upon the Dimetisiom of Iron and Steel Bars. 77 



coil twenty-two inches long and one-tiiird of an inch in interior 

 diameter. This coil was secured in a perpendicular position; 

 and the rod of iron, of which 1 wished to ascertain the incre- 

 ment, was suspended in its axis so as to receive the magnetic 

 influence whenever a current of electricity was passed through 

 the coil. Lastly, the upper extremity of the rod was fixed, 

 and the lower extremity was attached to a system of levers 

 which multiplied its motion three thousand times. 



" A bar of rectangular iron wire, two feet long, one quarter 

 of an inch broad, and one-eighth of an inch thick, caused the 

 index of the multiplying apparatus to spring from its position 

 and vibrate about a point one-tenth of an inch in advance, when 

 the coil was made to complete the circuit of a battery capable 

 of magnetizing the iron to saturation, or nearly so. After a 

 short interval of time, the index ceased to vibrate, and began 

 to advance very gradually in consequence of the expansion of 

 the bar from the heat which was radiated from the coil. On 

 breaking the circuit, the index immediately began to vibrate 

 about a point, exactly one-tenth of an inch lower than that to 

 which it had attained. 



" By multiplying the advance of the index by the power of 

 the levers, we have jy^oo^'^ of an inch, the increment of the 

 bar, which may be otherwise stated as ygoVou^'^ of its whole 

 length. 



" Similar results were obtained by the use of an iron wire, 

 two feet long and one-twelfth of an inch in diameter. Five 

 pairs of the nitric acid battery produced an increment of the 

 thirty-thousandth part of an inch ; and when only one pair 

 of the battery was employed, 1 had an increment very slightly 

 less, viz , the thirty-three thousandth part of an inch. 



" This increment does not appear to depend upon the thick- 

 ness of the bar; for an electro-magnet made of iron, three feet 

 long and one inch square, was found to expand under the mag- 

 netic influence to nearly the same extent, compared with its 

 length, as the wires did in the previous experiments. 



" I made some experiments in order to ascertain the lalo of 

 the increment. Their results proved it to be verij nearly ■pro- 

 portional to the intensity of the magnetism and the length of the 

 bar. 



" Trial was made whether any effect could be produced by 

 using a copper wire, two feet long and about one-tenth of an 

 inch in diameter ; but I need scarcely observe that the attempt 

 was unattended with the slightest success. 



" A very good way of observing the above phaenomena is 

 to examine one end oi an eleclro-magnet with a powerful mi- 

 croscope while the oilier end is fixed. The increment is then 



