272 RECENTLY DISCOVERED GASES. 



tubes in which the gas passed over respectively hot copper oxid, con- 

 centrated sulphuric acid, and phosphorus pentoxid; it then passed by a 

 two-way cock into a small flask, holding about 30 cubic centimeters, 

 which was inclosed in a Dewar tube. By means of the other opening 

 of the cock, the flask was connected with a mercury gasometer. By 

 means of a U-shaped capillary and mercury trough, it was also possi- 

 ble, through a three-way cock, to collect the gas at will in glass tubes. 

 About 50 cubic centimeters of liquid air were poured into the double 

 walled tube, and, by means of a Fleuss air pump kept constantly in 

 action, the liquid air boiled at 10 to 15 millimeters pressure. The argon 

 liquefied rapidly as soon as subjected to this low temperature, and in 

 the course of half an hour it was completely condensed. Altogether 

 there were about 25 cubic centimeters of a clear, limpid, colorless 

 liquid, in which floated white flakes of a solid substance. By stopping 

 the pump the pressure over the liquid air was now increased, and the 

 argon boiled quietly, the first portions of the gas being collected in the 

 mercury gasometer. Changing now the three-way cock, the largest 

 portion of the argon passed back into the iron gasometer ; after nearly 

 all the liquid had boiled away and only the solid substance was left in 

 the flask, the last portions of the gas were collected separately. The 

 solid substance remained persistently in the flask; it was slowly vola- 

 tilized by means of a Topler pump, which stood in connection with the 

 apparatus. 



We first directed our attention to the lighter fractions, for these had 

 for us the greatest interest. The density of this gas was found to be 

 14.67 ; the ratio between the specific heats was as usual 1.66, and the 

 spectrum showed, beside the well-known groupings of argon, a large 

 number of red, orange, and yellow lines of varying intensity. Evidently 

 we had before us a new element, which was contaminated with argon. 



This gas was then liquefied in a similar apparatus to that first used, 

 but constructed on a smaller scale; a portion, however, remained 

 uncondensed. Even by raising the reservoir of the mercury gasometer 

 until an overpressure of an atmosphere was reached, it was impossible 

 to convert all the gas into a liqgjd, although the temperature of the 

 boiling air was reduced as low as possible by rapid pumping. By 

 repeated raising and lowering of the reservoir we finally passed all the 

 gas through the cooled space, in order to free it, as fer as possible, 

 from argon. The uncondensible gas was collected by itself, and the 

 remainder was evaporated into another gasometer. 



You can well imagine how eager we were to know what the density 

 of this purified gas would prove to be. It was immediately weighed. 

 Our satisfaction can well be realized when we found that its density 

 was 9.76. Since, however, its spectrum at low pressure still showed 

 argon lines, though weak, we were compelled to admit that this number 

 was certainly too high. It was impossible that this gas should not 



