1898.] on Liquid Air as an Analytic Agent. 821 



delivered, the same material was found by Professors Kamsay and 

 Travers to exist in argon, and has been recognised and named by 

 them Neon, a new element. 



It is, therefore, possible to separate helium from other gases by 

 liquefaction when it is only present to the extent of one part in one 

 thousand. From this it would appear that heliima is less soluble in 

 liquid nitrogen than hydrogen in liquid air, and is of greater volatility 

 than the constituents of the other gases which were condensed. If the 

 sample of the uncondensed gas from the first liquefaction of the Bath 

 gas were again treated in the same way, a much more concentrated 

 specimen of helium could be obtained. Provided helium were wanted 

 on a large scale, then a liquid air apparatus, similar to that in use at 

 the Royal Institution, transported to Bath, and worked with the gas 

 from the King's Well, could be made to yield a good supply, as the 

 gas contains 1 • 2 parts in 1000. 



Argon, which is present in the proportion of 1*4 per cent., con- 

 denses with the nitrogen; but if the liquid be allowed to slowly 

 boil away, a residuum may be obtained containing about 7 per cent, 

 of argon. Argon, when frozen, solidifies to a perfectly clear glass. 



Absorption of Rontgen Radiation at Low Temperature by 

 Different Bodies. 



The transparency of bodies to the Rontgen radiation is an inter- 

 esting study, although we are not in a position to draw definite con- 

 clusions from the results. As a general fact we know the opacity 

 of elements in the solid state increases with the atomic weight. 



In the experiments small tubes of the same bore were filled re- 

 spectively with liquid argon and chlorine, potassium, phosphorus, 

 aluminium, silicon and sulphur, and exposed at the temperature of 

 liquid air (in order to keep the argon and chlorine solid) in front 

 of a photographic plate shielded with a sheet of aluminium to an 

 X-ray bulb (see Fig. 2). The order of increasing opacity of the 

 shadow of each substance was observed, and the sequence in the list 

 given above represents the results. A tube containing silicon was 

 a little more transparent than the potassium or chlorine. Sodium 

 and liquid oxygen and air, nitrous and nitric oxides proved much 

 more transparent than chlorine. Tubes of potassium, argon and 

 liquid chlorine presented no very marked difference of density on 

 the photographic plates. 



From these experiments it would appear that argon is relatively 

 more opaque to the X-rays than either oxygen, nitrogen, or sodium, 

 and that it is on a level with potassium, chlorine, phosphorus, alu- 

 minium and sulphur. This may be regarded as supporting the view 

 that the atomic weight of argon is twice its density relative to 

 hydrogen. 



Thermal Transparency at Low Temperatures. 



Pictet, after an elaborate investigation, concluded that below a 

 certain temperature all substances had practically the same thermal 

 Vol. XV. (No. 92.) 3 h 



