38 



KESEARCHES ON 



n. 



former, or when its wire is too short to afford any accurate determination of it, the 

 following method may be used, which we have found to supply better results than 

 any other, and may be added to the others already known. 



Let two experiments like those described above be made by introducing a different 

 number of arbitrary turns p and pi into the circuit, we shall have the following 



equations (supposing — = /) : — 



from which we obtain 





Pi 



and consequently knowing r, 



9 = 



f 



"We subjoin experiments made in different days, and under circumstances very 

 different from each other, because those made one after the other agreed together 

 to tenths of a turn. This is not surprising, because from the form of the value 

 of /, the errors equally affect the numerator and the denominator, and in practice 

 it is easy to make observations, so that if there is a little error, it should be always 

 on the same side. Besides, with this method there is no need of calculating the 

 resistance of B, which is rather long. 



Experiments. 



No. of Obs. . 



P/ 



P^ 



r* 



r\ 



/ 



1 



1.0 



98.2 



3.0 



33.8 



3.156 



2 



9.8 



35.2 



6.6 



14.3 



3.299 



3 



6.0 



28.3 



5.7 



13.0 



3.081 



The difference between the maximum and minimum of these values is = 0.217, 

 and their common mean = 3.179, which in practice we assume as / = 3.18. 



II. Determination of the Forces. — After having settled the needle perfectly in the 

 centre of the globe or circle, the current is transmitted through their wire, and the 

 deviation of the needle carefully observed. If the needle does not point exactly 

 to any division, it is easily reduced to it by the rheostat. As this angle is of great 

 importance in calculations, in order to determine it more accurately and avoid the 

 error of eccentricity of the needle, which is always sensible, by reason of the small- 

 ness of the circle, and also the errors caused by a slight aberration of the currents 

 from the plane of the meridian, I proceeded as follows : — 



1. The points were carefully marked to which the opposite ends of the needle 

 corresponded when no current was running through the wires, and both positions 

 noted. 



2. Transmitting the current through the wire, the deviation of the opposite 

 ends was noted, and their mean taken, which is the true angle of deviation as 

 measured from the very centre of the circle. 



