MAGNETIC SURVJEY OP GREAT BRITAIN. 55 



and the corresponding points of the other three quadrants be- 

 come in turns the points of support in the customary processes 

 of the reversals of the poles and circle. If this operation is 

 gone through at any part of the earth on or near the line of no 

 dip, the whole of the quadrant is thereby subjected to exami- 

 nation. In such situations, consequently, this method affords 

 the means of examining the whole circumference of the axle ; 

 and in all other localities, as much of the circumference as 

 amounts to four times the complement of the dip. Whatever 

 portion in the latter cases remains unprovided for, may be tested 

 by converting the needle, temporarily, into one on Mayer's 

 principle. This can easily be done by the application of a little 

 wax ; the quantity of which may be varied at pleasure, so as to 

 correspond with the weights of different sizes, by which, in 

 Mayer's method, the angles of inclination, from which the dip 

 is computed, are varied in successive observations. By one or 

 other of these processes the true dip at any station can be 

 obtained from any and every inclination of the needle; and every 

 part of the circumference of the axle can consequently be tested. 



In what has been said, it has been presumed that there 

 is no magnetism in the circle itself, as, should such exist, it 

 would certainly become the source of discordance in the results 

 derived from different azimuths, or from different weights, in- 

 dependently of any defect in the axle ; and so far, therefore, 

 the agreement of the results in such trials (should they be found 

 to agree) indicates with great probability the freedom of the 

 circle from magnetism as well as the goodness of the axle. But 

 Mr. Lloyd has employed and has described in a subsequent 

 part of this report an independent and much more delicate mode 

 of examination for magnetism in the circle. 



The customary provision of ttvo needles for each apparatus 

 does not alone afford security against the errors which may be 

 occasioned by either of the defects to which I have now al- 

 luded. In respect to the axle, if the results of the two needles 

 are accordant, it is thus far satisfactory, that it certainly is not 

 probable that both needles should have accidentally exactly the 

 same imperfection ; but if they differ, the observer has no guide 

 as to which is to be preferred ; whilst their mean result cannot 

 usually be more than an approximation to the true dip, for it is 

 also improbable that the two needles should have an exactly 

 equal amount of error in opposite directions. As a means of 

 detecting magnetism in the limb, two needles are of no more 

 avail than one ; because both are directed to the same point of 

 the circle when observed with at the same station, and, if a dis- 

 turbing influence exists, both will be subjected to the same error. 

 If, howevei', one of the needles is temporarily fitted on Mayer's 

 plan, — and the dip is obtained in successive experiments from 



