NO. 7.] 



HORIZONTAL INTENSITY. 



75 



not give sufficiently serviceable material for investigation. For the cases in 

 which magnet VI has been used, therefore, I have simply had to take the 

 value found in Hamburg in 1893 for the temperature-coefficient: 



a = 0-000638 . 



THE FINAL VALUES OF THE CONSTANTS /* AND C. 



The assumption that the magnetic moment of magnet V, and thus the 

 factor (U also, have remained constant throughout the expedition, is made, as 

 already mentioned, the basis of the final calculation of the temperature- 

 coefficient of the magnet. As a check on the correctness of this assumption, 

 I have again calculated (i with the final value of a for each of the 8 certain 

 normal days with the following result: 



The values found agree, as the table shows, very well with one another, 

 and I have therefore assumed the mean of all the determinations as the 

 final value of p for the respective distances e and E. 



No observations permitting of the determination of /.t were made with 

 the small declination-needle as deflected magnet during the expedition. For 

 this mounting of the apparatus, I have therefore kept to the determinations 

 in Hamburg in 1893, and in Wilhelmshaven in 1897. I have recalculated 

 the latter, introducing the improved ultimate value of a and obtained: 



