124 The Rev. J. Bashforth on the Conducting Powers of Wires 



an angle of 90°, must have been considerable. Mr. Joule 

 seems to have forgotten, that when a body oscillates under the 

 action of any force, that force varies as the square of the numbers 

 of nbrations only under vert/ peculiar circumstances. In this 

 experiment a bar of bismuth, 1|^ inch long, vibrated between 

 the two temporary poles of the electro-magnet \^ inch apart. 

 The force which caused the bismuth to vibrate is supposed to 

 have been dependent on induction; and consequently during 

 every vibration of the bismuth bar through 90°, the force which 

 Mr. Joule was desii'ous of estimating must have gone through 

 very wide variations. It is certain that the angle of vibration, 

 instead of being 90°, ought to have been the least possible for 

 the purpose of obtaining any satisfactory comparisons of forces 

 resulting from a variation in the strengths of the galvanic current. 

 The number of vibrations per minute observed in Exp. I. with 

 assumed currents 1, 2 and 4, were 48, 63 and 96. In Exp. VI. 

 the number of vibrations with a strength of current 2 was found 

 to be 63, as before. On this Mr. Joule remarks, that these three 

 numbers are evidently as the square roots of 1, 2 and 4. Now 

 48, 67'87(=63 + 4"87) and 96 are really as the square roots of 

 1, 2 and 4. If we assume that the forces vary as the square of 

 the observed number of vibrations, we get 1, 1'72 and 4, instead 

 of 1, 2 and 4. On plotting Mr. Joule's results of Exp. I., I 

 obtained three points very nearly in a straight line ; and I am 

 thus led to a law which the experiment does very nearly satisfy, 

 namely, 



301-1-1x16=48-1, 311 + 2x16 = 63 + 1, 



and 311 + 4x16 = 96-1. 



I merely mention this as a fact, without any intention to propose 

 it as a general law. 



The assumed currents of 1, 2 and 4, in Exp. II., gives 4|, 

 9^, and 17 vibrations through a quadrant per minute ; and hence 

 if we supposed the forces of the electro- magnet to vary as the 

 number of vibrations, we should have 1, 2"23 and 4 for the 

 strengths of the galvanic current, instead of 1, 2 and 4. Thus 

 if we sujopose that the commonly received laws hold good, and 

 from Mr. Joule's numbers of observed vibrations, endeavom- to 

 deduce the strengths of the galvanic currents employed, in 

 Exp. I. we get 1, 1-72 and 4, in Exp. II. 1, 2-23 and 4, instead 

 of 1, 2, and 4 in both cases. It is remarkable, that in both 

 experiments consistent results are obtained in the first and last 

 observations ; but that the middle observations in the same two 

 experiments deviate very considerably, in opposite directions, from 

 the desired results. 



On Exp. III. Mr. Joule remarks, that " In this instance we 



