124 PHENOMENA, ATOMS, AND MOLECULES 



ing gas. The addition of nitrogen decreases the heat conductivity of the gas 

 because of the decreased production of atomic hydrogen, and thus the 

 voltage is lowered since less heat is withdrawn from the arc. In the second 

 case, the decreased voltage drop per unit length along the arc is more than 

 compensated for by the increased length of the arc which is brought about 

 by the addition of nitrogen. The lower mobility of ions in the heavier gas 

 probably causes the arc to be carried forward more easily by the blast of 

 gas and thus increases its length. 



PART TWO 



A STUDY OF the heat losses from tungsten filaments at very high tem- 

 peratures in at atmosphere of hydrogen led the writer to conclude in 191 1 ^ 

 that hydrogen is largely dissociated into atoms at temperatures of 25oo°K 

 or more. The total heat loss from the filament, after subtracting that due 

 to radiation, increased in proportion to the y^^ power of the temperature at 

 temperatures over 2700° K, whereas the normal heat loss by convection, as 

 determined for example in nitrogen, should have increased with the i.S*'' 

 power of the temperature. Further work showed ^ that by heating a 

 platinum or tungsten filament above I300°K in hydrogen at low pressures, 

 atomic hydrogen was formed which had very remarkable properties. It 

 would dissolve at ordinary temperatures in platinum and would be con- 

 densed on glass surfaces at room temperature and at this temperature com- 

 bined instantly with oxygen or phosphorus and reduced oxides such as 

 WO3, CuO, Fe203, ZnO or Pt02. More accurate measurements of the 

 heat losses from tungsten filaments in hydrogen at various pressures ^ 

 gave 90,000 small calories as the heat of combination of 2 grams of atomic 

 hydrogen, and showed that the degree of dissociation at atmospheric pres- 

 sure increased from about 2 per cent at 2400°K to about 34 per cent at the 

 melting point of tungsten. 



In attempting to obtain the Balmer spectrum of hydrogen without 

 contamination by the secondary spectrum, R. W. Wood ^ built very long 

 vacuum tubes in which he passed currents of amperes through moist 

 hydrogen at a few millimeters pressure. He observed many remarkable 

 phenomena. Short pieces of tungsten or platinum wire mounted in a side 



* Langmuir, Trans. Amer. Electrochem. Soc, 20, 225 (1911) and Jour. Amer. 

 Chcm. Soc, 34, 860 (1912). 



^ Langmuir, Jour. Amer. Chem. Soc, 34, 1310 (1912) ; Freeman, Jour. Amer. 

 Cheni. Soc, 35,92? (1913). 



^ Langmuir and Mackay, Jour. Amer. Chcm. Soc, 36, 1708 (1914). 37, 4^7 (1915) 

 and 38, 1 145 (1916). 



^ R. W. Wood, Phil. Mag., 44, 538 (1922). 



