270 
between the rolls, prevents a perfect weld, and renders the 
iron unsound. In the manufacture of steel, what is termed 
unsound iron blisters very much in the converting furnace. 
This is probably because the carbon, penetrating into the 
pores of the iron when it meets with the protoxide, forms 
carbonic acid gas, and thus blows up the blister. On the 
table is a piece of Swedish bar steel, exemplifying a very 
large blister. 
Clearness, the absence of greys or small cavities. — These 
cavities are possibly also occasioned by particles of protoxide 
of iron which have been volatilized at a high temperature. 
This is the more probable as scrap iron, although very tough, 
which I attribute to the interlacing of its fibres, is very full 
of greys ; and in making very tough iron, required to be free 
from greys, the scraps are scoured bright by a mechanical 
process previous to being worked ; thus cleaning off the pro- 
toxide, and the red or per-oxide. Gun barrel iron is required 
to turn clear in the interior, otherwise the friction against 
these small cavities rapidly destroys the evenness of surface. 
Toughness This quality in a great degree appears to 
depend on the mechanical arrangement of the molecules of 
iron, because a bar of granular iron of good quality, if 
further drawn out, will become fibrous. Moreover, the most 
fibrous iron will lose its fibre under circumstances which 
appear to favour a re-arrangement of the molecules; for 
instance, the iron being subject to too great a degree of heat, 
or burned, as it is termed, or as I said previously, being cold 
swaged, that is, gently hammered from a yellow heat until it 
becomes only lukewarm. It might be a curious question 
how far the quantity of latent caloric may vary in various 
irons, and how far their properties are modified thereby. It 
is now an admitted fact that fibrous iron may become 
granular under the influence of vibration, and, therefore, I 
think it is presumable that toughness depends on the mole- 
