Page:The fundamental laws of electrolytic conduction.djvu/72

 of the cylinder measures 54 mm.; the height of the lower part, 70 mm.; that of the upper part, 25 mm., both inside measurements. The glass is 4 mm. in thickness, as it must be taken rather thick. As the cathode extends only to the middle of the upper cylinder, the effect of diffusion is destroyed in our experiments. During the comparatively short duration of the electrolysis, this will be active only between the layers in the upper cylinder, and will have no effect on the mass in the lower one; it can, therefore, be the cause of no error. Moreover, the motion, which, according to Wiedemann, the electrolyte as a whole experiences from the anode to the cathode, cannot vitiate our results, as it cannot take place under the above conditions. The only error, so far as I can see, which enters into my method and cannot be avoided, arises from the fact that the metal which is separated out by the current has a different volume from that of the salt which is carried away from the upper part. This change in volume is replaced by liquid flowing in or out. The values which we obtain for the transference will be incorrect by the contents of this quantity of liquid. Our error is, however, very insignificant, and may be at least approximately computed. We shall see that even in the case of very concentrated solutions it does not amount to as much as the unavoidable error of analysis. This will be all the more true in the case of the dilute solutions, for, as is readily seen, the error must, in general, diminish proportional to the dilution.

Fig 5 Besides the apparatus, a voltameter was introduced into the circuit. I chose for this purpose the convenient and accurate arrangement described by Poggendorff, called a silver