KINETICS OF CERTAIN SILVER SALTS 277 



drawn out at B while the nozzle C was dipping under the sur- 

 face of the solution to be used. By this process, the main- and 

 side-tubes were filled up to the level of the top of the latter; 

 the sizes of apparatus we employed contained 40 and 60 cc. 

 respectively. The nozzle was now likewise closed with a rubber 

 cap, and the whole apparatus suspended in one of Profes- 

 sor Ostwald's thermostatic water-baths, 1 in such a fashion 

 that the nozzle protruded over the rim, while the liquid 

 was entirely within the water-bath. After it had acquired 

 the temperature of the bath, the circuit was closed. As soon 

 as the electrolysis was concluded, the nozzle was opened and 

 suitable portions of the liquid were filled into tared vessels, 

 by blowing steadily through B. These portions were weighed 

 and analyzed. The portion remaining in the apparatus was 

 estimated by the latter's gain in weight, and likewise analyzed. 

 Provided there has been no mixing through diffusion and con- 

 vection during the electrolysis, the first portion taken from 

 the apparatus, if the liquid is properly divided, will consist 

 of the stratum around the anode, which has become more 

 concentrated, and of a sufficient quantity of the unaltered 

 middle layers to insure a rinsing of the adhering parts of the 

 lowest stratum. The succeeding portions, being composed 

 wholly of the middle layers, must show an unaltered compo- 

 sition, while what is left in the apparatus includes the diluter 

 strata about the cathode. To prove the reliability of our ex- 

 periment, we must find in the first place that the middle 

 layers had remained unchanged, and secondly, that the 

 gain of the lowest layer exactly counterbalanced the loss of 

 the highest, for the mean composition of the whole liquid 

 undergoes no change. After finding that the first condition 

 was invariably fulfilled, we saved time by giving up the sep- 

 arate examination of the middle layers, and divided the whole 



1 As described recently by Professor Ostwald, Zeitschr. 2, 565. 



