RECENTLY DISCOVERED GASES. 273 



contain argon, since at the temperature used argon possessed a measur- 

 able vapor pressure. 



We have, therefore, estimated that the density of the pure gas is 

 9.65. Here our work for tne time was ended by the beginning of the 

 summer holidays. 



On our return we resumed the study of this gas, which we will here- 

 after designate by its name of " neon.*' Its spectrum was photographed 

 by Mr. Baly, one of my assistants, by means of a spectrometer which 

 we had constructed during the vacation. To our astonishment, the 

 lines of helium were easily recognized. A comparison photograph 

 showed this beyond all question. Hence the density of the gas was in 

 all probability too low, owing to the presence of the helium. Since 

 now the temperature used was insufficient to liquify the neon, and 

 since the argon had been removed as far as possible, we had to face 

 the problem of how one could free neon from its accompanying impari- 

 ties. A means was found in its solubility. It is well known that the 

 solubility of those gases which do not react chemically with the solvent 

 follows in general the same order as their condensibility. According 

 to this helinm should have a lesser solubility than neon, and neon than 

 argon. The solubility of these gases in water is, however, too slight 

 to be available for their separation. We have, therefore, used liquid 

 oxygen as a solvent. This mixes with all three gases and boils at a 

 temperature not far from the boiling point of argon. We therefore 

 mixed the gas with sufficient oxygen to be almost wholly condensed at 

 the temperature attained by boiling air at the lowest possible pressure. 

 The uncondensed portion, about one-hfth of the whole, was separated 

 and collected as that richest in helium; the middle portion we con- 

 sidered as purified neon, while the remainder consisted of a mixture of 

 argon and neon; naturally, all these portions contained oxygen in 

 larger or smaller quantities. 



After the removal of the oxygen, which was accomplished by passage 

 over hot copper filings, we determined the density and refractivity of 

 the middle portion. The density in two determinations was 10.04 and 

 10.19; the second figure was obtained after passing the electric spark 

 through the gas mixed with oxygen in the presence of caustic potash 

 and subsequent removal of the oxygen by phosphorus. The entire 

 quantity weighed was only 30 cubic centimeters at a pressure of 250 

 millimeters. The weight was 0.0095 gram. I mention these figures in 

 order to show with what an exceedingly small quantity of gas it is 

 possible to carry out a very satisfactory density determination. 



The refractivity of this portion with reference to the air as unity was 

 0.338. This portion still showed the spectra of argon and helium, and 

 was, therefore, submitted to a second purification, in which the heavier 

 part was more completely removed than the lighter. Even this purifi- 

 cation, however, did not remove all the argon, but its quantity was 

 sm 98 18 



