ANB ITS TEMPERATURE CORRECTION. 73 



and X the horizontal component of the eai-th's magnetic force, the equation of 



equilibrium is 



»j=X8inM (4.) 



The variations of u are observed, wliile the deflecting magnet has its tempe- 

 rature altered 30° or 40° Fahr., by means of hot or cold water ; by differentia- 

 ting equation (4) and dividing by it, these variations are connected with— -by the 

 equation. 



Am ^ ,, - 



=coiuAu (5.) 



m ^ ' 



X and the magnetic declination being constant. 



11. The chief objections to this method are the following : — 



1st, The circumstances under wliich the magnet is placed are considerably 

 different from its usual condition. It is necessary to raise or lower the tempera- 

 ture 30° or 40° in water, within a few minutes, to obtain satisfactory results, 

 whereas the most rapid changes in the magnetometer-box will probably be under 

 2° in an hour. It seems doubtful to me whether it has been proved that the 

 changes of magnetic moment occur as rapidly as those of temperature in all cases. 



2d, In the event of there being any other source of error due to temperature, 

 it is altogether omitted by this method. 



3d, If the correction has not been determined before adjusting the instru- 

 ment, the series of observations is broken up by the necessity of removing the 

 needle. 



12. As it is desirable that the observations of the balance magnetometer 

 should be made as valuable as possible, I shall proceed to consider how this may 

 be best done, as it is my opinion that they will be found ultimately capable of 

 giving diurnal and annual changes with considerable fidelity. 



13. The observations of At, the varying angle formed by the needle and the 

 horizontal, will at present obviously give comparative observations for the varia- 

 tions of vertical force, without reference to the value of the coefficient tan e, until 

 a good approximate value of the latter can be obtained, if the observations in 

 micrometer divisions can be corrected for temperature. In order to do this, it 



would be necessary to convert the value of — , obtained by deflection experiments 



into micrometer divisions, if this value be q. 



We cannot, however, use T, and therefore the method of deflections is, in this 

 way, insufficient ; besides, if the alterations in the value of T from temperature 

 should be caused by changes in the position of the centre of gravity, this change 



VOL. XVI. PAET I. T 



