An experiment with a pair of Robison ball-ended Magnets. 185 
An experiment with a pair of Robison ball-ended Magnets. By 
G. F. C. Searle, M.A., F.R.S., University Lecturer in Experi- 
mental Physics. 
\Read 28 January 1907.] 
§ 1. The advantages of the ball-ended magnets devised by 
John Robison in 1769 were discussed by me in a paper * read 
before the Cambridge Philosophical Society on Nov. 10, 1902, and 
further experience has proved the utility of these magnets in the 
instruction of students. I now describe an experiment in which 
two Robison magnets are employed. 
In an ordinary magnetometer the pivoted magnet is usually 
so short that its length may be neglected and thus the axis of the 
magnet may be regarded as a tangent to the line of horizontal 
magnetic force which passes through the pivot on which the 
magnet is supported. 
If we use a long Robison magnet for the magnetometer we 
can no longer neglect the length of the magnet when the lines of 
force in the neighbourhood have sensible curvature, but, at the 
same time, we gain the advantage of being able to take a fairly 
exact account of the length of the magnet, since the poles of a 
Robison magnet may be regarded, with considerable accuracy, as 
situated at the centres of the spheres. 
G 
Fig. 1. 
A Robison magnetometer is easily constructed in the following 
manner. A brass U-pieco, A (Fig. 1), formed of a strip of brass 8 mm. 
wide and 1 mm. thick, is soldered to a Robison magnet, and this U- 
piece is provided with a boss B through which a sewing needle 
passes, the needle being secured by a small set screw. The point of 
the needle rests in an agate cupf G carried by an arm of brass D 
projecting from a wooden block E as shown in Fig. 2. The block 
rests upon a drawing board and its height is such that the 
* “Notes on a Vibration Magnetometer and on the Ball-ended Magnets of 
Robison,” Proc. Camh. Phil. Soc. Vol. xii. Part 1, p. 28. 
f Agate centres can be procured from instrument makers at a small cost. 
A substitute is easily made by heating the sealed end of a piece of small glass tube 
and then sucking out the air from the other end so as to form a small cup at the 
heated end. 
VOL. XIV. l'T. II. 
13 
