54 



BIRKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, IQO2 1903. 



The table shows that the scale-value for the variometer for horizontal intensity has remained constant 

 all the time, and has not altered perceptibly with the temperature. 

 We therefore employ the mean value, viz. 



ei, = 4-613 y per mm. deviation. 



DETERMINATIONS OF SENSIBILITY FOR MATOTCHKIN SCHAR. 



21. The registering apparatuses here were in operation from the 3Oth August, 1902, to the nth March, 

 1903. The first month, from the 3oth August to the 3oth September, was spent in trial registering: for 

 it proved to be very difficult to get the balance compensated for variations in temperature. Compensation 

 of the balance was effected on September gth,. loth, nth, I2th and ayth, and October 6th. The sensi- 

 tiveness of the balance was altered on the 23rd September and the gth October, being increased 

 both times. 



The H. I. variometer acted almost without change; it only now and then underwent small corrections 

 with regard to the position of the base-line. These cannot, however, be supposed to have had any 

 special influence upon the sensibility. The declinometer acted without alteration all the time. Astatising 

 magnets were not employed. 



It will be seen from the diagram below, that the thread in the declinometer was very stiff, the 

 effect of this being that v. is very large, and therefore has to be determined very exactly. At the same 

 time the H. I. variometer has a sensibility, which, especially considering the violent storms that occur 

 here, must be characterised as disproportionately great. It seems as though the threads for the two 

 variometers have been interchanged. 



TABLE VI. 



Scale- values for Matotchkin Schar. 

 H = o.i 1 13 



It will be seen that ea and /, keep fairly constant, and exhibit no decided variation with time and 

 temperature. We make use of the mean, putting 



,'i =1.62 



The value of e v are found from a curve, which together with the observed values, is shown in the 

 following figure. 



