188 Proceedings of hidicma Academy of Science 



up 110 V. alternating current. The voltage was taken from the primary 

 voltmeter reading and the transformer ratio. The gas pressure in the 

 tube was measured by means of a manometer filled with "stanolax", 

 a heavy paraffin oil with a density of 0.857. Some of the results ob- 

 tained are given in figure 1. It is to be noted that there is a critical 

 voltage above which there is no pressure decrease under the given con- 

 ditions. At the lower voltages the first part of the decrease in pres- 

 sure, when allowance is made for the ionization pressure " is exactly 

 caused by the ozone formed as is shown in fig-ure 2. Towards the re- 



I 



+ eo 



■¥30 



-30 



7~/rrje /n n in. 



Fig. 2. A series of runs over varyiriK len^lh of discharjic time was made at about 

 6100 volts. 



versal point there is an accumulation of oxides of nitrogen which pre- 

 vent the further formation of much ozone. This part checks Mc- 

 Eachi'on's results. 



Disciission. In the corona discharge ozone is the chief product 

 foiTTied at first, provided the intensity of the discharge and the tem- 

 perature are not too great. A little nitrogen is oxidized, which grad- 

 ually accumulates with time, and not only reduces the ozone concen- 

 tration, but prevents further ozone formation. There are two possi- 

 bilities, one is catalytic decomposition of the ozone molecules in the gas 

 space by oxides of nitrogen; another is the adsorption of the oxides of 

 niti'ogen, (these may possibly be very heavy molecules), on the walls 



' Kunz. Phys. Rev. 8, 285 (1916) ; 10, .18;5 (1917) : 19. 165, 244. :iOO (1922). 



