( 708 ) 



Tliese i\i\U\ may Ite sutruMciit lo form iiii opinioji about the instru- 

 ment iu theory and to give an idea of its litness for practical wori^; 

 yet on this hitter }»oint tiie api>licati(»ns alone can give full and 

 convincing evidence. 



In what follows we inle.jid to mention some of these api)lications. 



Where the object is to measure xovy feeble currents no other 

 galvanometer seems to e()ual the instrument we are considering. It 

 is obvious that theoretically there is no limit to the sensitiveness of 

 any arbitrary galvanometer for constant currents. One can indefinitely 

 increase the period of oscillation of the magnets as well as the 

 (lislance of the scale and so obtain any desired sensitiveness in theory. 

 l)ut i)ractical difficulties soon (U-aw a limit. One among other diffi- 

 culties is the inconstancy of the zero-point, which is influenced by 

 many circumstances and which causes the more trouble the more 

 the period of oscillation increases. 



This is probably the reason why an electrometer is preferred to a 

 galvanometer when \ei'y feeble currents have to be measured, e. g. 

 when great insulating resistances have to be examined or the ionising 

 power of radio-active substances. 



In the celebrated investigation by Mr. and Mrs. Clrie '), which led 

 to the discovery of radium, the radio-activity of various materials 

 was judged by their power to render air conductive; and the conduc- 

 tivity of the air was measured by means of an electrometer. The 

 electrometer had to be chargetl by a current, which passed through 

 a, conducting layer of air, the rate at which the electrometer was 

 charged being a measure for the current. 



Evidently it was not an easy matter to measure currents in this 

 way ; so Mr. and ^Irs. Curie preferred a method of compensation 

 by means of a rod of piezo-electric quartz. Tlie charge received l)y 

 the electrometer through the layer of conductive air was compensated 

 by a contrary charge deri\'ed from the quartz-rod. To effect this the 

 rod was subjected to a steadily increasing pull by continuously adding 

 weight to a scale suspended on the quartz-rod. In this way the 

 image of the mirror of the electrometer had to be kept at zero, the 

 increase of the pull during the time being the measure for the current 

 and in this case also for the coiuhiclivity of the air. 



It is much easiei- to make these measurements with the string 

 galvanometer. 



I connected the instruiiient with two brass plates .4^ and A^, fig. 1. 



1) See e.g. Mme SKLODO^v^:KA Glrie, Recherches sur les substances radioactives. 

 Annnlos ÓV Cliimir el .!.■ l']iysi(|ue 7, T- 30, p. 99. 1903. 



