XII 



INTRODUCTION. 



these experiments, an hour of the day was chosen when there is least variation. The results 

 were as follows : 



Experiments to determine zero of magnet scale. 



After the tenth observation the direction of the alt-azimuth telescope was changed. 



This zero was possibly subject to slight changes from temperature, accidental blows to the 

 magnet, &c. ; but on both occasions, when carefully tested under nearly the same circumstances, 

 there was no appreciable difference. Throughout the journey across the Andes and pampas of 

 the Argentine republic, Lieut. MacKae adopted 41.0 as the zero, probably because the even di- 

 vision could be noted with greater certainty rather than from actual observed change in the 

 magnetic axis, though the difference from the zero adopted is not greater than would occur be- 

 tween very accurate and very rapid determinations as his, of necessity, were. 



Being placed upon its support in the line indicated by the small magnet of the alt-azimuth 

 instrument, and at such elevation that the collimator magnet and telescope of the latter may be 

 in the same horizontal plane, the declinometer requires but two adjustments, namely, to level 

 its base, and to bring the direction in which the force of torsion of the suspending fibres act, 

 accurately into the plane of the magnetic meridian. The first of these is immediately effected by 

 suspending the plummet and turning the foot-screws until the point of the former hangs over 

 the centre of the base, and the suspending thread in the centre of the tube. After the plummet 

 has ceased gyrating, suspend the stirrup and magnet in its place without torsion of the thread, 

 elevate the magnet until the scale divisions are seen distinctly in the telescope, and turn the 

 table top until the central division coincides with the vertical wire. Remove the magnet, intro- 

 ducing the torsion bar with the weak magnet in its place, and when the latter has come to rest, 

 should its vertical plane differ from that occupied by the collimator magnet, estimate the angle 

 formed, (if the scale is out of the field of the telescope,) and turn the torsion circle through an 

 equal amount in the opposite direction, and the middle division is again on the vertical wire. 

 If, on inversal of the bar, the same division remains bisected, the plane of detorsion coincides 

 with the magnetic meridian, and the instrument is ready for use. Otherwise, the zero of the 

 scale must be determined by the same process as that of the collimating magnet already men- 

 tioned ; and when ascertained, the torsion circle is to be turned, and the azimuth of the telescope 

 changed until such coincidence is perfected. 



The alt-azimuth instrument having been carefully levelled, and the line of collimation of its 

 telescope coinciding approximately with the magnetic meridian, the mode of observation is 

 shown in the following transcript from the note-books : 



WEDNESDAY, AUGUST 21, 1850. 



Locality, a vineyard 1,500 feet E. by N. from the northern extremity of Santa Lucia. Tem- 

 perature in the shade from 60. to 60. 3 ; barometer 28.268 inches; attached 58. 2 ; atmo- 



