474 Professor Dewar [April 6, 



of hydrogen is proportionally as much below the boiling-point of air 

 as the latter is below the ordinary temperature of this room. 



In order to observe the individual behaviour of the constituents 

 of the air at temperatures below their ordinary boiling-points it is 

 advantageous to place liquid nitrogen and oxygen in separate vacuum 

 vessels, so connected that they may be simultaneously exhausted, as 

 is represented in Fig. 4. On starting the air-pump both liquids enter 

 into rapid ebullition. As the exhaustion gets higher the temperature 

 of each liquid gets lower and lower, and it the melting-point is finally 

 reached in either liquid it must shortly begin to solidify. This con- 

 dition is quickly brought about in the case of the vessel A containing 

 the liquid nitrogen, which passes rapidly into the condition of a dense 

 white snow ; but no amount of time spent in maintaining a good ex- 

 haustion (5 to 10 mm. pressure) has any effuct in changing the liquid 

 condition of the oxygen iu B. Oxygen in fact remains liquid at tem- 

 peratures where nitrogen is solid. The snow of solid air produced by 

 the evaporation of liquid hydrogen, in the previous experiment, might 

 thus bo made up of solid nitrogen and a liquid rain of oxygen. To 

 show that the temperature of boiliug hydrogen solidifies oxygen, 

 some of the latter liquid is placed in a vacuum test-tube O (Fig. 3), 

 and liquid hydrogen H is poured on its surface, when the liquid oxygen 

 is quickly transformed into a clear blue solid ice. Both oxygen and 

 nitrogen, and we shall see later, hydrogen can be changed into the 

 condition of transparent ice as well as into the snowy state. A 

 closed vessel filled with any gas at atmospheric pressure, of such a 

 form that a portion of the surface in the shape of a narrow quill 

 tube, can be cooled in boiling liquid hydrogen like B, Fig. 5, shows 

 condensation of the gas to the solid state ; the only exceptions being 

 helium and hydrogen itself. Here are two vessels of the same 

 shape as A, B, Fig. 5. The first contains helium showing no con- 

 densation when the part B is cooled ; the second is filled with hydro- 

 gen, which equally shows no change of state under the conditions of 

 the experiment. It is easy, however, to make the hydrogen vessel 

 show liquefaction. For this purpose the experiment with the 

 hydrogen is repeated, only before doing so the part A is heated 

 to about 300° 0. over a Bunsen burner, in order to increase the 

 pressure of gas in the interior to above two atmospheres. Now 

 liquefaction is seen to take place with great facility. No change 

 is produced by similarly increasing the pressure in the helium vessel. 



The extraordinary command liquid hydrogen gives us over the 

 transition of state in matter may be best illustrated by the use of 

 a new kind of cryophorus. Wollaston's celebrated instrument 

 operates by forcing the evaporation of water in a closed vessel 

 by condensing its vapour in a part of tho receiver at a distance 

 from the fluid, thereby causing a lowering of temperature in the 

 latter until freezing takes place. Hence the name cryophorus or 

 cold-bearer. Instead of using water we may now show that the same 

 principle may be applied to the solidification of nitrogen at a distance, 

 instead of water. The sole difference in this case is that the 



