398 



Dr. C. Chree. Collimator Magnets and the 



head is turned through ±180°. There are several possible criticisms. 

 In the actual vibration experiment, the thread is twisted at most only 

 a degree or two, and it is open to doubt whether a value found for © 

 from twists through ± 180° is strictly applicable, even supposing the 

 conditions otherwise identical. 



There is usually more than one silk fibre in the suspension, and this 

 increases the probability that the value found statically for © may not 

 apply exactly to a vibration experiment. 



Another criticism is that experiment really gives ©/m'X', where m' 

 and X' are the values of the magnetic moment and of the force during 

 the torsion experiment, instead of 0/mX, where m is the moment at 

 0° C. and X the force during the vibration experiment. 



To judge of the effect of this, we require the relation between the 

 error in ©/w.X and the consequent error in X. For this we have 

 from (5) 



SX = -XJS(@/mX). 



In order that SX should equal ±5 x 10~ 6 at Kew we would require to 

 have 



8 (®/mX) .= ± 5 x 10~ 5 , approx. 



When one uses a suspension sufficient for the collimator magnet 

 alone, without the auxiliary bar, one can with an average magnet get 

 @/mX as low as 4 x 10~ 4 at Kew. Now an error of 10 per cent, 

 in ©/mX, through neglecting the variation of m with temperature or 

 the fluctuations of X at a fixed station is quite out of the question. 



On the other hand, suspensions such as are frequently used to carry 

 the auxiliary bar as well as the magnet, and are intended to stand a 

 good deal of rough handling, may make ®/mX at Kew as large as 

 25 x 10 _i . In such a case error may occasionally arise through not 

 discriminating between m and m. 



In practice the most probable sources of error are inaccuracy in the 

 torsion experiment and variation in ©, owing to variation of moisture, 

 between the vibration and torsion experiments. 



§ 35. Temperature Coefficients. — In the method of determining tem- 

 perature coefficients in vogue at Kew the collimator magnet is fixed 

 inside a wooden box, rigidly attached to the pillar which carries the 

 unifilar. The calculation assumes the mirror magnet — which is sus- 

 pended as in the deflection experiment — to be exactly perpendicular to 

 the collimator ; but, owing to the position of the latter being fixed, this 

 is in general only approximately true. Suppose the deflected magnet 

 to make an angle u with the magnetic meridian, and an angle |-7r ± \f/ 9 

 instead of tt/2, with the collimator magnet. The deflecting force F is 

 given in terms of X by the equation 



F/X = sin u cos ip. 



