It oil. 
tained in assaying approached not to the colour and purity 
of line glass, there remained a portion of the iron still un- 
revived, in the state of a fused ozyde, conveying colour 
and opacity to the mass ; that at certain degrees of co- 
lour, certain degrees of opacity existed, and proportionate 
quantities of the metal remained diffused in the scoria* 
When the colour of this was green, the quantity of metal 
united was small ; but as the green deepened, and be- 
came associated with light browns, the quantity of unre- 
vived iron was greater, and became much increased as the 
scoria assumed darker browns, or became totally black : 
in such instances I have found it contain, upon being re- 
assayed, 12 per cent. The assaying of iron ores is sus- 
ceptible of another extreme, by which experiment is 
equally clogged, and wherein it is very difficult to obtain 
an accurate result. In this case no perfect button of me- 
tal is found, but the portion of iron which the ore contained 
is in a vast variety of various sized globules of the richest 
crude iron, interspersed in, or covering the surface of a 
semi-vitrified opaque mass of scoria, of a greyish-blue, 
mottled, or whitish colour. The causes of which, and 
their strict analogy to similar results in the blast-furnace, 
shall be my chief object to explain. 
In a former paper I mentioned, that if iron-stones were 
smelted without the addition of any other substance, the 
product in iron would be proportioned to the quantity of 
lii ne contained in the respective classes. We shall see 
from the following experiments how far this is confirmed 
by means of the assav -furnace, and furnish to ourselves 
one important lesson, how far the various mixtures of 
ores affect the operation of smelting ? 
1st, Into a crucible, with a v eil fitted cover, I intro- 
duced 875 grains of a pulverised siliceous iron-stone in its 
raw mva- ? and applied such degree of heat as is usually 
given in such operations. In twelve minutes I found 
