214 Centenary Commemoration, 1799-1899. [June 7, 



by removing for an instant the cotton-wool stopper, when you see a snow 

 of solid air falling in the liquid. It is easy to arrange a method of 

 carrying liquid hydrogen in a small vacuum vessel in such a way as 

 to prevent the access of air. This is shown in Plate I., where the 

 vacuum vessel, after it has been filled by dipping it into the main 

 supply by means of a supporting wire, is surrounded with a glass 

 envelope, which becomes filled with an atmosphere of hydrogen 

 gas constantly maintained, thereby preventing the access of air. 

 That the density of the liquid is very small and is altogether 

 unlike liquid air is shown by dropping small pieces of cork, 

 which float readily in the latter liquid, but sink instantly in the 

 hydrogen (Figure B, Plate III.). The real density of the liquid is 

 only one-fourteenth that of water, so that it is by far the lightest 

 known liquid. This small density explains the rapidity with which 

 the liquid is cleared on the entrance of the air snow. The relative 

 smallness of the gas bubbles produced in the actively-boiling liquid 

 which causes an appearance of opalescence, is really due to the small 

 surface tension of the liquid hydrogen. The coefficient of expansion of 

 liquid hydrogen is some five times greater than that of liquid oxygen, 

 and is comparable with that of carbonic acid, about 5° from its 

 critical point. The latent heat of evaporation is about 190 units, 

 and the specific heat of the liquid is very high, and so far as my ex- 

 periments go, leads me to the value 6. This is in very marked 

 contrast to the specific heat of liquid oxygen, which is about 0*5. 

 The extraordinary lowness of its boiling point is at once apparent by 

 cooling a piece of metal in the liquid and then removing it into the air, 

 when it will be seen to condense for a moment solid air on its surface 

 which soon melts and falls as a liquid air. This may be collected 

 in a small cup, and the production of oxygen demonstrated by the 

 ignition of a red-hot splinter of wood after the chief portion of the 

 nitrogen has evaporated. If a long piece of quill tubing sealed at 

 one end, but open at the other, is placed in the liquid, then the part 

 that is cooled rapidly fills with liquid air. On stopping any further 

 entrance of air by closing the end of the tube, the liquid air quickly 

 becomes solid, showing in the interior a hollow spindle from con- 

 traction, in passing from the liquid into the solid form (Figure E, 

 Plate III.). On bringing the tube containing the solid from the liquid 

 hydrogen bath into the air we observe liquid air running from the 

 surface while the solid air inside is seen to melt (Figure D, Plate III.). 

 Here is a tube into which liquid oxygen has been poured. On placing 

 it in liquid hydrogen it freezes to a clear blue ice. Liquid nitrogen 

 under similar circumstances forms a colourless ice. If instead of 

 an open tube in free air we employ a closed vessel of about a litre 

 capacity to which the quill tube is attached, then, on repeating the 

 experiment, the same results follow, only the volume of the liquid 

 air formed agrees with the total quantity present in the vessel. 

 This suggests that any air left in the closed vessel must have a very 

 small pressure. This is confirmed by attaching a mercurial gauge to 



