Iron. 
147 
More suddenly, in a degree somewhat proportionate to 
the more sudden contraction of the pipe, a considerable 
quantity never enters the furnace, but, striking on the ex- 
terior wall, is thence repelled. 
A discharging-pipe constructed as in Fig. 3. would 
obviate, in a great measure, the imperfections of the two 
former : the length of the tapered piece is 12 inches, of 
the straight pipe, 6 inches ; extreme diameter as in the 
others, 5 inches ; diameter of straight pipe, 3 inches* 
From such a pipe it is conceived that the blast will pro- 
ceed to the greatest possible distance unimpaired in com» 
pression and velocity. So far, therefore, as the absolute 
force of the blast and breadth of the furnace will permit, 
decomposition will be prevented on the level of the pipe, 
and the manufacturer freed from the evils which I have 
above detailed, as attendant upon decomposition in that 
quarter.” (Mushet.) 
For drawings of the ground plan, sections and elevations 
of a water-blast I refer to Mr. Mushet’s paper, 6 Phil. Mag* 
362, to Roebuck’s paper in the same volume p. 324, co- 
pied into Rees’s Encyclop. articles Blast and Blow. 
Let it then be remembered, that among the points sug- 
gested thus far as to the smelting of iron, are the follow- 
lnga 
1st, That as the principal object is to procure metallic 
iron out of the ore, the first step is to get rid of all stony 
heterogeneous matters, by sorting the ore when necessary, 
by roasting, by washing. 
2dly, That the pieces of ore prepared for the furnace, 
should not much vary, or much exceed the size of an egg. 
3dly, That roasting, in contact with charcoal dust, has 
the following effects : it drives off sulphureous and arse* 
nical particles, if there be any : it drives off superfluous 
moisture ; it drives off carbonic acid gas by causing the 
Bxygen of the ore to unite partially with the charcoal, and 
