250 DESCRIPTION OF A MAGNETIMETER, BEING A NEW 



the bar by the needle, will be mistaken for the magnetism of posi- 

 tion, and give an erroneous result. Thus, in the preceding ex- 

 amples of dip, the needle in use being too large and powerful, the 

 results, especially the latter, are probably a little too low. The 

 nature of this error will more particularly appear from the next 

 proposition. 



Note. — When a sufficiently small needle is not at hand, the amount 

 of error in the observed dip, occasioned by the magnetic influence 

 of the needle on the bar, may, in all cases, be determined by ex- 

 periment. For we know that the plane of the magnetic equator 

 coincides with a true horizontal plane, in an east and west magne- 

 tic line, consequently, when the instrument is placed east and 

 west, the plane of no-attraction should be horizontal. But, if the 

 instrument be placed in this position, with the compass-needle at 

 right angles to the end of the bar, as in Figure 2. it is found that 

 the infusion of magnetism from the needle makes the plane of no- 

 attraction something above the horizontal. This angle, measured 

 from the plane of the magnetic equator, (instrument E. and W.) 

 is the correction to be applied to the co-dip, observed with the 

 compass-needle at the same distance from the end of the bar, with 

 the instrument N. and S., as in Figure 3. Thus, with the instru- 

 ment N. and S., and the north-pole of the compass-needle eight- 

 tenths of an inch from the end of the bar No. 1., the mean angle 

 of no-attraction was 26°; and with the instrument E. and W., 

 compass same distance, the angle of no-attraction was 7° 30 above 

 the horizontal. Now, were the magnetic dip 90°, this would be the 

 correction to be made use of; but as, in the present dip, the moveable 

 plate of the instrument, when E. and W., traverses obliquely to the 

 magnetic equator, while in the N. and S. direction it moves perpen - 

 dicularly to it, the observed correction must be reduced, so as to 

 give the angle formed with the magnetic equator. In this opera- 

 tion, radius is to the sine of the dip, as sine of the observed cor- 

 rection (or angle of no-attraction with the instrument E* and W.) 

 to sine of the true correction, to be applied to the co-dip, observed 

 with the instrument N. and S. Hence the above observed error 



of 



