158 Proceedings of Indiana Academy of Science 
Experimental: Preliminary experiments were made with a small 
Tesla coil. One side of both the secondary and primary was grounded, 
the other end of the terminal ending in a platinum wire supported on 
the inside of a bell, similar to the method proposed by H. Guilleminot,* 
so that the corona discharge, which is produced under suitable electrical 
conditions, may ozonize the air. The Tesla coil was supplied with energy 
from a small induction coil. The concentrations of ozone were very 
small and the power yield was only 2-5 g. per killowatt hour on a 
basis of gross input into the induction coil. A more suitable arrange- 
ment for securing a corona discharge is a wire passing through the 
axis of a tube, as previously used for the study of the direct current 
corona. At first a No. 29 platinum wire was placed in a five-inch 
aluminum tube. The results obtained were quite variable and it was 
thought that the variability of the direct current which excited the 
induction coil was responsible; but on applying storage battery cur- 
rent, the improvement was not great. An oscillographic study of the 
secondary current and of the voltage from a tertiary coil wound on 
the induction coil showed a very inconsistent wave form. -Various make- 
and-breaks were tried without success. A 5,000 v. Thordarson type H-1 
transformer was substituted and with a fairly constant source of alter- 
nating current gave results which were very much more consistent. 
Description of Electrical Set-up. A very steady source of alter- 
nating current at 110 v. was applied through variable resistance to a 
Thordarson transformer of type H-1 whose ratio was 1:50. <A con- 
denser made of three concentric aluminum tubes, having a capacity of 
0.0007 microfared, was placed in parallel with the high voltage ter- 
minals. A zinc spark gap with variable adjustment was used in an 
atmosphere of ether vapor. The Tesla coil was composed of four turns 
of wire in the primary and 200 in the secondary with an air core. The 
frequency was measured with a Kolster decremeter calibrated by the 
Bureau of Standards. The wave length varied according to the size 
of the wire in the tube and according to whether the measuring ap- 
paratus was connected to the tube or to the wire. With No. 29 wire, 
the wave length was 185 m. for the tube and 2000 m. for the wire, 
while with No. 16 wire, these values had increased to 260 and 300 m. 
respectively. The frequency, then, was in the range between 1 and 1.6 
million cycles per second. Variations in the width of the spark gap 
and in the power-input had no effect on the wave length. Attempts 
to change the frequency by changing the capacity of the primary or 
secondary circuit of the Tesla coil or by changing the inductance of 
the secondary had no effect on the frequency but resulted in reducing 
seriously the secondary current as well as the corona discharge. Evi- 
dently an optimum set of dimensions of the circuit can be obtained for 
each tube according to its capacity. No attempt was made to work 
out the relationships involved. The electrical circuits used in this work 
were such as to give optimum discharge in the tube used. In other 
4H. Guilleminot, Compt. rend. 136, 1653 (1903) deseribes an apparatus but gives 
no data as to its chemical possibilities. 
5F. O. Anderegg, J. Am. Chem. Soc. 39, 2581 (1917). 
