« PreviousContinue »
large anode type will therefore give as good results as the Richards type and is more convenient to use.
The above considerations led the author to the construction of a modified Richards voltameter. Instead of removing the liquid from the bottom, some of the silver crystals deposited by previous runs were placed in the bottom of the cup and the liquid was left undisturbed during the experiments. At first some inconvenience was experienced, due to the rapid growth of the silver toward the anode and consequent sudden changes in the resistance of the voltameter, which prevented the proper adjustment of the current to a constant value. But this trouble was overcome entirely by surrounding the lower part of the porous cup by a small glass cup cut from a test tube or a small beaker. The excellent agreement with Richards's form (see Table 8) shows that also this modification will give the correct silver equivalent.
It is therefore proposed to abandon in future experiments, especially for the absolute measurement of current, the usual form of the silver voltameter and employ either Richards's type or one of those proposed by the author. The differences in the weight of silver obtained by two of these voltameters in series show much smaller variations than those by the old form, and, as will be seen in the following, determinations made by means of independent experiments agree very closely. SILVER NITRATE SOLUTION-DEPOSIT ON PLATINUM AND ON SILVER.
It has been shown by Kahle that an old solution will yield too heavy deposits, its effect increasing as more and more silver is electrolysed. The Reichsanstalt advises, therefore, to use the solution only until 3 grams of silver are deposited from 100 cm of solution. It seemed to be of interest to see whether in the Richards form the solution outside of the porous cup could be used repeatedly. Eight grams of silver were deposited by two runs from the same solution, amounting to about 150 cm, and the solution left standing for several weeks. In the experiments following two Richards voltameters were placed in series, one with a fresh solution, one with this old solution. The latter gave on the average 0.055 per cent more silver. Now, it was thought that it might be possible to obtain a good solution by leaving it in contact with silver crystals. After several weeks more this solution was tried again and gave on the average 0.025 per cent heavier deposits than the Richards form. While an improvement is to be observed, it is clear that keeping the solution simply in contact with silver will not entirely eliminate the complex ion. It is therefore advisable to use always freshly prepared silver nitrate solution. Variations in
4825—No. 1-04 3
concentration have been shown by former investigators
be of no influence upon the weight of the deposit. In our case the electrolyte was always a 20 per cent solution.
Gore“ and Kahle found a somewhat larger deposit when the kathode was used a second time before removing the silver collected in a previous run; and Richards and Heimrod corroborate their statement so far as the filter-paper voltameter is concerned, but found no difference with the porous-cup voltameter. Patterson and Guthe obtained the same deposit on platinum and on silver. Table 9 gives the results of tests made with this point in view.
It is apparent that it makes no difference whether a clean platinum kathode is used or whether the silver from previous runs is left on it. Of course the deposit should not become too heavy, on account of the increasing danger of mother liquid remaining included in the interstices between the crystals.
The explanation for the heavier deposits on silver in the filter-paper voltameter is to be sought in the action of the silver upon the anode liquid.
ELECTRO-CHEMICAL EQUIVALENT OF SILVER.
In conclusion, the results are given of experiments which were suited for the calculation of the electro-chemical equivalent of silver in terms of the electrical standards of the Bureau. According to legal specifications the E. M. F. of a Clark cell at 15° C. is 1.434 volts. The unsaturated Weston cell was found to have an E. M. F. of 1.01954 volts. The resistance standards have been repeatedly intercompared with those of the Reichsanstalt.
In Table 10 the amount of silver deposited by one ampere hour and under the condition that the terminal potential difference of the resist
a Gore: Nature, 27, p. 327; 1882.
ance was equal to the E. M. F. of the Weston cell is given for the different forms.
It is seen that the agreement between these independent experiments is a surprisingly good one, the largest differences amounting to less than 1 in 10,000 in the first three types, as long as the current is equal or larger than 0.5 ampere. For 0.2 ampere the deposits are invariably greater. This is not surprising. The current in these cases was closed five hours, and in spite of all precautions some of the anode liquid probably diffuses through the porous cup during that time. This explanation seems the more plausible from the results obtained in the Richards form with 0.2 ampere. In the first experiment of the four the anode liquid was removed every ten minutes, in the next two every fifteen to twenty minutes, and in the last every half hour. The increase of the weight of the deposit with the opportunity of the liquid to penetrate the porous partition is apparent. The current should therefore be closed not longer than two hours, this of course depending to a certain extent upon the grain of the porous cup.
To obtain the electrochemical equivalent of silver we have to divide the deposit, 4.09888 grams, by 1.01954 X 3,600. This gives 1.11675 mg per coulomb; but the silver was weighed with platinum weights, and
therefore a correction of +0.0069 per cent should be applied for the buoyancy of air. Thus the final result for the silver equivalent is
1.11683 mg per coulomb. With a filter-paper voltameter the corresponding results would be 0.048 per cent higher, or 1.11736 mg. Kahle's comparison between the E. M. F. of the Clark cell and the electrochemical equivalent of silver reduced to the same value for the E. M. F. of the Clark cell would give 1.11726 mg.
In the following table we find a comparison of the values obtained by different observers, all being reduced to the same units. In the case of Perot and Fabry," who used a Clark cell at 0° C. and found its E. M. F. to be 1.4522 volts, using 1.118 mg as the electrochemical equivalent, the difference of 0.0164 volt, given by the Reichsanstalt, has been used to reduce to 15° C., instead of the ratio given by them. If the latter is used their results are identical with von Ettinghausen's.
The large differences in these results can hardly all be due to the silver voltameter alone. No doubt the Clark cell comes in for its share. Carhart's value is based on only two comparisons, with very small silver deposits, and Rayleigh's is the result of only one experiment.
Richards has made an attempt to reduce the absolute measurements of an electric current to the same basis by calculating the electrochemical equivalent under the supposition that the porous cup voltameter had been used, and finds good agreement between the results
a Perot and Fabry: Annal. de fac. de sci., Marseille, 8, p. 201; 1898.
of Lord Rayleigh and Mrs. Sidgwick, Fr. and W. Kohlrausch, Kahle, and Patterson and Guthe. They all agree within 3 in 10,000 and give the mean value of 1.1175 mg.
Since that time two more absolute determinations bave been made. The recent one by van Dijk and Kunst using the filter paper voltameter gives 1.1182 (or 1.1177) mg for the silver equivalent; but Pellat and Luduc obtained 1.1195 mg, or, reduced to the same basis, 1.1188 mg. Thus, the latest determinations of the electrochemical equivalent still show a difference larger than 1 in 1,000.