March 16, 1871] 
NATURE 
389 

of the Bessemer flame, though skilfully and laboriously 
conducted, have been curiously barren of philosophical 
results, and for practical purposes altogether a failure. If 
the above questions were answered it would be different. 
During the first part of the blow a large proportion of 
the graphitic carbon of the grey iron becomes converted 
into the condition of “combined carbon” such as exists 

FIG. I.—“ BOMBS” PRODUCED IN THE BESSEMER PROCESS. 
in steel and white cast-iron. What then is the condition 
of this carbon if it yields a cyanogen or hydro-carbon 
spectrum? Nitrogen has been found in considerable 
quantities in cast-iron which is rich in carbon. Steel 
makers know that organic compounds containing nitrogen 
as well as carbon are far more efficacious in cementation 
than carbon in a state of comparative purity. Thus, bone 
dust is more effectual for case-hardening than wood char- 
coal, and ferrocyanide of potassium still more effectual 
than bone dust. Every steel maker is the proud possessor 
of a profound secret, a “ physic” which he furtively buys 
from a distant druggist or drysalter, and having dis- 
guised its yellow colour by grinding it with lamp black, 
locks the physic in a strong box and the secret in his 
own bosom. As this profound secret, like so many others, 
is perfectly well known to all whom it may concern, I 
perpetrate no breach of confidence in describing the 
“physic” as the ferrocyanide of potassium. Its value 
is unquestionable, but ow it acts is still a mystery. 

FIG, 2. 
Some chemists have maintained that a nitride of iron is 
essential to the production of steel, and others have hinted 
at the existence of a cyanuret. As we know that carbon 
exists in the same condition in white iron as in steel, a 
question of considerable interest is offered for spectro- 
scopic solution, viz., Is cyanogen burning in the Bessemer 
flame? Again, we have another set of workshop facts 
and laboratory experiments which go to show that for the 
production of steel, hydrogen in the form of hydro-carbon 
is necessary. It is well known that coal gas, paraffin, and 
other hydro-carbons, are more efficient cementing agents 
than pure carbon. Dr. Percy found that the charcoal of 
sugar, which retained some hydrogen or hydro-carbon, 
readily converted iron into steel, but that the same char- 
coal failed to produce steel under similar circumstances, 
after it had been deprived of its hydro-carbon. It is well 
known that wood charcoal which has been several times 
heated in the cementing furnace, loses some of its power 
of cementation, and this has been attributed to the driving 
off of the hydro-carbon contained in the fresh charcoal. 
Again, it is found that when, by means of an acid, we 
dissolve the iron of steel or white iron away from its 
carbon, the residue is not simple solid carbon, but an un- 
mistakeable liquid hydrocarbon, an oil which, like other 
| its existence. The information thus afforded is 


hydrocarbons, burns with a smoky flame. In this case it 
is possible that the hydrogen may be supplied to the car- 
bon by the water or the acid. If so it presents an in- 
teresting case of the formation of what we usually regard 
as organic matter from inorganic materials, 
If, on the other hand, the hydrocarbon exists ready 
formed in the steel and the white iron, the conversion of 
grey iron into white iron, ze. of graphite into this hydro- 
carbon, is a still more remarkable case of the same kind. 
It is true that the hydrogen may be detected by a direct 
combustion analysis, but this does not reveal the mode of 
| analogous 
to what we obtain by similar means respecting nitrogen. 
The last change that occurs in the flame, that which 
announces to the foreman the time for stopping the blow, 
requires but little explanation; but it is, nevertheless, 
instructive if thoughtfully examined. The contraction of 
the flame and loss of brilliancy is evidently due to the 
exhaustion of the carbon. The change which occurs is 
very similar to that which is observed when air is admitted 
to a jet from which coal gas is burning. 
If to such a jet, supplied with a constant quantity of 
coal gas, atmospheric air be admitted, so that it shall mix 
with the gas before burning, the white or luminous portion 
of the flame will contract in proportion to the quantity of 
air supplied, and if a gradually increasing quantity of air 
be admitted, this contraction will progress until the white 
flame is totally extinguished. Mr. Jonathan Wilkinson, 
of Grimesthorpe, who has been recently investigating this 
subject with a view to practical photometric applications, 
finds that the quantity of air required thus to extinguish 
the white flame is proportionate to the quantity of carbon 
combined with the hydrogen, or to the illuminating power 
of the gas,—that for every standard “candle” of illumi- 
nating power about o'2 of air is required. Thus fifteen- 
candle gas will require three times its own volume of air 
at the same temperature and pressure for the extinction of 
the white flame, seventeen-candle gas 3'4 of air, and so on. 
In the case of the Bessemer flame, we have a constant 
supply of air to a diminishing supply of carbon, and 
therefore we may expect that there should occur in 
the white portion of the flame due to hydro-carbon a 
change corresponding to that which would occur in 
Mr. Wilkinson’s photometric flame, if, instead of a constant 
supply of coal gas mixed with an increasing supply of air, 
he maintained the air in constant flow and gradually 
closed the gas-cock. In this case there would not only 
occur a gradual diminution of the brilliancy but also of 
the dimensions of the flame. 
Such is the change which takes place in the Bessemer 
flame towards the end of the blow, and it so far confirms 
the hypothesis that a considerable portion of the white 
flame is due to hydro-carbon. If it were due to the com- 

S's 
FIG. 3. 
bustion of iron the white flame should increase towards 
the end of the blow, for it is then that the iron, when no 
longer protected by the more combustible carbon, begins 
to burn in a serious degree, just as I have shown that the 
full combustion of the carbon takes place after the bulk of 
the silicon has been oxidised. 
W. Matrieu WILLIAMS 
' 
