and Experimental Researches therewith. 53 



pared are loud. I therefore prefer to balance the metal under 

 examination by means of a similar mass placed in the opposing 

 coil, reading on the sonometer the differences of sound, which 

 are then slight. Experience has shown that the most accurate 

 results are obtained when the sonometer is replaced by a gra- 

 duated strip of zinc about 23 millims. wide, 200 inillims. long, 

 and tapering from a thickness of 4 millims. at one end to a 

 fine edge at the other, and superposed in a horizontal plane 

 over the opposing coil ¥, the metal to be tested being in a 

 plane midway between a and b on the left of the plate. 



The delicacy of the readings may of course be greatly in- 

 creased by diminishing the angle between the two faces of the 

 strip ; but there are many points connected with its use which 

 would be too long to describe in this paper. 



As a rule three Daniell elements will be found quite suffi- 

 cient ; and even this weak force is so exquisitely sensitive that 

 it will find out the smallest fraction of difference in weight or 

 structure of metals. Thus two silver coins such as a shilling, 

 both qu'te new, and both apparently of the same weight, will 

 be found to possess a difference of weight which the instrument 

 at once indicates. 



The following experiments will show its exceeding sensitive- 

 ness and its wide field of usefulness as an instrument of research. 



I. If we introduce into one pair of the induction-coils any 

 conducting body, such as silver, copper, iron, &c, there are 

 set up in these bodies electric currents which react both upon 

 the primary and secondary coils, producing extra currents 

 whose force will be proportional to the mass and to its spe- 

 cific conducting-power. A miligramme of copper or a fine 

 iron wire, finer than the human hair, can be loudly heard and 

 appreciated by direct measurement, and its exact value ascer- 

 tained. We can thus weigh to an almost infinitesimal degree 

 the mass of the metal under examination : for instance, if we 

 take two English shilling pieces fresh from the Mint, and if 

 they are absolutely identical in form, weight, and material, 

 they will be completely balanced by placing one each in the 

 two separate coils, provided that for these experiments there 

 is an adjustable resting-place in each pair of coils, so that each 

 coin may lie exactly in the centre of the vacant space between 

 the primary and secondary coils. If, however, these shillings 

 are in the slightest degree worn, or have a different tempera- 

 ture, we at once perceive this difference, and, if desired, 

 measure it by the sonometer, or by lifting the supposed 

 heaviest coin at a slight distance from the fixed centre line : 

 the amount of degrees that the heaviest coin is withdrawn 

 will show its relative mass or weight as compared with the 

 lightest. I have thus been able to appreciate the difference 



