452 EXPEDITION TO POINT HAKKOW, ALASKA. 



suspended magnet, then the difference of the two readings of the scale, that is, before and after 

 the small weight was added, or for weight IV and for weight W+w will correspond to ^i,, of tho 

 horizontal force. To give the instrument any desired sensitiveness compute the angle of dellection 

 z corresponding to it, and set the torsion circle accordingly, then by means of the upper suspension 

 screw, with its two sets of opposing screw-threads, the suspension threads arc to be brought to 

 that distance, which will bring the middle of the scale (50) on the vertical thread of the telescope. 

 I "sing the second method a weight has to be provided corresponding to the desired sensitiveness, 

 and the suspension threads must be regulated in order that the additional weight may produce a 

 change of a certain number of divisions of scale when it is added and taken off. 



The instrument is provided with a mechanical compensation for changes of temperature. In 

 view of the extreme low temperatures which are likely to bo experienced at Point Harrow, how- 

 ever, and under the presetit circumstances, it will be better to deduce the corrections for any out- 

 standing amount, not compensated, differentially from the observations of the horizontal force 

 themselves, than to attempt a complete mechanical compensation. The latter operates as follows: 



Kefcrring to accompanying Jigure, suppose- 

 the temperature increases, tho effective force 

 of the magnet will diminish, the differential 

 expansion of glass and zinc (which materials 

 form the compensation) will push the :<; end 

 in, which brings the suspension threads closer 



/ . X 1 iT^U together, and thus diminishes the torsion 



' || '_ ; M **e force balancing /fin the same ratio tliat If 



may***, ( itself diminishes. Increasing scale readings 





~. should correspond to increasing horizontal 



nurrtrf 



magnetic force, or correspond to a movement 



I 



of the north end of the magnet toward tho 

 north. The narrow space dividing the fixed 



from the movable mirror is in the plane of the optical axis of the telescope. The instrument is 



placed under a zinc cover. 



"3. THE VERTICAL FORCE OR BALANCE MAGNETOMETER. 



"Put the knife-edge supporting the magnet in the magnetic meridian and level support: tho 

 magnet will then be free to oscillate in the magnetic prime i 



vertical; balance the magnet and its appendages (mirror, I 



knife-edge, balancing weights, compensation bar, &c.) hori- *<*? j n?0u 



/on tally by means of two weights on opposite sides of the 

 knife edge; next bring the center of gravity of the system 

 to that particular positioii close to and beloic the knife-edge "~ "~" 

 which corresponds to the desired sensitiveness; this is done 



nwv aZIe 



by raising or lowering the central ball or weight. Set the 



mirror so that the middle of the scale (50) is reflected on Ji**& I 



the thread of the telescope when the magnet is level; at the same time this center division must 



remain bisected, as seen in the. fixed mirror. 



Let 1'= the vertical component of the earth's force, d the horizontal distance of center of 

 gravity of the system from the plane of support passing through the knife-edge, 1J"= the weight of 

 magnet and appendages, m = the magnetic moment of the magnet, then I"HJ= 1JV. Now, suppose 

 the magnet inclined through the small angle i/, and let ft = distance of center of gravity of the 

 system below plane of knife-edge; then 



V_h 



"To determine the ratio * we oscillate the magnet and appendages in its vertical plane and 



let T= time of an oscillation in that position. We then take the magnet off its support and sus- 

 pend it (with its appendages) by a single thread (determining torsion and allowing for it), as in 



