xlviii Introduction to the Makeestoun Observations, 1844. 



constant, aud therefore, tliat a X = 0, or that it varies in one direction only, in- 

 creasing continuously, or diminishing continuously, throughout the period selected, 

 and, therefore, that the sign of A X is the same for both equations. In the hitter 

 case, it is evident that by taking the mean of the vahies of q from the two equations, 

 the last members will nearly destroy each other. It has been supposed that the 

 variations of X are altogether independent of the variations of the temperature, a 

 supposition which is borne out by every method of examination of the results. The 

 details of a series of comiiarisons are given, pages li., lii., and liii., Introduction, 1843, 

 from these it appears : 



70. 1««, That the value of q is the same, when a sufficient number of compari- 

 sons have been obtained, whether it has been obtained from comparisons of daily 

 means, at 1, or 2, or 3, .... or 14 days' interval. 



71. 2(7, That the value of q is the same, whether the differences of temperature 

 have been due to natural or artificial causes, and when the differences of temperature 

 of the mao-net have had an opposite sign from those for the temperature of the ex- 

 ternal air. 



72. From the second result, it follows, that the variations of the horizontal 

 component of the earth's magnetism are wholly independent of the temperature of 

 the air, and from both results it appears probable that they are independent of the 

 temperature of the soil.* 



73. The following Table contains the sums of differences of the daily mean 

 temperature of the bifilar magnet, and the value of f[ which has resulted from each 

 series of comparisons. The scries of comparisons for 1845 have been made since 

 the publication of the series for 1844, for the purpose of verifying the constancy of 

 the result. 



Table 8. — Determinations of the Temperature Coefficient of the Bifilar Magnet. 



The series of observations for 1844, giving each result an equal weight, give y' = 1'95 sc. div. 



1845, <i = 1'91 - 



1844, giving the results the weights 2 A <, give ^ = 1'92 •■• 



1845, g' = 1-95 - 



* See foot-note, p. 39.5 of the present volume. 



