( 211 ) 



We must remark, however, that the galvanometer loses in sensi- 

 tiveness when its deflections become quicker. If the time of a 

 deflection which is always kept at the limit of aperiodicity is 

 reduced a times, the sensitiveness diminishes a^ times. 



Now it sometimes happens that very rapid alternations of current 

 must be measured, which are so small, that they require a sensitive 

 position of the galvanometer in order to be perceptible and measur- 

 able and in these cases the condenser method can no longer be 

 applied . 



If under these circumstances the deflections of the galvanometer 

 are less rapid than the oscillations of the current which must be 

 studied, the curve, recorded by the quartz wire, will no longer 

 render these oscillations accurately. We must then, as has also been 

 done with many capillary-electrometric curves, calculate from the 

 properties of the instrument used and the data of the directly 

 recorded curve, the value of the real oscillations. This calculation 

 may lead to the construction of a new curve, which now in all 

 particulars is the correct expression of the oscillations. 



The manner in which capillary-electrometric curves may be cal- 

 culated, is known ^). Do the curves, recorded by the string galvano- 

 meter, fulfil the same conditions and is it possible to apply the same 

 calculation to them? 



The answer to this question must in general be negative. We 

 shall try in this paper to analj^se the curves of the string galvano- 

 meter and to show that they admit of equally accurate constructions 

 as the capillary-electrometric curves, but that the manner in which 

 they are calculated is different. 



In this connection some propei'ties of the galvanometer, more 

 especially the tension and mass of the string and the resistance to 

 the motion of the string must be discussed. We shall try to express 

 the value of these three quantities in absolute measure, by which a 

 clearer light will be shed on their significance for the practicability of 

 the instrument generally- and for the modifications which might 

 advantageously be made in order to render the galvanometer service- 

 able for special purposes. 



2. The principles of the method. 



In the analysis of the curves, recorded by the string galvanometer 

 we start from the assumption that the resistance offered by the air 



1) For literature on tliis subject vide ; Pflüger's Arcliiv f. d. ges. Physiol. 

 Bd. 99, p. 472, 190:3. 



15* 



