416 



tube, just as all other coiiiiectioiis in the apparatus are made. There 

 was 110 sign of aiij C'O, diffusion inward from the water of the 

 thermostat tliroiigli the riibber-comiectioiis and corks, nor of an 

 0.^ -absorption through the rubber. Blind ex|)eriments, lasting 24 lionis 

 gave no measurable change of titration standard of the lye at tempe- 

 ratures between 20° and 30° C, while the manometer vi, remained 

 al zero Ihroiighoiit that lime. 



e. The oxijgen-supphj and measurement. 



ill order to prevent ozon-formation, a 10 7» natronsolution is to 

 be preferred to diluted sulfuric acid for the electrolysis. 



Ill fig. 6 C is a glass cylinder with natron-lye in which the 

 plalina-eleclrodes />, and /j, are placed. By means of thin platina- 

 wiie these electrodes are fastened by inelling in the glass-tubes 1 

 and 2 respectively. The tubes 1 and 2 pass through caoutchouc- 



Fig. 6. 



