OF ARTS AND SCIENCES. 209 



iron disk, and thus mask the heating effect peculiar to it. That the 

 generation of heat in the wire of the magnets had no influence was 

 proved by passing a steady current through the magnets for several 

 minutes, when no deflection was observed in the galvanometer. To 

 avoid the heating effect of the cores, the smaller iron disk was placed 

 in a glass tube whii-h exactly fitted it, as in the preceding case. The 

 ends of the tube were tightly closed, and it was placed in a larger 

 tube, 2.7 cm. in diameter, parallel to the axis. A stream of water 

 was passed through the outer tube. After several attempts, however, 

 it was found that the water pressure was so variable that the arrange- 

 ment would not answer. The larger tube was then tightly packed 

 with asbestos, a layer of about one centimeter in thickness surround- 

 ing the smaller tube. It was thought that this arrangement would 

 certainly prevent any communication of heat during the three minutes 

 that the experiment lasted. 



The result of this experiment was the same as before. The deflec- 

 tion due to the heating was unchanged by changing the positions of 

 the magnets. The deflection was about 0.1 cm. smaller than in the 

 preceding cases, probably on account of the smaller size of the disk. 



The experiments were then repeated upon a copper disk of about 

 the same size as the larger iron disk. No heating effect whatever was 

 observed. It must be remembered, however, that in this case the 

 thermo-electric element consisted of copper and German silver, the 

 electromotive force of which is much less than that of iron and Ger- 

 man silver. 



The battery used in these experiments consisted of a modified form 

 of Bunsen cell. The outer vessel contained dilute sulphuric acid ; the 

 inner vessel contained a solution of 150 gr. of bichromate of potash 

 and 450 gr, of sulphuric acid to one litre of water. This cell was 

 found to be fully as constant as the Bunsen. 



The inteusities of the magnetic fields were measured, after the ex- 

 periments were over, by placing a small coil of copper wire before the 

 poles and reversing the current. The current was somewhat less than 

 when the cells were first set up. 



The intensities of the magnets were respectively 2,033 and 760 

 times the horizontal intensity of the earth's magnetism. In the 

 C. G. S. system the intensities were 343.57 and 128.44. 



The results of the preceding experiments show that the chief heat- 

 ing in a mass of iron, that is alternately magnetized in different direc- 

 tions, is not due to molecular motion ; at least, if this motion is similar 

 to that assumed by "Weber's theory. 

 VOL. xviii. (n. s. X.) 14 



