180 Proceedings of Indiana Academy of Science 
that different tubes of the same design and of similar materials will 
yield different results. 
4. The usual pressure characteristic consists of four parts: 
1. The pressure decrease. 
2. The pressure increase (usually quite rapid) up to at- 
mospheric pressure. 
_3. The pressure increase (having the shape of a temper- 
ature rise curve) above atmosphere. 
4. The abrupt pressure decrease when the power supply 
is disconnected. 
Part 4 always oceurs; parts 1, 2 and 3 are to be found together 
when working with fresh air only, while part 3 only will be found 
when the air has been sufficiently contaminated. 
When the pressure is decreasing the discharge is more quiet and, 
as a rule, the temperature rise less than when the pressure is increas- 
ing. The character of the discharge is also quite different after re- 
versal than before, thus indicating a very definite change in the struc- 
ture of the air. This statement is also borne out by the fact that 
changes in pressure are accompanied by corresponding changes in cur- 
rent flow when the voltage is held constant. 
5. The pressure decrease is not proportional to the nitric oxides 
absorbed. The pressure decrease is always greater than can be ac- 
counted for by the products absorbed, thus indicating the presence of 
some heavier molecule. It is true that some of the pressure decrease 
may be due to some of the gas being driven into the glass, but that 
this would account for the discrepancy observed seems doubtful. 
Whether this heavy molecule is a combination of O and N or a heavy 
molecule of O or N it is not possible to state. It is likely, however, 
that more than one such combination will be found in the effluent gases. 
6. Although not conclusive, the data does show that in most cases 
the pressure decrease with alternating current is more rapid than with 
the induction coil. The 10, 20, and 30 minute runs indicate that alter- 
nating current may also be expected to produce a higher concentra- 
tion of nitric oxides in a given time, and especially is this true for the 
shorter periods of discharge. 
Complete data showing the results of tests not only on the Siemens 
tube but tubes of other design, together with much other material of 
interest relating to the corona discharge, will be found in a bulletin 
of the Engineering Experiment Station of Purdue University, to be 
published in the near future. 
The author wishes to express his appreciation for the assistance 
given him in the prosecution of this work by various members of the 
staff of Purdue University and particularly to Dr. Anderegg. 
Engineering Experiment Station, Purdue University. 
