613 



of the suspended magnet is seen to coincide with the central 

 division of the scale of the telescope. The verniers of the 

 circle being then read, the deflecting magnet is reversed, and 

 the telescope is moved until there is a new coincidence. The 

 verniers being again read, the diff'erence of the two readings 

 is double the angle of deflection sought. It is necessary to 

 eliminate the changes of the Magnetic Declination, which may 

 occur between these two readings ; and for this purpose the 

 wooden box and one of the spare magnets may be employed 

 by a second observer. But the same elimination may be 

 made as effectually by a single observer, by taking a series of 

 readings with the deflecting magnet alternately in the two 

 positions. Finally, the observation is to be repeated with 

 the deflecting magnet at the same distance on the other side 

 of the suspended magnet, and the mean of the two results 

 taken as the deflection corresponding to that distance. 



The quantity sought may be inferred from the angle of 

 deflection at a single distance, with as much accuracy as is 

 generally attainable in observations made in the open air, or 

 in a tent ; and, in such cases, it will generally be found more 

 advantageous to multiply the observations at the same dis- 

 tance, in the manner already mentioned, than to repeat them 

 mttwoov more distances. The distance should be about 

 five times the length of the magnets. 



The preceding arrangement is suggested chiefly in regard 

 to the economy of time. But, when the observer has suffi- 

 cient leisure, it is desirable that the time of observation of 

 the two elements should be as near as possible to the epochs 

 of their principal maxima or minima, the periodical variation 

 being then least. For this purpose the observations should 

 be so arranged, that the middle of the observation of Inten- 

 sity may fall between 10 and 10^ a. m. ; and that of the ob- 

 servation of Dechnation between 1 and \\ p. m. In this case, 

 then, the preceding arrangement should be nearly reversed. 

 The observer should commence with the observation of de- 



