1898-1902. No. 6.] TERRESTRIAL MAGNETISM. 71 



we obtain 



These values are manifestly too low. The total intensity at the 

 places in question must be assumed to have a value of about 0.6 c. g. s., 

 or more. 



That the error cannot originate from the values for the horizontal 

 intensity, is evident from the fact that if H is calculated with W = 0.6 

 and the corrected values of I, H is found to be from 800 y to 1100 y 

 greater than the mean value observed. 



It must therefore be assumed that the values of J calculated ac- 

 cording to the formula (5), are much too high, which again may be 

 accounted for by the fact that the constant p, which is dependent upon 

 the magnetic moment of the inclination needle, has changed its value. 

 No very great degree of accuracy, moreover, can be attributed to the 

 above-mentioned determination of p and a. 



Under these circumstances I have found it absolutely impossible 

 to set down any reasonable value for the index-error in the present in- 

 stance, and have therefore had to content myself with giving the ob- 

 served inclination values without correction. These, however, cannot 

 be so very different from the true values, as is indicated by the fact 

 that together with the observed values of the horizontal intensity, they 

 give the following not unreasonable values for the total intensity: 



