The Reduction of Osrnic Acid by Lipoids. 17 



The purified liydrogen was next passed tlnough a haixl glass 

 tube containing the black substance carefully weighed into a 

 porcelain boat. The gas was allowed to pass through this for ten 

 minutes to displace air, and the tube containing platinized asbestos 

 was heated to redness, and was kept heated throughout the rest of 

 the experiment. The gas was allowed to pass for another ten 

 minutes, and the tube containing the substance M^as then connected 

 with a carefully weighed I'^-tube with glass stopcocks, containing 

 phosphorus pentoxide. The combustion tube was then heated ta 

 redness and maintained at a red heat for fifteen minutes, with 

 hydrogen passing through. A guard tube containing calcium 

 chloride was attaclied beyond the weighed phosphorus pentoxide 

 tube to prevent any possible diffusion of moisture from the air. 

 Since dry hydrogen or air was always passing through the 

 apparatus when the stopcocks of the P2O5 tube were open, it was 

 not necessaryto use a PaOg guard tube, and calcium chloride was 

 sufficient. 



The tube was allowed to cool in the current of hydrogen until 

 it was possible to handle it, and the hydrogen was then displaced 

 by a current of dry air passing for half an hour. The possibility that 

 some water was formed by catalytic action of metallic osmium 

 when the hydrogen was displaced by air, was negatived by repeating 

 this part of the operation with a weighed phosphorus pentoxide 

 tube attached. No increase in weight of this tube was found. 

 After the hydrogen had been displaced by air the boat and the 

 drying tube were weighed. 



Since the entrance of air into the apparatus was of uo account 

 except after the tube containing platinized asbestos, ordinary 

 rubber stoppers were used up to this point. The connexion 

 between the drying tube after the platinized asbestos and the hard 

 glass tube containing the boat was made by a small rubber stopper. 

 Minute quantities of air might leak into the apparatus by diffusion 

 through this stopper. Experiments showed that when the tube 

 was heated under the same conditions as in the actual experiment, 

 but without the boat, there was, in fact, a very slight gain in weight 

 of^the phosphorus pentoxide tube. After the whole apparatus 

 had become perfectly dried by long passage of hydrogen, this 

 increase in weight was practically constant for the period of opera- 

 tions described, and amounted in total to 0-00081 grm. This 

 weight was subtracted from the gain in weight of the tube in the 

 actual experiment. We are inclined to the opinion, from previous 

 experience in similar work, that this correction is probably smaller 

 than would have been the case if glass joints had been used 

 under the conditions of the experiment, so that more complicated 

 apparatus would not have offered any advantages. 



The results of the single experiment, which it was possible to 



c 



