Magnetism on a permanent Electric Current. 321 



Length, as weighed =5*68 centims. 



Width „ =1-08 centim. 



Area, including projections . . . =7*15 square centims. 



Weight = *193 grm. 



.*. Thickness (taking sp. gr. =7*71) = '00347 centim.* 



With this plate the following results were obtained: — 

 M. C. ~W"^ 



April 29th, 6680 k x tan 38 37 -127 x 19 9 



■„ 49 13 - 130 „ 



Mean . = -1285 x10 s 



„ „ 49 13 -130 „ 



Platinum. 



One strip of this metal has been used. 



Length, as weighed =6*32 centims. 



Width, „ =1-078 centim. 



Area, including projections . . . =7*57 square centims. 



Weight = *457 grm. 



.*. Thickness (taking sp. gr. =22*1) = '00274 centim. 



With this strip only one series of observations was made ; 

 and that was rather a hasty one. I found: — 



April 28th, 6830 k x tan QQ° 2' 417 x 10 10 



Nickel. 

 There was some difficulty in obtaining a strip of this metal 

 of proper shape for the experiment. The piece used was ob- 

 tained by stripping off the nickel plating from a piece of brass, 

 upon which the deposit had been purposely laid in such a 

 manner as to make it easy to remove. The strip thus obtained 

 was narrow and irregular in shape ; and its thickness cannot 

 readily be determined at present ; so that 1 do not attempt to 



* The plates of very thin rolled iron used were furnished me by Prof. 

 Rowland, who is indebted for a supply of the same to the courtesy of 

 Prof. Lang-ley, of Allegheny Observatory. 



t It is evident that the values of this ratio thus obtained for iron are to 

 some extent, perhaps to a great extent, fictitious ; for of course the 

 strength of the magnetic field within the iron plate itself is the effective 

 strength in the experiment, and this is probably very dilferent from the 

 value of M as determined by means of the test-coil. It seems best, how- 

 ever, for the present to employ this latter value of M, which must bear 

 an intimate relation to the true value, and which has the great advantage 

 of being easily determinable. Nickel has hardly been examined quaiiti- 

 tativelv as yet; and platinum is not sufficiently magnetic to present any 

 difficulty of this sort. 



