NIPHER 5'^'^ ANN. REP. MAGNETIC SURVEY OF MISSOURI. 525 



The value of P^ as determined in the first of the tables, gives 

 P 

 for the value of i ^ the values 



P , 



log 



These values were used in the reduction of the deflection series 

 made for the determination of the magnetic moment of magnet 

 Co. 



In the second table, the decrease (a) in the value of log m is 

 determined, and it is found that for the summer of 1882, and at 

 a temperature of 80.7, the value of (a) is 



Q-02 1 13 



^ '-'•000,009 ' 



2032 



or l0g/« = 9-86415 + 0-000,009 ^^ 



where d is estimated in days from July 13. The value of log m 

 was therefore considered constant during the summer. 



In all of the intensity determinations the time of vibration was 

 determined by means of a Waltham watch belonging to Mr. 

 Ringling. This watch had been cleaned just before leaving St. 

 Louis, but it had not been rated. The error of the watch was 

 determined, at intervals during the summer, by comparison with 

 clock-beats from the observatory of Washington University, 

 transmitted daily to the telegraph lines of various raihvays in the 

 State. Its rate during the summer was a loss of 20 seconds per 

 day, fluctuating however between 17 and 22 seconds. The cor- 

 rection on the time of vibration was -\- o.^^^^ second at all the 

 stations, the time of vibration at the stations not varying suffi- 

 ciently to change the value of this correction. The effect of 

 neglecting this correction altogether would be equivalent to the 

 effect of an error of half a degree in temperature. 



The intensity determinations were all made with magnet Cg in 

 the University declinometer, C j 7 being used as a deffected mag- 

 net. The moment of inertia of Cg was obtained from the table 

 given in the 4th report.* The values for H are not corrected for 



* Trans, vol. iv. No. 3, p. 468. 



