MAGNETIC OBSERVATIONS 313 



The annual rate above deduced is further confirmed by the fact that it is of practically the same 

 order as values obtained at other Arctic stations. Thus, for example, for Cape Thordsen and 

 Mossel Bay, Spitzbergen, Dr. Solander* has deduced the annual rates of increase in magnetic 

 inclination as o.'4 and o.'8 for the periods 1883.2 to 1899.7 and 1873.5 to 1899.7 respectively. 



HORIZONTAI, INTENSITY 

 Instrument, Methods, and Results 



Observations for horizontal intensity were made, as already stated, with magnetometer 

 No. IIII loaned by the United States Coast and Geodetic Survey. For the various standardi- 

 zation observations with this instrument before and after the work of the expedition, as also 

 for the constants and methods of reduction used, see pages 6 to 11. Determinations were made 

 generally on Monday afternoons and Tuesday mornings, but it was not possible before the 

 advent of daylight to adhere strictly to this program. In these observations for intensity the 

 time of a set of fifty oscillations was observed in the usual manner before and after deflection 

 observations. Sometimes these were repeated ; at other times the final set of oscillations was 

 lost, owing to the failure to transit of the selected scale division. 



The fortieth division, being near the middle of the scale, was the one whose transit was 

 chosen. The initial amplitude of the oscillation was usually made about one degree in order 

 to increase the probability of the transit of the fiftieth oscillation. The amplitude of swing 

 was noted at the beginning and end, in scale divisions, but is not published for lack of space. 

 The time was noted by mean time chronometer Negus No. 1809, which was carried to the 

 observatory for the purpose. It was always placed at the foot and touching the north side of 

 pier j^ (figure 2). This chronometer was compared daily with the sidereal chronometer Negus 

 No. 1769, which was used in the astronomic observations. Determinations of the torsional effect 

 of the four suspension fibers were made before the first and after the last oscillation sets. The 

 temperature was noted at the beginning, middle, and end by a Centigrade thermometer, the 

 bulb of which projected into the magnet-house of the magnetometer. This same thermometer 

 was used in the deflection observations, readings being taken before and after each set. 



Magnet No. 2 was suspended during deflection observations. A solid brass bar with a 

 carrier supported magnet No. 4 at thirty and forty centimeters from the center. Eight 

 settings were made and the time and temperature noted as usual. Magnet No. 2 was kept in 

 magnet box No. 2 outside of the hut (shown on plat) and was brought into the observatory 

 immediately after preceding oscillation observations. From fifteen to thirty minutes elapsed 

 before it was observed upon. After deflection, and before the subsequent oscillation observa- 

 tions, it was returned to magnet box No. 2. 



The observation specimens, pages 314 and 315, which together yield an absolute value of 

 the horizontal intensity, H, will serve to make clear the method of observation and computation. 

 In the tabulation of results given below, each value of H is deduced from two or more such sets 

 of oscillation and one set of deflection observations made in the order stated above. 



By the aid of the specimens the accompanying tabulation of the condensed original 

 notes may be readily interpreted. The times given are local mean reckoned from midnight 

 through twenty-four hours. The mean deflection angles observed at thirty and forty centi- 

 meters are given in the two columns under heading u. The columns t and t ' show the mean 

 temperature-xeadings for the deflection and oscillation observations respectively. The mean 

 time of one oscillation corrected for the rate of chronometer appears under column headed T' . 

 The effect of ninety degrees of torsion in the suspension is given in the column v. The column 

 headed H giv&s the finally reduced values for horizontal intensity expressed in gammas, one 

 gamma being o.ooooi C. G. S. unit. The resulting values of the magnetic moment of magnet 

 No. 4 at ^ degrees and at twenty degrees Centigrade are shown in the last two columns respect- 

 ively. Reference to pages 8 and 10 will serve to explain the remaining headings. 



21 *See reference, page 305. 



