NO. 7.] 



HORIZONTAL INTENSITY. 



73 



in 1895, October 17th, 



December 12th, 

 1896, January 28th, 

 March 19th, 

 June 18th. 



If to these are added the observation-days after the return of the expe- 

 dition -- March 2nd, 1897, in Hamburg, and April 17th and 19th, 1897, in 

 Wilhelmshaven we have 10, what I will call normal, days on which 

 determinations of the horizontal intensity have been made with the apparatus 

 mounted in the same way, under comparatively quiet magnetic conditions, in 

 temperatures varying from 28 G. on the 28th January, 1896, to + 11.6 C. 

 on the 17th April, 1897. 



THE FINAL VALUES OF THE TEMPERATURE-COEFFICIENT. 



In order to study the temperature-coefficient more carefully, I calculated 

 fi from the observations on the 10 normal days, by formula (6), both with 

 the value a = 0-000307 (Hbg.) found in Hamburg in 1893, and with the 

 mean value, a = 0'000514 (Ex), found in the above-described manner by 

 the observations during the expedition. The result was as follows: 



If the 4 series of values of /< in the table are plotted graphically as a 



I | i' 

 function of ~ , 4 fairly uniform curves are produced, which distinctly 



10 



