and Spheres by Electricity. 



521 



mechanical process termed " spinning." Two hemispherical 

 cups were thus formed, and were reduced to the requisite degree 



Sectional view. Scale ^rd. 



Perspective. 



of lightness and to equal weights by repeatedly filling them with 

 weak nitric acid ; they were then soldered together, and their 

 remaining slight inequalities removed by mechanical means. 



In all cases yet observed, the motion has been attended by a 

 peculiar crackling sound at the surfaces of contact, and by heat- 

 ing of the rolling metal ; and in experiments upon a large scale 

 with thick tubes, strong vibrations accompanied by emission of 

 musical sounds were observed, similar in a moderate degree to 

 Trevelyan's experiment with heated metals. In a dark place 

 electric sparks appeared occasionally at the points of contact, 

 especially as the tube or sphere rolled /ro??i the spectator. The 

 motion did not occur either when the rolling metal was too light 

 to maintain sufficient contact with the rails, nor when it was 

 too heavy ; also it did not occur when the electric current was 

 too feeble, nor when it was too strong and pai'tly fused the 

 metals at their points of contact. Tubes of small diameter acted 

 less powerfully than larger ones. 



The revolution of the ball does not appear to be an effect of 

 Phil. Mag. S. 4. No. 103. Suppl. Vol. 15. 2 M 



