upon the Dimensions of Iron and Steel Bars. 79 



capacity of each division of the graduated tube ; consequently 

 a very minute expansion of the former would have produced 

 a perceptible motion of the water in the capillary tube : but, on 

 connecting the coil with a Daniell's battery of five or six cells 

 (a voltaic apparatus quite adequate to saturate the iron), no 

 perceptible effect whatever was produced either in making or 

 breaking contact with the battery, whether the water was sta- 

 tionary in the stem, or gradually rising or falling from a change 

 of temperature. Now had the usual increase of length been 

 unaccompanied by a corresponding diminution of the diameter 

 of the bar, the water would have been forced through twenty 

 divisions of the capillary tube every time that contact was made 

 with the battery. 



Having thus ascertained that the bulk of the bar was in- 

 variable, I proceeded to repeat my first experiments with a 

 more delicate apparatus, in order, by a more careful inves- 

 tigation of the laws of the increment of length, to ascend to 

 the probable cause of the phaenomenon. 



A coiled glass tube, similar to that already described, was 

 fixed vertically in a wooden frame. Its length was such that 

 when a bar one yard long was introduced so as to rest on 

 the sealed end, each extremity of the bar was a full inch 

 within the corresponding extremity of the coil. The appa- 

 ratus for observing the increment of length consisted of two 

 levers of the first order, and a powerful microscope situated 

 at tlie extremity of the second lever. These levers were 

 furnished with brass knife edges resting upon glass. The 

 connexion between the free extremity of the bar of iron and 

 the first lever, and that between the two levers^ were esta- 

 blished by means of exceedingly fine platinum wires. 



The first lever multiplied the motion of the extremity of 

 the bar 7 '8 times, the second multiplied the motion of the 

 first 8 times, and the microscope was furnished with a micro- 

 meter divided into parts each corresponding to ,jzrV(ytl^ of an 

 inch. Consequently each division of the micrometer passed 

 over by the index indicated an increment of the length of the 

 bar amounting to jjTjVsij^'' of an inch. 



The quantities of electricity passing through the coil were 

 measured by an accurate galvanometer of tangents, consisting 

 of a circle of thick copper wire one foot in diameter, and a 

 needle half an inch long furnished with a suitable index. 



The quantities of magnetic polarity communicated to the 

 iron bar were measured by a finely suspended magnet 18 

 inches long, placed at the distance of one foot from the centre 

 of the coil. This magnetic bar was furnished with scales 

 precisely in the manner of an ordinary balance, and the weight 



