Deter mi nation of the Earth's Horizontnl Magnetic Force. 385 



§ 16. I next arranged the eighty-two magnets of group C in sub- 

 groups, according to the value of /x, with the following results : — 



Table V. — Analysis of Magnets of Group C according to Size of 

 Induction Coefficient. 



Value of n. 



Number in sub- 

 group. 



Mean 



Mean 



TO. 



Mean to. 



Mean 



7T2K. 



Mean 

 Wq. 



Meon 

 10 8 ?'. 



Mean 10 5 x P/r 2 . 



Mean /u. 



When +. 



When -. 



>6 



26 



6-32 



883 



140 



2762 



433 



160 



800 



359 



6 to 5 -5 



24 



5*71 



898 



157 



2704 



330 



133 



994 



245 



5'5 „ 5 



19 



5-25 



854 



163 



2796 



294 



132 



754 



241 



5 „ 4 



13 



4-69 



837 



179 



2774 



296 



148 



787 



931 



There is here no trace of a connection between /x and K, which sup- 

 ports the conclusion drawn from the absence of apparent connection 

 between m and K. There is probably a slight tendency when fx is very 

 decidedly below the mean for m to be slightly low ; but the tendency 

 in m/fjL to increase as /x diminishes is conspicuous. 



The large value for P, when negative, when /x lies between 5 and 4 

 has no real significance, the mean being based on only two magnets. 

 The one clear and important relationship brought out by Table V is 

 between fx and q. Large values of these two constants unquestionably 

 have a tendency to occur together in the magnets of group C. Out of 

 the twenty-six magnets whose /x exceeded 6, no less than eighteen had 

 a q above the mean ; while of the thirteen magnets whose /x fell short 

 of 5, only two had a q above the mean. As the result seems important, 

 I submit the following analysis, showing the distribution of the different 

 values of /x in group C : — 



Table VI. 



All values 



From 7 to 6. . 

 „ 6 „ 55 

 „ 5-5 „ 5.. 

 „ 5 „ 4.. 



10 6 q 



200 



250 



300 



350 



400 



500 



600 



700 



11 



17 



19 



13 



7 



8 



1 



2 



2 



3 



3 



6 



4 



5 



1 



2 



3 



3 



7 



5 



2 



2 











3 



6 



5 



2 







1 











3 



5 



4 







1 















The table is to be read thus : Of the eighty-two magnets, four had a 

 value of 10 6 q between 150 and 200, and of these, two had a ll lying 

 between 6 and 5*5, while two had a li lying between 5*5 and 5. 



Table VI certainly supports the conclusion drawn from Table V. 



As a further check on this conclusion, I arranged the eighty-two 

 magnets in two sub-groups, according as q was above or below the 

 mean, with the following result : — 



