbon-aceous matter, but alfo to mix it with a largeproportion (feldom lefs than thrice its weight)of fome alkaline or eafily verifiable matter toferve as a flux to promote the fufion of all theheterogeneous contents of the ore, and to afforda thin-flowing medium, through which the glo-bules of reduced metal can readily fall by theirfuperior gravity, and at laft be collected in aSingle mafs or button at the bottom of thecrucible.

In all the common proceffes of reduction,therefore, which are ufually given, the ad-ditions to the ore are of two kinds, the purpofeof which fhould be well diftinguifhed, namely,the carbonaceous matter, which is effential tothe difoxygenation of the metallic oxyd, and thefaline or fufible flux, the ufe of which is oftenhighly convenient and even neceflary, but alfooften needlefs or even detrimental. The in-variable fault of faline fluxes as generally ap-plied, is, that they always diffolve a portion ofthe metallic oxyd before it has time to pafs to themetallic ffate, and retain it permanently, thusrobbing the metallic, button of a part of whatotherwife would unite with it: and hence italmoft invariably happens that common affaysmade in the dry way with faline fluxes return alefs proportion of metal than the ore reallycontains. Nor is this lofs trifling in many in-ftances, fince the accurate Klaproth found adifference of 9 per cent, between the yield ofcopper from a certain ore allayed in the dry way,and the fame ore treated in a different manner,in the moift way. As a proof of the folution ofa part of the metallic oxyd in the faline flux, itmay be added that the fcoria, or flux after melt-ing, is always found deeply coloured with thatprecife tint which would be given, as in thecommon preparation of coloured glaffes, by thedefigned admixture of the oxyd of the metalufed. Thus the fcorite of reduced cobalt oreare of a deep blue; of fine copper, green orbrown according to circumftances; of iron.grafs-green, and the like. The lofs of metal is Rillgreater when the ore retains a portion of ful-

phur, which laft uniting with the alkali, a£t*with ftill more power on the metallic oxyds.

As naked alkaline fluxes are peculiarly liableto this inconvenience, fome eminent docimafticchemifts employ with conliderable fuccefs fluxesof equal power, but with lefs difpofition todiffolve the metallic oxyds. Of this kind arevery fufible glaffes made of much alkali andfilex without litharge or other metallic admix-ture, fluor fpar, lime, and above all, borax. Thislaft produces a thinner fufion of vitrifying mix-tures than an equal quantity of any other fub-ftance whatever, and hence a fmaller dofe ofthis than of the naked alkalies, will anfwer allthe purpofes of a flux, and of courfe the lofsby folution of the metallic oxyd is much lefs.

A reduction of the fulphuretted ores is fome-times conveniently effected by a Angle opera-tion which gives a button of metal of confider-able purity. The roafting, which is always atedious operation, is faved by mixing the orewith two or three times its weight of nitre, andprojecting it into a hot crucijrle. The mixturedephlagrates on reaching the crucible, the ful-phur is burnt and converted to fulphuric acidwhich unites with the potafh of the decompofednitre, and the metal, now freed from tire ful-phur, becomes fully oxygenated by the nitricacid and ready for reduction. This is effectedby throwing in a reducing flux of tartar andpitch or any fimilar matter, and applying aftrong heat for the requifite time.

It appears, however, that though nitre isoften a good and expeditious method of freeingores from their fulphur, it is better not to per-form the whole reduction in a Angle operation,but firft to feparate the metallic oxyd afterdephlagration (which may be done by wafhing)and then to reduce it by a proper flux.

The redudtion of the pure oxyds or carbon-ated oxyds of. copper, whether natural or arti-ficial, is effected without lofs, and in by farthe belt method, fimply by heating intenfely incontact with charcoal in a covered crucible.This is in fact a very dofe imitation of the me,thod of reduction in the large way of theroafted copper ores, only as thefe latter ftillcontain fulphur, arfenic, iron, and other foreignmatters, the procefs requires to be repeatedmany times before the copper is in a puremalleable Rate, in each of which, the impuritiesfeparate in the form of a thick pally i'cum, asalready deicribed, under the article of reductionin the large way.

In final! experiments,, where faline fluxes arc