402 



EEPOKT — ^1886. 



10 cc.) + 78 cc. of water. Tube XIV. was •562 centims. in diameter ; the others 

 were not measured. The tubes were arranged thus : — 



Fig. 8. 



so that the current through any one of them could be measured, with short interrup- 

 tion to the others. The anode vessels contained BaC]2, the cathode vessels NaCl. 

 The E.M.F. of the applied battery was measured by an Ayrton and Perry volt- 

 meter as 29 volts. 



Table op Eesults. 



Late on February 24 the curreui. was stopped because the advancing ions had 

 nearly met, and it was desired to observe whatever phenomenon might accompany 

 their meeting. The tubes were left in position for a day or two, and by February 27 

 it was found that, while the precipitation boundaries due to the current still 

 remained, three other fainter outlines had made their appearance, one, a fresh one, 

 crawling along from the BaCl.j end, the other two advancing beyond the current- 

 formed precipitate. The three are labelled in the table a, b, c respectively, and 

 their positions are indicated ; but no meaning is yet attached to them. 



The result of these experiments is to give a rough absolute determination of the 

 rates of travel of Ba and of CI through the jelly, and to show that the speed of CI 

 is pretty exactly twice that of Ba. Now can this be assumed to be probably due 

 to the fact that barium is a dyad, while chlorine is a monad ? or is it due to the 

 fact that the atomic weight of Ba, 137, is almost twice that of CI.,, viz., 71 ? The 

 simplest mode of examining this question is to replace the chlorine by bromine or 

 iodine, whose atomic weight is quite different, while its valency is the same. 



Similar experiments with iodine instead of chlorine. 



The cathode vessels were filled with a solution of potassic iodide instead of with 

 sodic chloride solution. An E.M.F. of 29 volts was again applied to the same four 

 tuhes, filled with the same jelly as before, and the results are subjoined. The 

 precipitation on the advancing iodine side was, however, double ; one faint outline 

 in advance, which it was suspected might be due to a trace of chlorine impurity 

 in the KI, and the main precipitation, which was considered to be real Agl. 

 Successive positions of the faint advance cloudiness are recorded in the following 

 table in parentheses below the numbers giving the corresponding main advance : — 



