vacuum to vaporize the liquid rapidly. The most common form of these vessels is that of a globe. Such a vessel is known as a Dewar globe or bulb.
It has been found that liquid air can be kept very well by putting it in a tin or galvanized iron vessel, which in turn is placed in a larger one, and then filling the space between the two with felt. Under these conditions vaporization takes place quite slowly, and it is possible to transport the liquid comparatively long distances. It has, for example, been transported from New York to Baltimore and Washington. In one case with which the writer is familiar two cans were taken from Mr. Tripler's laboratory in the morning, delivered at the Johns Hopkins University in the afternoon, and used to illustrate a lecture in the evening. After the lecture there was enough left for certain experiments that were carried on during the rest of the night.
Tripler, Linde, and Hampson have all succeeded in devising forms of apparatus by means of which air can be liquefied without the aid of other cooling agents than the expanding air. In principle the methods employed by these three workers are essentially the same. It appears from the published statements that at the present time Tripler's plant is the most efficient. While a few years ago a half pint or so of liquid air is said to have cost five hundred dollars, now five gallons can be made for about twenty dollars, and probably much less. The general working of Tripler's apparatus can be made clear by the aid of the accompanying drawing, Fig. 2. A1, A2, A3 represent steam compression pumps. Air is taken through I from above the roof of the laboratory. In the first pump it is compressed to sixty-five pounds to the square inch. It, of course, becomes heated as it is compressed. In order to cool it down again it is passed through a coil, B1, which is surrounded by water of the ordinary temperature. This compressed and cooled air is then further compressed in the second pump, A2, to four hundred pounds to the