MAGNETIC SURVEY OP GREAT BRITAIN. 



109 



To compare with this, we have the following observations 

 with the needles L. 1, L. 3, L. 4, in Dublin. 



Table XXXV. 



Hence we obtain the following corrections : 



Needle L. 1, correction = + 12'*0 

 „ L. 3 „ = + 5'-3 



„ L. 4 „ = + 13'-4 



In L. 3 and L. 4, needles whose poles are unchanged, the 

 errors here deduced are, of course, those which result from 

 the moment of the needles' weight, combined with that arising 

 from the disturbing action of the limb. 



The weights due to these corrections are at once deduced 

 from the principles of the preceding pages. When the results 

 of one needle, at a given station, are compared with those of 

 others, and that we seek their difference, it is manifest that 

 /» = 1,5'= 1, (14), and that, consequently, 



W = Ei^ + Eg^ ; 



Ej denoting the probable error of the mean result of the 

 given needle, and Eg that of those with which it is compared. 

 When we look no further than the actual difference of the re- 

 sults at the one station, it is manifest that 



E, 



fi_ E2 — ii. 



> ■'-'9 — > 



n^ ^ Tic 



c, and €3 denoting the probable errors of a single observation, 

 in the needles compared, and Wj and Wg the number of obser- 



