for Voltaic Electricity, ^c. 123 



Table IV. 

 Copper wire. Iron wire. 



E = 125-1845, 11=0-58582. E = 23-370, Il = -104. 



The agreement between calculation and experiment above 

 shown is far nearer than could have been expected, considering 

 the many sources of error. The experiments must have been 

 very carefully conducted. 



Lenz in 1832 showed by experiment with convolutions of wire 

 of diameters 0"73 inch, 6"57 inches, and 28 inches, that the elec- 

 tromotive power which magnetism produces in them remains the 

 same {Mem. Acad. St. Pet.; translated in the Scientific Memoirs, 

 vol. i. p. 607). Again, in 1838, Lenz and Jacobi published ac- 

 counts of experiments, which showed that the magnetism excited 

 by the galvanic current in a long bar of iron, by comparatively 

 short spirals surrounding it, was, for a given strength of cun-ent, 

 independent of the diameter of the spiral. It appears that 

 Mr. Joule was guided by this principle in the construction of 

 his powerful electro-magnets. Still it must not be concluded 

 that a wide convolution is on the whole as advantageous practi- 

 cally as a smaller one. The resistance of a wire varies as its 

 length ; consequently the wider the spirals, the more the strength 

 of the galvanic current is diminished for a given number of con- 

 volutions. Thus with Mr. Dresser's battery, and the iron wire 

 of Exp. I., the efi"ect of one circuit, 1 foot long, would be 330 ; 

 of one 2 feet long, 280 ; of one 3 feet long, 240, &c. ; and of 

 one 28 feet long, 60. 



It appears that Mr. Joule took no precautions for ascertaining 

 the exact strengths of the galvanic current with which he expe- 

 rimented. The arrangements of the sixteen cells which he adopted 

 were certainly calculated to produce currents as 1, 2 and 4 ; but 

 the disturbing causes arc so numerous, that in every case an 

 actual measurement was absolutely uecessauy. Again, the num- 

 ber of observations in each experiment is far too small ; for 

 within the limits of errors which Mr. Joule appeal's to allow, it 

 would be possible to confirm a great variety of laws. In Exp. I. 

 the arc of vibration is not given, but in Exp. II. it is said to have 

 been as large as a quadrant of a circle. Such a circumstance as 

 this deserved some exjdanation. The resistance of the air to the 

 bar of bismuth, making seventeen vibrations per minute through 



