= 
Siemens Ozonizer Lay 
more than eight times the concentration of NO was obtained at the 
reversal point on run D-26, which was made under the same conditions 
as D-25. 
Several runs were made to determine whether or not the pressure 
rise could be duplicated, and it was found that it could be reproduced 
as desired, using either the induction coil or the alternating current. 
Whether or not the same result would have taken place in a new tube 
of different construction is not known. 
The experiments made on fresh air compared with those made 
with air which has stood: in the tube for a considerable period show 
a very marked change. When the air is not fresh in the tube, but has 
stood for some time in the tube following the last application of the 
high voltage discharge, the pressure rises when voltage is applied, and 
no appreciable quantity of ozone or nitric oxides are to be found. The 
corona discharge is more noisy and appears to consist of many sparks 
and points. This condition is much like that after reversal when fresh 
air is used, for immediately upon reversing the discharge changes from 
a quiet blue glow to noisy streamers and condensed discharges, which 
condition increases as long as the pressure increases. Although the 
shape of the pressure rise curve differs somewhat when using stale air 
compared to the rise above the initial pressure when using fresh air, 
yet, all the evidence obtained goes to suggest that in some manner the 
air standing in the tube is carried through the equivalent of a reversal 
and subsequent pressure rise. The air standing in the tube may well 
be affected by some traces of the products of the previous run, this 
action being catalytic in its nature. Such traces may remain occluded 
in the glass or in a very dilute state in some air pocket. 
Run No. D-28 was made in an effort to secure more information 
concerning the way in which the contamination took place. This run 
followed D-27 with an interval of 23 hours. The air was swept out 
of the tube following D-27 in the usual manner, allowing about 15 
minutes for absorption. The tube was then closed up for 3 hours, after 
which fresh dry air was blown through at a rapid rate for 3 hours. 
The tube was again closed up, and the next day, after an interval of 
17 hours, run D-28 was made. Reversal took place at 30 minutes after 
the pressure had decreased 28 mm. Titration at this point gave a con- 
centration of 0.4 per cent NO, while but a trace of ozone was observed. 
From this experiment it appears that the effect of contamination 
is reduced by sweeping out the products after a run, using a large 
quantity of air. In case the tube stands for a considerable length of 
time, even though the tube has been carefully swept out, the yield is 
materially affected as in D-28, where the yield was about one-eighth of 
what it was in D-29, which was a check run with fresh air. Thus, 
even small traces of the previous runs serve to greatly reduce the 
yield. Curves showing the pressure changes for D-28 and D-29 will 
be found in Fig. 6. 
The results from the use of a discharge tube not only vary with 
the design but also vary greatly with tubes of the same design. A 
second tube built of pieces of the same tubing and having practically 
all the same dimensions as the tube described in the first part of this 
12-20320 
