NATURE 

PAPERS ON IRON AND STEEL 
V.—THE BESSEMER PROCESS (continued ) 
‘ea the previous papers I have described the phenomena 
presented during the different stages of the blow, and 
have endeavoured to explain the chemical actions upon 
which they depend. The next stage, that of adding 
the molten spiegeleisen to the iron which has been fully 
acted upon by the blast, also presents some interesting phe- 
nomena which have not hitherto been fully examined. 
In a paper on “Burnt Iron and Burnt Steel,” read 
b2fore the Chemical Society 6th April last,* I showed that 
the “burnt iron” of the workman is really what its name 
implies, viz., iron which has been more or less oxidised 
throughout its substance, and that “burnt steel” is quite 
different,—_that the presence of combined carbon in suffi- 
cient quantity effectually protects iron from oxidation by 
heat. 
These conclusions are strikingly illustrated in the Bes- 
semer process. In spite of the excessively high tem- 
perature and the abundant supply of oxygen during the 
blast producing most violent combustion of the material 
ithe converter, I have found no “burnt iron” during the 
early or middle stages of the blow. This only appears at 
quite the latter stages when the carbon is nearly all burnt 
out. At the termination of the blow, the material left in 
the converter is burnt iron of a very exaggerated type in 
all cases where the burning out of the carbon has been 
carried to its full extent. 
Mr. Bessemer failed in his attempts to produce malleable 
iron by his process, and all subsequent attempts have 
equally failed, even when the very finest qualities of he- 
matite pig-iron have been used. I am not aware that any 
explanation of this has yet been given, but have no doubt 
that it depends upon the principle above stated, viz., that 
some combined carbon is absolutely necessary to preserve 
the iron from oxidation, and thus, as the carbon is re- 
moved, the iron begins to oxidise throughout, and we have 
an incoherent mixture of iron and particles of oxides, 
which crushes under the hammer or the rolls, is neither 
malleable nor ductile. 
The degree of rottenness depends upon the extent to 
which the blow has been carried, and the iron thus pro- 
duced varies froma quality which simply cracks at the 
edges when hammered or rolled, toa mass that crushes 
into granules like a piece of coarse sandstone. If in- 
attention or some hitch in the machinery prevents the 
immediate turning over of the converter, and the blow is 
continued a few minutes too long, the amount of oxidation 
is so considerable that the mass in the converter loses 
its fluidity, and becomes a spongy and pasty mixture of 
mieited iron and infusible oxide, which is rather trouble- 
some to the manufacturer. 
By the simple method described in the paper above- 
r2ferrea to, I have been able at once to detect the presence 
of entangled particles of oxide in the midst of the iron 
remaining in the converter at the end of the blow. They 
are even visible un tne fracture of overblown iron. 
The presence of this free oxide explains some otherwise 
inexplicable phenomena which accompany the pouring of 
the spiegeleisen. A furious ebullition of the molten metal 
takes place, jets of burning carbonic oxide spurt up vio- 
lently from all parts of the surface ; the converter is filled 
with the blue flame which pours forth from its mouth, 
producing the weird illumination I have already described. 
The outpouring flame so completely occupies the whole 
dimensions of the mouth of the converter, that no air can 
possibly enter, and thus all the oxygen required for the 
combustion which is going on must be derived from the 
material inside the converter. Some of the carbon of the 
spiegeleisen is thus burning at the expense of the oxide 
of the original charge, and this oxide is thereby reduced. 
The sole function usually attributed to the spiegeleisen 
* An abstract of this paper will be found in Nature, April 20, p. 497. 

is that of converting the iron into steel; but if the above 
be correct, it performs, in addition to this, the important — 
service of reducing the free oxide of the rotten burnt iron, 
and thereby rendering it malleable. We shall now under- 
stand why Mr. Bessemer and others have failed to produce 
malleable iron by directly oxidising the silicon, carbon, &c., 
of the pig-iron in the converter. It may be asked how 
then does the puddler remove the carbon from pig 
iron? My answer is simply that he does it by a far less 
violent process of oxidation ; that towards the end of his 
work when the iron is “coming to nature,” 7.e, when the 
proportion of protecting carbon has become very small, 
he takes especial precautions by closing the dampers, and 
otherwise diminishes the rate of oxidation as much as 
possible, and thus he is able to work down to less than 
zis per cent. of carbon without burning his iron. 
The more violent oxidising agency of the Bessemer 
blast demands a greater quantity of carbon for the pro- 
tection of the iron, and accordingly it is found that about 
0°25 per cent. is the minimum limit of carbon which is 
practically obtainable without sacrifice of malleability. I 
have determined the carbon of many hundreds of samples 
of Bessemer steel which has been specially made as 
“ mild” as possible, where it was a primary object to reach 
the minimum proportion of carbon, and have never found 
any sound metal to contain less than 0°20 per cent. The 
usual range of this (which issometimes called “ Bessemer 
metal” being scarcely steel although not true iron) is from 
0°25 to 0°30 per cent. of carbon. I do not here speak of 
the limits of absolute possibility, but of the practical limits 
of the process as at present conducted. 
I should add that, in the course of subsequent working 
the proportion of carbon is reduced, but the extent of 
this reduction is very variable, depending on the number 
of re-heatings, the amount of surface exposed, and the 
kind of furnace in which the reheating is conducted. By 
using a reducing flame the oxidation of the carbon may 
be wholly prevented, but in the ordinary reheating or mill 
furnace and in the exposure of rolling, &c., a certain 
amount of oxidation commonly occurs. Rails and tyres 
usually contain two or three hundredths per cent. less than 
the ingots from which they were made, thin plates and 
sheets lose a larger proportion, even as much as one-tenth 
per cent. inextreme instances. I have removed the whole 
of the carbon from the surface of a hard steel plate by 
exposing it for several days to the low red heat of an 
annealing furnace. W. MATTIEU WILLIAMS 

THE CAUSE OF LOW BAROMETER IN THE 
POLAR REGIONS AND IN THE CENTRAL 
PART OF CYCLONES 
is none of the treatises on Meteorology or Physical 
-- Geography is there to be found any satisfactory ex- 
planation of the observed low barometer in the polar 
regions, or in the centre of acyclone. Observations show 
that in the Antarctic region there is a permanent depres- 
sion of more than one inch below the average height 
nearer the equator, and in the Arctic region a depression 
of about halt that amount ; and also that for several days 
frequently the barometric pressure of the central part of 
a cyclone is one or two inches less than that of the exte- 
rior part. Mr. Buchan, in his excellent treatise on Meteo- 
rology, attributes the low barometer in the polar regions 
to the effect of the vapour in the atmosphere. If the 
amount of vapour in the polar regions was greater than 
in the equatorial, this effect, so far as it would go, would 
be in the right direction; but just the reverse is the case; 
for it is well known that the amount of vapour in the warm 
equatorial region is much in excess of that in the cold 
polar regions. Attempts have also been made, without 
success, to account for the depression in cyclones by the 
effect of centrifugal force. 
[Futy 20, 1871 


