TJtlE KOYAL AllTiLLERY INSTITUTION. 1 ; U 
called because specks of graphite mottle the surface of the white iron. 
There is a great difference in the appearance of pig-iron made at different 
works; even the pig-iron from the same firm may vary much, though 
the number of pig may be the same. Any variation in the blast, fuel, 
fee., in the furnace will tell on the iron,* and even out of the same 
running there may be a difference in the pigs. In all cases the “ sow ” will be 
found to differ in appearance from the pigs which she feeds. Being larger, 
the metal cools more gradually, favouring both the formation of large 
crystals and the separation of graphite. For the manufacture of shells, it is 
desirable to have a dense, close-grained iron; not too hard to turn, as it 
would then injure the tools and largely increase the cost of manufacture, as 
the rate of driving the turning-lathes would have to be lowered. It will be 
seen in the manufacture that a cut of nearly *1 in. is taken from the exterior 
of the shell. It is necessary to go to that depth, as the exterior of the shell 
is very hard, and would injure the tools. It is probable that turning takes 
off more than would be imagined from the strength of the casting. It is 
known by experiment that a cast-iron girder not planed down is stronger 
than one which is reduced to the same dimensions by means of planing. In 
fact, taking off the outer skin must, to some extent, diminish the strength of 
a casting, as the soundest and most dense part of the iron is removed.! If 
a broken shell of some considerable thickness is examined, the fracture will 
clearly exhibit the closer structure of the exterior of the shell, and also of the 
interior. The walls cooling first on the outer and inner surfaces, the central 
part is liquid while the walls are set, and the contraction on cooling causes 
the central part to be more open in the grain, as it cannot receive any fresh 
iron to make up for the contraction; the walls, on the other hand, can feed 
themselves at the expense of the still liquid centre. Hence the diffi¬ 
culty of getting a very thick casting sound. This also shows how thickening 
a part of a casting, especially if there is a sudden alteration in form, may 
actually weaken instead of strengthen a shell. Instances may now and then 
occur where the thick bottom of a common shell is quite spongy, owing to 
the quicker cooling surfaces feeding themselves at the expense of the base of 
the shell. 
We cannot trust to the numbers of the pigs to choose the fit iron 
for shells. The pigs are broken up and selected, the viewer being able 
to tell by experience what kind to take as suitable for the shell, and what to 
put on one side for re-melting in company with other iron so as to bring it 
to the required standard. Scrap iron— i.e., iron produced from breaking up 
old shells—is also melted up and run into pigs for the purpose of selection, 
and then mixed in the proper proportion for the shells required.]: Be- 
melting is found to harden iron. This seems to confirm the idea that a very 
high temperature may produce soft iron, as the temperature in there-heating 
furnaces is not generally so high as in the smelting furnaces. 
The metal is melted in cupola furnaces, of which there are eight in the 
* The position of the tuyeres also affects the iron. 
f Some trials have been made of casting shells to their final dimensions, and the result was good. 
Some expense would be caused by the greater accuracy necessary in moulding. The whole question 
is a large one, and much consideration has been and still is given to it in the Department, 
t Shells made of iron of known quality are not run into pigs, hut are placed at once in the shell 
furnaces. 
18 
