From the first determination of the electromotive force, made immediately after the casting, I obtained 8=4.008 and Hence the electromotive force directly after the casting was sensibly greater than in the later experiments. About ten days after the casting the reduced deflections of the thermoelectric current were : For No. 1, 134.91 This value is but little greater than that obtained in the later experiments. Probably the molecular change which brought about the diminution of the electromotive and thermoelectric forces had for the most part already commenced when the first thermoelectric experiments were made. Alloy No. 3 and Copper. The constituents of this alloy were 2 parts bismuth and 1 part tin. The diminution of the electromotive force of this alloy was particularly great; for the first determination gave a=94·34, a value nearly twice as much as the one found above. The thermoelectric force was also at the commencement much greater; for the deflections became when reduced: In order to verify the presence of the alteration in question, special experiments were made with this alloy. Three rods were cast at the same time, and soldered to copper wires. Two of these pairs of wires were inserted in the air-thermometer. In the following days the electromotive force was determined, and also simultaneous thermoelectric experiments made with the third pair of wires. Experiment 26, on the 28th of February, directly after the casting. With the current-intensity stan 20° the mean deflection 34-43 was obtained. The reduced deflection of the thermoelectric current was 120.24. Experiment 27, on the same day, about 5 hours later.-Deflection with stan 20° 10′, 30-33. Reduced deflection of the thermoelectric current=113.39. Experiment 28, on the 29th Feb.-To ascertain the value of B, deflections were taken with two different intensities of the We obtained : current. whence B 5.600. This experiment gave the thermoelectric force =103.37. Experiment 29, on March 6.-With stan 20° 13' the deflection in the air-thermometer was 22:35. Thermoelectric force =84.27. = Experiment 30, March 9.-Deflection 19.08 when stan 19° 35'. Thermoelectric force =74-61. The following Table makes evident the constant diminution of both forces. The first column gives the day of the experiment; the second contains the values of a which have been calculated, with the value of ß=56, from Exp. 26-30. The reduced deflections of the thermoelectric current are given in the third column; the fourth is obtained by division of the numbers in the third column by those in the second. Consequently both forces diminish in nearly the same proportion. The alteration proceeded most quickly directly after the casting, the diminution being already sensible after a few hours. It might be supposed that the galvanic current (which, it is well known, calls forth both temporary and permanent changes in the conducting-wire) is the cause of the phenomenon here observed. That this is not the case, however, is shown by the experiments with the wires first cast, in which the change had taken place although they had lain several weeks unused, as well as Experiments 26-30, in which one of the pairs of wires had not been exposed to the influence of any stronger current than the thermoelectric current produced by the cooling of the place of soldering. It must also be remarked that the diminution of the electromotive force cannot be ascribed to any change of structure of the copper wire, because this had been drawn a very long time before the experiments. Alloy No. 4 and Copper. The constituents were 12 parts bismuth and 1 part tin. Calculation gives ẞ=6.866 and a=262-61 from. Exp. 31, Reduced deflection:-For No. 1, =269·91 168.0 169.4 168.6 168.7 The first determination gave the electromotive force = 256.69 and the thermoelectric =266·89, values which differ but little from those obtained later. In order to learn whether this alloy undergoes in the course of time a similar change as the preceding, it was recast and its thermoelectric force investigated. Immediately after casting the reduced deflection was 297-63; a day later, 292-63; and six days after casting, 274-12. These experiments show distinctly that the thermoelectric force (and consequently also the electromotive force) of this alloy diminishes with time. Copper and Alloy No. 5. The constituents were 32 parts bismuth and 1 part antimony. The electromotive force. In the first determination there was obtained ẞ=6.083 and a=307.61 with s = tan 16 14 The two values of a do not differ much from each other. The deflection, reduced, becomes 295.24. The experiment with the pair of wires No. 2 could not be carried out, because in the preparation of the experiment it was injured at the joint. In the first determination values of the thermoelectric force but little differing were obtained for the two |