581 
Cast IroUy Steel, and Malleable Iron, 
skeleton grains will attract each other and become united. 
The attractive forces of these grains for carbon are developed 
only when they are at a white heat, and it is an error, which 
has remained even in the last edition of Turner’s Chemistry, 
that iron will weld at a red heat. When the grains of iron 
at a white heat really come into contact with carbon, or when 
a sufficient quantity of carbon has not been burnt away 
during the process of puddling, the skeleton grains, instead 
of adhering to each other, come into a state of fusion during 
the compressing force of the strokes of the hammer, assume 
a crystallized form, and produce a sort of cold-short iron. 
Chemically-pure iron I could never make to weld. I pro- 
cured chemically- pure iron by reducing oxide of iron (from 
which all traces of silica were carefully separated by means 
of repeated solutions and evaporations) in a current of hydro- 
gen, and inclosed it hermetically in a platinum tube pre- 
viously filled with dried hydrogen. Several of these tubes 
were exposed to different degrees of heat, from the dark red- 
heat to the white-heat, and hammered, in order to give con- 
sistency to the inclosed powder of iron ; but it never showed 
any signs of continuity or vvelding afterwards, and under the 
microscope itself appeared unaltered. 
We have just shown, that silicon principally imparts to the 
iron the property of welding ; but we see likewise in the last 
specimen of iron examined, that even a considerable quantity 
of carbon contained in the iron does not impart to it the pro- 
perty of hardening after being heated and cooled in water. 
Besides this last specimen of iron, I melted pure iron with 3 per 
cent, of charcoal prepared from sugar, in a clay crucible, and, 
watching the heat very carefully until it had become quite 
liquid, I poured it into a common ingot mould for cast steel. 
This iron when broken presented a large round granulation 
of a bluish-white colour, resembling the specimen c before 
mentioned; under the hammer it forged extremely soft and 
tough, like Taberg iron, but would not harden at all when 
dipped red-hot into water, notwithstanding it contained 2*5 
per cent, of carbon. But it contained scarcely any traces of 
silicon ; and I found by keeping the liquid mixture for a 
longer time at a higher degree of heat, the silicon increased, 
and with it the proj)erty of hardening. 
We have mentioned in a former paragraph, that iron when 
heated in a reverberating-furnace in contact with siliceous 
matter, imbibes a portion of silicon ; but the question now is, 
why the specimen e, before-mentioned, did not imbibe silicon 
from the bottom of the furnace, but combined instead with 
the carbon of the [decomposed flame? The answer is ob- 
vious. In the cast iron from which the specimen e was pre- 
