354 Prof. A. Schuster and Mr. A. W. Crossley. [Jan. 21, 



We traced the cause of the anomalous results shown in this table. 

 The anode of the small bowl had by repeated use been gradually 

 dissolved ; the current density was consequently increased. Under 

 these circumstances the current becomes unsteady, polarisation effects 

 make themselves apparent, and the deposits are no longer trust- 

 worthy. The deposits taken when the current density is too great 

 have generally a yellow colour. We are reminded of some old expe- 

 riments in which by increasing the current density black deposits 

 were obtained on the kathode, which at one time were supposed to 

 be a hydride of silver. Poggendorff is generally stated to have 

 proved that the black deposit is not a compound, but silver in a 

 finely divided state. On referring to Poggendorff's paper, his experi- 

 ments do not seem convincing, and he has expressed himself with 

 more caution than those who quoted him afterwards. He states, 

 however, that the black deposit often suddenly changes into a light 

 one. Some observations made by Mr. Hoskins Abrahall in the 

 Owens College Laboratory, as well as our own experiments, lead us 

 to believe that it is the current density at the anode more than that 

 at the kathode, which introduces the anomalous results. When the 

 "deposits are thus untrustworthy, the current, as far as we are able to 

 judge, is always unsteady, so that no danger arises when the silver 

 voltameter is used for the calibration of instruments. 



At this stage of the inquiry we introduced a second bell-jar and a 

 second voltameter of approximately the same size as the small one 

 previously used. The balance was also changed, and the weighings 

 were taken on a new short-beam balance. This balance was unsteady 

 in its indications after first setting up, and a sudden change of zero 

 while one of the basins was being weighed renders the result of Experi- 

 ment 17 doubtful. The numbers obtained in this experiment are 

 therefore included in square brackets in our tables. Table YII gives 

 the comparison of the deposits in air and in vacuo taken in basins of 

 nearly the same size. A glance at the numbers can leave no doubt 

 as to the reality of the increase in the deposit under reduced pres- 

 sure, although the amount of the increase is a little uncertain. There 

 is only one case (Experiment 23) in which the deposits are prac- 

 tically identical, and in that case it was noticed that the deposit in 

 vacuo was yellow an indication that the current density was probably 

 just a little too large. The average difference between the deposits 

 is about one part in two or three thousand. 



