AS A MEASURING INSTRUMENT. 405 



process described, no matter how involved the relation may be, the 

 values of both the elements can be determined as accurately as the 

 readings of the scale will allow, and at as short intervals as may be 

 desired, so that all necessity for uncertain interpolations is removed. 



I have thus far, in common with all physicists who have investi- 

 gated this subject, considered only one-half a R of the resultant curve. 

 But it is easy to imagine that the half which lies on the other side of 

 the magnetic meridian cannot, on account of a possible want of sym- 

 metry in the form of the coils, or in the position of the needle, be 

 unconditionally assumed as identical with the first half. If, therefore, 

 it be desired to use the deflections of the needle on the other side of 

 the meridian for comparative measurements of the force of the cur- 

 rent, caution requires that the form of the second half of the curve, 

 and the scale of intensity based on it, should be determined in the 

 same way as the first. 



It must also not be forgotten that the effect of the spiral on the 

 needle varies with the height of the latter within the former, though 

 in the middle, where the maximum lies, a slight change in the height 

 has no very perceptible influence on the result. For the sake of 

 security it is therefore advisable not to change the height of the 

 needle after having constructed the scale of intensity, and if it should 

 have been accidentally moved, to construct the scale anew. The 

 fixed index may serve to detect any change of height. 



The same must be done if there is any reason to fear that the mag- 

 netism of the needle has changed perceptibly. The magnetism of a 

 simple needle may, it is true, grow stronger or weaker without any 

 effect on the measurements, for the action of the current on the needle 

 then increases or diminishes in the same ratio. But the distribution 

 of magnetism in the needle must not be changed thereby, i. e. , the 

 magnetic intensity of every point of the needle must change in the 

 same ratio, which cannot be assumed with certainty. If the galvan- 

 ometer be furnished, as is usual, with two needles acting in opposite 

 directions, its indications may be changed without any change in the 

 magnetic distribution by a variation in the relative magnetic intensity 

 of the two needles. 



For this reason it is absolutely necessary to test the scale of inten- 

 sity of the instrument occasionally, especially after the passage of 

 powerful currents through it; and therefore the mode of testing and 

 correcting it must be a simple one. In this respect the method 

 described leaves nothing to wish for. Prolix as the description of it 

 may seem, in practice it is as simple as it is convenient. Half an 

 hour is quite sufficient to make the measurements necessary for the 

 construction of a table like that given on page 402, which, if the 

 instrument has a steady position, and the force of the current does not 

 vary, are as accurate as could be desired. 



Besides its convenience, this method has the great advantage of 

 assuming nothing whose correctness is not perfectly evident. The 

 method is certainly an empirical one, depending upon experimental 

 data of a complicated nature; but in the use of these data it is strict 



