In 1892 I visited London, where I met Professor Dewar, who invited me to witness an exhibition of the manufacture of liquid oxygen--and incidentally liquid air--at the Royal Institution. To me it was a most wonderfully interesting event. I saw air, taken from the room, gradually liquefy in a small glass test tube open at the top. When the tube was withdrawn from the refrigerating chamber it boiled by the heat of the room, and rapidly evaporated. We lighted a splinter of wood and blew it out, leaving a live spark on the end of it, and held it over the mouth of the tube, knowing that if anything like pure oxygen were evaporating the splinter would relight and blaze (an old experiment with oxygen gas). At first the splinter would not relight, because the evaporating gases were a mixture of oxygen and nitrogen in the proportions to form air. But owing to the fact that nitrogen evaporates sooner than oxygen, a second trial was successful, for the splinter immediately began to blaze, showing that the gas evaporating then was pure, or nearly pure, oxygen. When the liquid oxygen was poured into a saucer and brought into proximity with the poles of a powerful magnet the liquid immediately rushed out of the saucer and clung to the magnet poles; showing that oxygen is magnetic. Since that time other experimenters have succeeded in making liquid air on a comparatively large scale, and the process is simple when we consider some of the old methods. Mr. Tripler of New York, who has made liquid air in great quantities, does it substantially as follows: First, he compresses air to about 2500 pounds to the square inch. Of course the air is very hot when it is first compressed because all the air in the tank has been reduced in bulk about 166 times, and all the heat that was in the whole bulk of air is concentrated into one-166th of the space it occupied before it was compressed. It is 166 times hotter. There are two sets of pipes running from the compressor to a long upright tank called the liquefier. These pipes pass through running water, so that the compressed air is quickly cooled down to the temperature of the water (about 50 degrees Fahrenheit). The pipes--at least one set of them--run the whole length of the liquefier, and most likely are coiled. This set of pipes contains the air to be liquefied. A second set of pipes runs to the bottom of the liquefier, where there is a valve. By opening this valve a jet of compressed air is allowed to play on the other set of pipes, when intense cold is produced by the sudden expansion of the air. This cold air rushes up around the pipe containing the air to be liquefied and escapes at the top, thus absorbing the heat until the temperature is reduced to 312 degrees below zero. Then the air liquefies and runs into a receptacle, where it may be drawn off at pleasure. It will be seen that a large part of the compressed air is wasted in cooling the remainder sufficiently to liquefy.