58 MR. F. E. SMITH, MR. T. MATHER. AND DR. T. M. LOWRY 



filter pump, but if the liquors contain much impurity it is advisable to rinse the 

 crystals cautiously with a few cubic centimetres of iced distilled water. 



(5) If it is desired to dry the crystals, the best method is to make use of a 

 HEMPEL'S vacuum desiccator charged with stick potash, and cautiously heated over a 

 water-bath to accelerate the drying. As a rule, however, it is best to use the moist 

 crystals for preparing solutions, the exact strength of which can, if necessary, be 

 determined by evaporating a known weight of the solution. 



(6) The water used in the earlier experiments for the final crystallisation of the 

 nitrate and for the preparation of the solutions was a specially pure sample prepared 

 at Hendon by Mr. W. R. BOUSFIELD, by a process of continuous fractional distilla- 

 tion, and stored in a large Welsbach bottle. Its electrical conductivity had been 

 measured and found to be only 1 reciprocal megohm per centimetre cube. Subsequent 

 experiments showed that commercial distilled water could generally be used without 

 introducing any error. 



D. Effects produced by Repeated Electrolysis. 



That an increase in the value obtained for the electrochemical equivalent of silver 

 may result from repeated electrolysis of a silver nitrate solution was first observed by 

 NOVAK,* and later by RODGER and WATSON,! KAHLE,^: VAN DIJK and GUTHE.|| 

 RODGER and WATSON record as successive relative values the numbers 



9983, 9987, 9990, 9999, 9995, 9993, 9995, 10002, 10005, 10006, 10002. 



In an attempt to confirm these observations we repeatedly electrolysed two solutions 

 of silver nitrate, the one being contained in a platinum bowl and the other in a silver 

 bowl. The resulting values of the electrochemical equivalent were, 



when the platinum kathode was used : and when a silver kathode was used : 



1-11827 (14a) 1-11825 (146) 



1-11822 (156) 1-11826 (15a) 



1-11822 (166) 1-11825 (16a) 



1-11827 (17 c) 1-11833 (176) 



1-11838 (18a) 1-11840 (186) 



1-11834 (21a) 1-11830 (21c) 



It thus appears that there may be a small increase, but nothing comparable with that 

 observed by RODGER and WATSON. 



* NOVAK, ' Proc. Roy. Bohemian Ac. Sci. Prague,' 1, pp. 387-432, 1892. 



t RODGER and WATSON, 'Phil. Trans.,' A, 186, p. 631, 1895. 



{ KAHLE, 'Zeitschr. lust.,' 18, pp. 229-267, 1898. 



VAX DiJKand KUNST, ' Ann. der Phys.,' 14, p. 569, 1904. VANDlJK, Ann. derPhys.,' 19, p. 249, 1906. 



|| GUTHE, ' Phys. Rev.,' 19, p. 138, 1904 ; ' Bull. Bureau of Stands.,' vol. 1, p. 355, 1904. 



