108 On the Rotation of Metallic Spheres by Electricity. 



the expansion by the heat, or by both of these, then the phseno- 

 menon of motion (as obtained) becomes immediately intelligible ; 

 for during this period the ball has by its momentum moved for- 

 ward a mmute distance, and the expansion, instead of remaining 

 simply a liftinff power, becomes a propelling one by occurring at 

 a small distance behind the line of centre of gravity of the ball ; 

 or, in other words, the maximum of expansion is always a little 

 behind this point. 



It has occurred to me, that, as the heat produced is that of con- 

 duction-resistance, the velocity of motion might perhaps be in- 

 creased by using a ball composed of inferior conducting metal ; 

 and I have constructed balls similar in size to the one described, 

 but composed of iron, also of nickel-silver. The iron ones were 

 superior to those of copper, and the German-silver ones were 

 better still. The nickel-silver balls invariably overtook the 

 others if placed with them upon the same rails. A copper ball, 

 weighing 376 grains, acted quite slowly, moving only about 19 

 feet per minute, while others of iron or of German-silver, weighing 

 from 800 to 1000 grains each, moved at speeds varying from 30 

 to 42 feet per minute ; and the increase of velocity by increase 

 of conduction-resistance appeared to be capable of still further 

 extension . 



To further increase the velocity of motion, I have introduced 

 into the apparatus the principle of action discovered by Peltier, 

 termed electro-thermancy, by constructing balls composed of one 

 hemisphere of iron and the other of nickel-silver, placing the 

 iron half of the ball upon a nickel-silver rail, and the German- 

 silver half upon an iron rail, and passing an electric current from 

 the iron rail to the German-silver one, to generate heat of electro- 

 thermancy in addition to that of condiiction-resistance at the 

 points of contact. In a number of experiments of this kind, I 

 found that by first passing the current from the iron rail (through 

 the ball) to the German-silver one in order to increase the heat, 

 and then from the German-silver to the iron so as to decrease it, 

 the velocities obtained were in the proportion of about 12| to 14; 

 but this requires further confirmation. 



It is highly necessary in comparative experiments to use balls 

 that are smooth and perfectly free from oxide or greasy matter. 

 Gilding the metals does not interfere with the action, and is an 

 advantage by preventing oxidation. 



8 Broad Street, Birmingham, 

 November 22, 1858. 



Note to Mr. Gore's Faper. 

 The following appears to us to be an analysis of the interest- 

 ing experiment so clearly described by Mri Gore :— Supposcj 



