Density of the Electrolyte at the Electrodes. 279 



per-cent. sulphuric acid. For the purpose of collecting the 

 streaks a capillary glass tube had one end blown out into 

 the shape of a small funnel, and was bent so that it could 

 be lowered into the electrolyte with the finger pressed on the 

 upper end, which projected from the cell, until the funnel 

 part was immediately below the electrode. After a reversal 

 of the current the funnel part was placed in this way under 

 the anode, from which the streaks were flowing. The finger 

 was then removed from the top end of the tube, and the 

 funnel acted as a reservoir for the denser liquid flowing 

 from the anode. As soon as gas appeared at the anode the 

 finger was replaced on the top end of the tube, the latter 

 being then removed and its contents emptied into a test-tabe. 

 This proceeding was repeated many times (at least fifty), and 

 the total liquid collected was tested for hydroxyl by the ad- 

 dition of a few drops of a solution of titanic acid in strong 

 sulphuric acid. No trace of a coloration could be perceived, 

 and thus the presence of hydroxyl was not proved *. 



* The test, which is a delicate one, has been greatly employed by 

 Eicharz in his experiments on the mode of appearance of hydroxyl at the 

 anode in sulphuric acid (Wied. Ann. xxxi. p. 912, 1887). He concludes 

 that the appearance of hydroxyl at the anode is caused by the purely 

 chemical decomposition of the " Ueberschwefelsaure " (Berthelot's S 2 7 -j- 

 H 2 0), which is first formed there. Traube, Berichte der deut. chem. 

 Qesell. xviii. p. 3348 (1888) gives the equation for this decomposition as 



S 2 6 (0 2 )H 2 +2H 2 0=2H 2 S0 4 +H 2 (0 2 ). 

 Eicharz confirms Berthelot's statement that no hydroxyl is formed at 

 the platinum anode unless the strength of the acid is above 60 per cent. 

 As we obtained the streaks in 10-per-cent. acid, the above fact lends 

 additional strength to the view that the streaks are not hydroxyl. They 

 might consist of H 2 S 2 8 (Ueberschwefelsaure), which Bertnelot has shown 

 to be a minor secondary product formed during the electrolysis of sulphuric 

 acid, whether dilute or strong (see Ann. de Chim. et de Fhys. xiv. 1878, 

 and xxi. 1880). The quantity so formed at a platinum anode is small, 

 and would apparently be still smaller at a palladium anode charged with 

 hydrogen if the following reasoning be accepted. Let us first consider 

 the case of an anode uncharged with hydrogen. Assuming that the ion 

 liberated at the anode is S0 4 , the simplest hypothesis to account for the 

 secondary reactions is that part of the S0 4 combines directly with the 

 H 2 S0 4 of the electrolyte to form H 2 S 2 8 , whilst the remainder of the S0 4 

 (probably by far the greater part) acts on the H 2 of the electrolyte, 

 forming H 2 S0 4 and liberating 0. Should, however, the anode be charged 

 with H, the S0 4 would have an additional tendency to combine directly 

 with the H, forming H 2 S0 4 . This tendency would probably greatly 

 exceed the other two, so that if sufficient occluded H be present, it would 

 be expected that neither O would be liberated, nor H 2 S 2 8 formed. It is 

 known that there is no evolution of O under these conditions ; and it 

 would be interesting to estimate the relative amounts of H 2 S 2 8 produced 

 at a palladium anode when charged and wben uncharged with H. The 

 H 2 S 2 8 could be quantitatively determined by the method used by 

 Eicharz (loc. cit. p. 917). 



U2 



