ON THE COMBUSTION OF IRON IN COMPRESSED OXYGEN. 
239 
operations, is known. In the directions for the use of Natterer’s apparatus, contained in 
the article “ Ivohlensaure ” ( 4 Handworterbuch der Chemie,’ Band iv.), the writer states, 
that before pumping commences, the space between the piston and the valve should be 
filled up with oil, so as to prevent the retention of any gas between the piston and valve, 
when the former, in compressing, is pushed to the extreme limit of its stroke. Any gas 
so remaining in such space (schadlicher Raum ), expands again on the return of the 
piston, and thus causes, if not an actual loss of power, at least a considerable retardation 
in the compressing process. In the operation above described, in which 60 atmospheres 
were compressed with impunity into the receiver, I omitted to follow this part of the 
directions of the ‘Handworterbuch,’ whilst in the subsequent experiment in which igni¬ 
tion occurred, a layer of olive oil, about O'l inch thick, was poured upon the piston, so 
as exactly to fill the space above mentioned. 
Now, a careful examination of the burnt parts of the apparatus, leaves no doubt that 
the combustion commenced in the space between the piston and valve, and that it was 
this layer of oil which first became ignited. The compression of oxygen to J^th of its 
volume should, according to thermo-mechanical laws, raise the temperature of the oxygen 
to upwards of 2000° C., but after making due allowance for the loss of heat to surround¬ 
ing surfaces, there still remains a temperature sufficiently high for the ignition of oil 
under favourable circumstances. If the oil be spread as a thin film upon the surface of 
a mass of metal, the rapid absorption of heat by the latter prevents the temperature of 
the oil from rising to its igniting-point; but, in the form of a layer O’l inch in thickness, 
no such rapid refrigeration can occur, and the surface of the oil, in contact with the gas, 
may become ignited by the rapid communication to it of the high temperature of the 
compressed oxygen. It is also not improbable, that traces of chlorate of potash, which 
are always carried over with oxygen when the latter is rapidly evolved,_ may have found 
their way into the pump, and contributed, to some extent, to the ready inflammability of 
the oil. However this may be, the result ought to be regarded as a caution against the 
use of combustible lubricants in the compression of oxygen or nitrous oxide. If ignition 
of the oil occur at high pressures, it will assuredly be communicated to the iron of the 
receiver, which evidently burns in oxygen compressed 25 times, with at least the same 
facility as tissue-paper in atmospheric air, the condition of the various parts of the ap¬ 
paratus, after the explosion, leading to the conclusion, that the combustion from the 
beginning to end occupied only a very short time, probaoly not more than three oi four 
seconds. The risk attending'the compression of oxygen and nitrous oxide, may be 
avoided by the employment of a non-combustible lubricant. For this purpose, a strong 
solution of soft-soap in distilled water, appears to answer very well. 
The facility with which a mass of iron thus becomes ignited, and the rapidity with 
which it burns in oxygen, at high pressures, suggests the possibility of employing shells 
of wrought or cast iron charged with compressed oxygen, for warlike purposes. The in¬ 
terior of such a shell would scarcely be more difficult to ignite than gunpowdei, and, 
once ignited, the pressure of the enclosed oxygen would, notwithstanding its absorption, 
be for some time augmented by the intense heat, whilst the walls of the shell would be¬ 
come thinner, until they finally burst into fragments of burning and semi-molten iron. 
The condition necessary to secure such a result, may be determined from the known ab¬ 
solute thermal effect of iron in oxygen. Andrews found that the union of one litre of 
oxygen with iron, produces sufficient heat to raise the temperature of 5940 grammes of 
water through 1° C. It hence follows, that 780 cubic inches of oxygen, by combination 
with iron, would evolve sufficient heat to raise 1 lb. of cast-iron to its melting-point. 
This amount of oxygen introduced into the receiver above described, would exert a pres¬ 
sure of 20’5 atmospheres, consequently in would require the union of a quantity of oxy¬ 
gen exerting a pressure of 125 atmospheres, to raise the w hole of the receivei to the 
melting-point of cast-iron. These conditions are not encouraging; for, although al ess 
amount of oxygen than that required for the complete fusion of the shell would suffice 
for the purpose required, yet, it would doubtless be necessary to augment the thickness 
of such a vessel when used as a projectile, and this would necessitate a corresponding in¬ 
crease in the bursting charge of oxygen. Ihus, little could probably be effected with 
less than 100 atmospheres of oxygen forced into the shell, a pressure, which, I fear, 
would prove not only dangerous, but unmanageable .—Journal of the Chemical Society. 
