Instructions Jar Observers of the Aurora Borealis. 37 



the direction of the magnetic needle; but it is desirable to esta- 

 blish this as nearly as possible^ and for this we must know the 

 summit of the arch more accurately than the unaided eye can 

 give it. Intersect a part of the lower edge at a considerable dis- 

 tance from the vertex ; read the horizon circle ; turn the instru- 

 ment round its vertical axis till it intersects the lower edge again ; 

 the half sum of this and the former reading, is the reading of the 

 vertex. In these observations the time must be carefully noted 

 at each intersection, and the error of the watch used, ascertained 

 as soon as possible. Of course, those who are familiar with in- 

 struments will find no difficulty in this; but persons who depend 

 for their time on dials, or even mail-coaches, are not to neglect 

 the verification of this element. In the latter case, they will as- 

 certain from what town the time is derived, and what error it is 

 liable to ; but it must always be remembered, that without a 

 knowledge of the time, even though a rough one, all else is use- 

 less. 



Very good observations may be made without divided instru- 

 ments, by noting the position of the arch with respect to three or 

 more stars at a noted time. By drawing a line on a celestial 

 globe through them, the vertex of the arch may easily be found 

 with sufficient accuracy, by those who are not familiar with sphe- 

 rical trigonometry, and, of course, its altitude and azimuth. 

 Should the observation be made on a journey, the place of the 

 road should be mentioned. 



Any elevated object, as a spire, chimney, angle of a building, 

 upright pole, Sec, may be used for the same purpose. Let the 

 observer change place with respect to it, till he sees the point of 

 it that he has selected coinciding with the vertex of the arch ; 

 let him mark the position of his feet, so that he may recover it 

 in the morning, and determine the horizontal distance of that 

 point to the place where he stood, and its elevation above his 

 eye ; the latter divided by the former is the tangent of the al- 

 titude. If a theodolite can be procured, the altitude as well as 

 the azimuth of this point may be determined at once. 



Should the streamers have the appearance of converging to a 

 point south of the zenith, the altitude and azimuth of that point 

 should be determined. If possible the deviation and inclination 

 of the magnetic needle should also be obtained. 



These, owing to the imperfection of our magnetic instruments. 



