The Preparation and some of the Properties of Pure Argon. 185 



account of all these gases has already been given in the ' Proceedings ' 

 under the names of " Neon," or " new," " Krypton," or " hidden," and 

 " Metargon" ; and at the meeting of the British Association at Bristol, 

 the discovery of " Xenon " or " the stranger " was announced. The 

 removal of these gases from argon has put us in possession of pure 

 argon, and the present paper deals with some of its properties. 



The Properties of Pure Argon. 



In order to prepare 15 litres of argon, it is necessary to deal with 

 about 1,500 litres of atmospheric air, of which approximately 1,200 

 litres consist of a mixture of nitrogen and argon. To absorb the 

 nitrogen contained in this quantity of gas by conversion into nitride, 4 

 kilograms of magnesium would be required theoretically, but in order to 

 cover loss through leakage and incomplete action, 5 kilograms of the 

 metal are employed. The absorption of the oxygen and nitrogen was 

 conducted in three stages. In the first, the oxygen was removed by means 

 of metallic copper ; in the second, the nitrogen was passed twice over 

 metallic magnesium ; and in the third, the gas, now rich in argon, was 

 finally freed from nitrogen and hydrogen by passage over a mixture of 

 anhydrous lime and magnesium powder heated to a red heat, and sub- 

 sequently over red-hot copper oxide. The apparatus employed is 

 shown in detail in the annexed figure. 



It was of course necessary to confine the gas over water between the 

 successive stages of purification, and finally to store the gaseous argon 

 in the same way. On account of the considerable solubility of argon 

 in water, this would have entailed no small loss if the quantity of water 

 with which it had been brought into contact had been large. We con- 

 sequently decided to make use of gas-holders of the gasometer type, in 

 which the water was contained in an annular space of small capacity. 

 Balance weights were attached to cords passing over pulleys, and 

 served to relieve the pressure on the gas due to the weight of the gas- 

 ometer. As the volume of the gas decreased after each successive 

 stage, the four gas-holders employed were of different sizes ; the capa- 

 city of A was about 180 litres ; that of B, 27 litres ; and of C and D, 

 each 18 litres. 



Atmospheric nitrogen was obtained by drawing air, freed from car- 

 bon dioxide by passage through caustic soda solution, over heated 

 metallic copper. A large iron tube F, 3 feet 6 inches long, and 3*5 inches 

 in diameter, containing 25 Ibs. of scrap copper, was connected with the 

 gas-holder A ; the tube was heated in a long fire-brick trough during 

 these experiments, but a gas-furnace is shown in the figure, which 

 has now been substituted for the more primitive arrangement. 



The time required to fill the gas-holder was usually about five hours, 

 and it was found, on analysis of the gas, that one single operation 



