W. G. Mixter — On Electrosynthesis. 55 



The eudiometers were filled with gas for the second experi- 

 ment without removing the residual gas of the first, namely 

 0'61 and 0'58 ce and the increase of 0'1 CC of residual gas in the 

 second experiment shows that almost no ozone was formed. 

 Moreover, the mercury was only slightly tarnished. During 

 half of the time of the first experiment the current was 

 reversed in the primary of the coil with no marked difference 

 in the relative rate of combination in the two tubes. The 

 direction of the current was not changed in the second experi- 

 ment. The lack of uniformity in the results from hour to 

 hour is only in small part due to errors of observation and is 

 probably owing to leakage of electricity from the wires or its 

 passage over the surface of the glass above the jacket tubes. 

 The marked discrepancies during the first three hours of the 

 first experiment show the liability to error in the method. 

 Leaving out these hours, the results accord fairly and the total 

 amount of combination is 20*6 CC and 204 cc respectively. 



It will be observed that the gases in the two tubes were 

 under nearly equal pressures. This is essential, for the higher 

 the pressure the more rapid the combination as shown by the 

 following : the results in each horizontal line were by the same 

 current acting on a hydrogen-oxygen mixture in two eudiom- 

 eters. 



c. c. combined. Meau pressure. c. c. combined. Mean pressure. 



6*22 133 mm 3*87 73 mm 



5*42 160 4*39 110 



21-95 186 13'3 91 



14-43 81 11-25 76 



The reason for these results is not apparent and the rate of 

 combination bears no simple relation to the pressure. 



Vapor of water, as already mentioned, is electrolyzed by the 

 spark discharge, and it is also dissociated by the feeble glow 

 discharge. Three experiments made with the vapor at low 

 pressures in apparatus like that figured gave small volumes of 

 permanent gas. Hence it may be that some of the water from 

 the oxidation of hydrogen is decomposed, but it is probable 

 that little if any is decomposed in a mixture of gases in which 

 water is constantly forming and which contains on account of 

 rapid drying a very small proportion of water. 



In order to determine the ozonizing effect on pure oxygen 

 of the glow discharge, one eudiometer was filled with 36 cc 

 (reduced to 0° and 760 mm ) of oxygen at 193 mm pressure and the 

 other eudiometer with 30'6 CC of hydrogen and oxygen. A 

 saturated solution of potassium hydroxide and iodide was put 

 into each eudiometer to dry the gases and absorb ozone. A 

 current from two storage cells on the primary of the coil was 

 used for an hour and a half. The oxygen contracted 0*3 CC , and 



