452 Prof. F. Kohlrausch on the Bifilar 



Corrections for the Length of the Rod. — It would not be 

 practicable to make the distance of the magnet from the needle 

 so great that the length of the bar might be neglected ; since 

 for a considerable distance there would exist no angle (j> 

 which would bring the needle into a transverse position. We 

 may take account of the length of the rod by multiplying the 

 above result by 



l+i^(3 + 5cos 2 0). 



In this a is the distance of the middle point of the- magnet 

 from the needle, X the distance apart of the poles of the 

 magnet — that is, in the case of bar-magnets, five sixths of the 

 length of the bar. 



IV. On the J3i filar Determination of Moments of Inertia. 



Gauss's well-known method of determining the moment of 

 inertia of a body from its oscillations, when loaded and in the 

 unloaded condition, eliminates the unknown directive force of 

 the oscillations. The employment of a known directive force 

 {e. g. of a bifilar* directive force) permits the employment of 

 a simpler method. 



I was induced to investigate this method of determination 

 by obtaining, by means of suspended weights, inconsistent 

 measurements of the moment of inertia of a magnetic bar, and, 

 in spite of the greatest care, a result too great, as could be 

 concluded from geometrical measurement of the carefully 

 worked bar. 



The method was carried out by means of the bifilar- suspen- 

 sion arrangement recently described by mef . We determine 

 first the moment of inertia of the bifilarly-suspended carrier 



* Kohlrausch, Wied. Ann. xvii. p. 744 (1882). 



t By adopting the extremely convenient method of M. Wild, in which 

 the oblique surfaces over which the wires run are replaced by surfaces at 

 right angles to which the wires are clamped (JRep.f. Meteor, d. K. Acad, 

 d. Wiss. St. Peternb. vii. No. 7, 1883). 



I may be allowed to remark here that I am quite unable to regard my 

 method of the " bifilar " measurement of the earth's magnetic intensity as 

 a combination of Gauss's method with that of "Wild, as M. Wild appears 

 to do (loc. cit. p. 2). I believe that I may claim for my method that it is 

 an altogether independent one. For the " bifilar galvanic method," out 

 of which the " bifilar magnetic " has been evolved, is found described, long 

 before the first of the publications of M. Wild on this subject, and with 

 account taken of the torsion of the threads, in the third edition of the 

 Leitfaden der praktiscfien Phydk, 1877, p. 184. The first bifilar-magnetic 

 method described by Wild in the year 1881 is, further, altogether different 

 from mine both in arrangement and in object. Wild's second method is 

 more nearly like mine, but it was unknown to me, since it was published 

 later than mine. 



