High Frequency Corona Discharge 163 
so that the energy expended in the tube is not proportional to the pri- 
mary energy. In Fig. IV, where air enclosed within the tube is sub- 
jected to discharge for different periods of time, the concentrations are 
increased with time but at a loss of efficiency owing to the simultane- 
ous decompositions of ozone, such as occurs also with very low fre- 
quencies or direct current.. In this way, concentrations up to 15 g. 
Te iL War THOur 
G 
i ry 
WL 
V76 s/ Eee een 
Fig. V. The rate of flow of air through the tube was varied. The length of 
discharge was 1 min., after which the tube was swept out completely. A No. 18 wire 
was used and the primary power was 75 2 watts. The spark-gap was set at 2.00 mm. 
x 
Pe 
G 
SILELYMO 77S 
oo/ 
bo eo) 
O02 
MIL 
oof 
per cubic meter (mg. per liter) were obtained. In Fig. V an optimum 
flow rate is indicated for the given conditions. Similar optima have 
been previously obtained with different types of 60-cycle ozonizers.° 
The size of wire has a marked effect on the optical and chemical 
properties of the discharge. Around a fine (No. 33) wire there is a 
8 Chassy, Compt. rend. 133, 789 (1901) with a Berthelot ozonizer and 60 cycle cur- 
rent. Anderegg, loc. cit., gives curves for a direct current discharge in oxygen. The 
shapes of these curves have been checked by A. C. Grubb and J. K. Stewart working 
independently in this laboratory with current rectified by a kenotron. 
®Ray and Anderege, J. Am. Chem. Soc. 43, 967-78 (1921), observed an optimum 
flow rate with the 60 cycle discharge in apparatus A. It was also noted with the very 
large apparatus E for both ozone and nitric acid production. 
