DESCRIPTION OF THE DRAWINGS.

A. A tube, 5 metres long, 14 millimetres external diameter, and 4 millimetres internal diameter, in which the oxygen condenses. It is furnished with a screw-tap, r, from which the liquid oxygen jets out. A pressuregauge, м, measures the pressure up to 800 atmospheres.

B. A tube, 4 metres long, in which is solid carbonic acid. The stock of carbonic acid is contained in a gasometer, G, of I cubic metre capacity. A three-way tap, H, puts it when desired into communication with the apparatus.

c. A howitzer shell, containing 700 grms. of chlorate of potash mixed with chloride of potassium. It is heated with gas.

P1, P2. Double-action exhaustion and force pumps, drawing carbonic acid from the tube B or the gasometer G, according to the position of the tap H. s. A tube, 60 millimetres diameter and 11 metres long, in which is condensed the liquid carbonic acid compressed by the pumps. This liquefied gas returns by the small tube t to the tube B.

R. A tube, 125 millimetres in diameter and 1'1 metres long, containing liquid sulphurous acid.

P3, P4. Double-action exhaustion and force pumps, exhausting sulphurous acid gas from the tube R.

Q. A tubular condenser of sulphurous acid compressed by the pumps. This body, when liquefied, returns by the small tube f to the tube R. The cold water for condensing the sulphurous acid passes through the aper

tures E E.

Exit for the vapourised carbonic acid caused by the suction of the pumps.

approach each other as much as possible, certain indispensable conditions are necessary, which may be expressed

thus:

I. To have the gas absolutely pure, with no trace of foreign gas.

2. To be able to obtain extremely energetic pressures. 3. To obtain intense cold, and to subtract heat at these low temperatures.

4. To utilise a large surface for condensation at these low temperatures.

5. To be able to utilise the rapid expansion of the gas from extreme condensation to the atmospheric pressurean expansion which, added to the preceding means, will compel liquefaction.

Having fulfilled these five conditions, we may formulate the following alternative :

When a gas is compressed to 500 or 600 atmospheres, and kept at a temperature of 100° or 140°, and it is allowed to expand to the atmospheric pressure, one of two things takes place :

Either the gas, obeying the force of cohesion, liquefies, and yields its heat of condensation to the portion of gas which expands and loses itself in the gaseous form; or, on the hypothesis that cohesion is not a general law, the gas must pass to the absolute zero and become inert, that is to say, an impalpable powder.

The work done by expansion will not be possible, and the loss of heat will be absolute.

Struck with the truth of this alternative, which is rendered certain by thermo-dynamic equations based on accurate data, I have sought to produce a mechanical arrangement which should entirely satisfy these different conditions, and I have chosen the complicated apparatus of which the following is a brief description :

I take two pumps, P, and P4, for exhaustion and compression, such as are used industrially in my ice-making apparatus. I couple these pumps in such a way that the exhaustion of one corresponds to the compression of the other. The exhaustion of the first communicates with a tube (R) of I'I metres long and 12.5 centimetres in diameter, and filled with liquid sulphurous acid. Under the influence of a good vacuum the temperature of the liquid rapidly sinks to -65°, and even to 73°, the extreme limit attained.

Through this tube of sulphurous acid passes a second smaller tube (s), of 6 centimetres diameter, and the same