338 



Mr. G. Gore on the Thermo-electric 



[Feb. 23, 



or, 3rd, to both these influences combined. The more ultimate cause, 

 however, of the phenomena in these cases must be sought for in the mo- 

 lecular movements produced by heat in the metals and liquids. 



The currents obtained with copper plates were no doubt influenced in 

 their amounts (if not also in their direction) by the oxidizing action of the 

 air upon the liquid and metal at their line of mutual contact ; for we know 

 that metals in contact with liquids oxidize much more quickly if oxygen 

 has access to their wet surfaces. And the currents were also influenced by 

 the action of unequal temperature upon this air-contact line ; for we know 

 that wet metals oxidize still more rapidly if heat is applied. 



Influence of line of contact of liquid and metal with the air. 

 That the length of line of contact of the liquid and copper with the 

 air is capable of producing electric currents was shown by the following 

 experiments : — 



Two strips of sheet copper of the annexed form, fig. 5, f inch wide, and 

 Fig. 5. Fig. 6. 



1 2 inches long in the longest limb, were cut from contiguous parts of a 

 sheet of copper, and, after being perfectly cleaned, were coiled into the 

 shape represented by the annexed sketch, fig. 6. They were then placed 

 in a flat-bottomed porcelain dish and connected with the galvanometer, 

 one of the spirals being supported at about J inch higher than the other 

 by means of a triangle of glass rod. The liquid to be examined was then 

 poured into the dish until it just (and completely) covered the lower spi- 

 ral, and the direction and amount of the permanent deflection noted. The 

 positions of the spirals were then reversed and the electrical effects again 

 noted. 



Experiment 1. — With a liquid composed of 100 grains of cyanide of 

 potassium dissolved in 12 ounces of water, whichever of the spirals was 

 only partly submerged and therefore had the longest air-line, was strongly 

 electro-negative to the wholly submerged one. 



Experiment 2. — With a mixture of one measure of strong nitric acid and 

 ten measures of water, deflections of somewhat less amount, but in pre- 

 cisely similar directions to those of experiment I, took place. 



Experiment 3. — With dilute hydrobromic acid the directions of the 

 deflections were also similar, but still less in amount. 



Experiment 4. — With a half-saturated solution of borax very feeble de- 



