Magnetometer Deflexion-Bars. Al 
diminishes slightly the deflecting force on the deflected 
magnet, anda “slight change of reading is observed in the 
magnetometer-telescope. Putting the weights off and on, 
one gets in a few minutes sufficient measurements to supply 
a good mean. As a check, it is well to observe at two 
distances. Also if one takes a distance as large as 30 ems.— 
usually the shorter of the two distances employed in deter- 
mining the horizontal foree—it is advantageous to employ 
two deflecting magnets simultaneously, situated at equal 
distances on the iwi@ arms, on carriages of equal height. 
When only one magnet or carriage is av Vvailable 2, a distance of 
25 ems. is a convenient one in this country, with a deflecting 
magnet of normal moment (say 800 to 1000 c.@.s.). The 
smaller the horizontal force at the place, the larger is the 
distance at which defiexions may conveniently be taken, and 
the more favourable, ceteris paribus, are the conditions for the 
experiment. One can of course increase the bending effect 
most easily by increasing the weights applied. But if one 
uses weights much larger than those used at Kew (660 grms. 
each approximately) there is a risk of tilting the magneto- 
meter over—which would probably entail disastrous results. 
It is also desirable that the experimental weights should 
together not much exceed the weight of the bar (usually 
under a kilogram). 
To a first approximation the deflexion-angle u due to a 
magnet of moment m, at distance 7, when the horizontal force 
is H, is given by 

BN 0) te Soe ee 
The deflecting and deflected magnets are supposed to be 
mutually per pendicular, as in the ordinary horizontal-force 
experiment. Answering toan increment 57 in 7, we have an 
increment dw in u; and by (1)—assuming m "and H un- 
changed—the two increments are connected by the relation 
3(6r/7) + cot u du=0, 
or ea FCO UM es he Ae) 
In the present case 7 is a distance read off from the deflexion- 
bar, uw is the inclination of the deflected magnet to the 
magnetic meridian, as given by the azimuth- circle of the 
magnetometer, and —su is the change in w answering to 
loading by given weights*. In practice —6du represents a 
number of scale-divisions, as read off on the magnetometer- 
telescope. 
~ 
* Necessarily 6r is positive and du negative, 7. e. the deflexion-angle is 
always diminished by loading. 

