TRANSACTIONS OF SECTION B. 829 



elements of higher atomic weight tbau argon, belonging to the same series. The 

 search for this element was described in last year's Address, and, it aviU be remem- 

 bered, the results were negative. 



Reading between the lines of the Address, an attentive critic might have 

 noticed that no reference was made to the supposed homogeneity of argon. From 

 speculations of Ur. Johnstone Stoney, it would follow that the atmosphere of our 

 planet might be expected to contain new gases, if such exist at all, with densities 

 higher than 8 or thereabouts. Dr. Stoney gives his reasons for supposing that tlie 

 lighter the gas the less its quantity in our atmosphere, always assuming that no 

 chemical compounds are known which would retain it on the earth, or modify its 

 relative amount. Therefore it appeared worthy of inquiry whether it was possible 

 to separate light and also heavy gases from argon. 



The beautiful machine invented by Dr. Hampson has put it in our power to 

 obtain, through his kindness and that of the ' Brin ' Oxygen Company, large 

 quantities of liquid air. We were therefore able to avail ourselves of the plan 

 of liquefaction, and subsequent fractional distillation, in order to separate the 

 gases. 



On liquefying 18 litres of argon, and boiling off the first fraction, a gas was 

 obtained of density 17 (0 = 16). This gas was again liquefied and boiled off 

 in six fractions. The density of the lightest fraction was thus reduced to 134, and 

 it showed a spectrum rich iu red, orange, and yellow lines, difiering totally from 

 that of argon. On re-fractionating, the density was reduced further to 10'8 ; the 

 gas still contained a little nitrogen, on removing which the density decreased to 

 9-76. This gas is no longer liquefiable at the temperature of air boiling under a 

 pressure of about 10 millimetres ; but if, after compression to two atmospheres, the 

 pressure was suddenly leduced to about a quarter of an atmosphere, a slight mist 

 was visible in the interior of the bulb. This gas must necessarily have contained 

 argon, the presence of which would obviously increase its density ; and in order to 

 form some estimate of its true density, some estimate must be made of the relative 

 amount of the argon. We have to consider a mixture of neon, nitrogen, and 

 argon, the two latter of which are capable, not merely of being liquefied, but of 

 being solidified without difficulty. Under atmospheric pressure nitrogen boils at 

 — 194°, and solidifies at -214°, and the boiling-point of argon is —187*, and the 

 freezing-point —190"; the vapour-pressure of nitrogen is therefore considerably 

 higher than that of argon. The mist produced on sudden expansion consisted of 

 solid nitrogen and argon ; and for want of better knowledge, assuming the vapour- 

 pressure of the mixture of nitrogen and argon to he the sum of the partial pressui'es 

 of the two, it is obvious that that of argon would form but a small fraction of the 

 whole. The vapour-pressure of argon was found experimentally to be 109 milli- 

 metres at the temperature of air boiling in as good a vacuum as could he produced 

 by our pump ; but as we have only to consider the partial pressure of the argon at 

 a much lower temperature, we do not believe that the pressure of the argon can 

 exceed 10 millimetres in the gas. This would correspond to a densit}' for neon of 

 9-6. 



The ratio between the specific heat at constant pressure and constant volume was 

 determined for neon in the usual way, and, as was to be expected, it approximates 

 closely to the theoretical ratio, being 1'655. We therefore conclude that, like- 

 helium and argon, the gas is monatomic. 



It may be remembered that the refractivity of helium compared with that of 

 air is exceptionally low — viz., 0-1238. The lighter gas, hydrogen, has a refractivity 

 ofO'4733. It was to be expected from the monatomic character and low density 

 of neon that its refractivity should be also low; this expectation has been realised, 

 for the number fotind is 0-3071. Argon, on the other hand, lias a refractivity 

 not difiering much from that of air — viz., 0-068. Since the sample of neon certainly 

 contains a small amount of argon, its true refractivity is probably somewhat 

 lower. Experiments will be carried out later to ascertain whether neon resembles 

 helium in its too rapid rate of diftusion. 



The spectrum of neon is characterised by brilliant lines in the red, the orange, 

 and the yellow. The lines in the blue and violet are few, and comparatively 



