extensive scale* I have constantly considered the assay- 
furnace as capable of affording conclusions applicable to 
the operations of the smelting-furnace, and that change 
or innovation should always have the concurring testimo* 
ny of truth to back them, though, on a small scale, be- 
fore they be risqued on one more momentous or extensive* 
Influenced by such motives, I early rejected, as totally 
inapplicable to the scale of manufacture^ the numerous 
tribe of salts, alkalies, and earths : these, in the applica- 
tion, are subject to no rule, nor guided by any immediate 
object of general utility, but are as arbitrary as their au- 
thors are numerous* On the contrary, I have directed 
; my endeavours to the use of such agents as effect separa- 
tion upon the large scale, and have been fortunate to find 
that the same solvents, when properly applied, are produc- 
tive of the most perfect and finished results* By the sim- 
ple application of lime or chalk, in various proportion s 5 
as a calcareous earth, and common bottle glass, in the 
place of silex, to constitute fusibility, I have been able to 
produce in the assay-furnace, all the various qualities of 
crude iron, as to strength and fusibility* In no case has 
the result of any assay been considered perfect, unless the 
vitrid mass found upon the surface of the metallic button 
exhibited a degree of transparency and purity of colour little 
inferior to flint glass, or slightly darkened by a faint shade 
of azure* In such vitrifications, purity of colour is the 
surest proof of the non-existence of iron in the state of a 
fused oxyde : the same degree of pellucidity renders it easy 
to detect the smallest globule of metal which by chance 
may have been suspended during fusion* In all experi- 
ments where a just association of mixtures has been pre- 
sent to produce this peculiar scoria, the quality of the iron 
will be found richly car bonated, and the button possessing 
a smooth, silvery, greasy-feel surface. On the other hand., 
experience has repeatedly shewn, that when the scoria ob-* 
