42 Dr. C. Chree on the Bending of 
Supposing a to be the value in minutes of arc of a scale- 
division, and (—8w) to represent the number of scale-divisions, 
the working equation answering to (2) is 
dr=17r(—du).2x'000291 cotu. . . +. (3) 
The angle u varies of course with changes in natural decli- 
nation or horizontal force, with change of temperature in the 
deflecting magnet, and with any artificial magnetic disturb- 
ances peculiar to the locality. Hxcept at times of pronounced 
magnetic storm, natural magnetic changes or changes of tem- 
perature present no real difficulty, as loadings and unloadings 
occupy so little time that a few repetitions secure a mean 
from which slow regular disturbing effects are satisfactorily 
eliminated. Large irregular magnetic disturbances, such as 
may be experienced near an electric railway, would unques- 
tionably be troublesome; but sites where such disturbances 
exist are in any case unsuitable for magnetometers. | 
The degree of accuracy obtainable by the method may be 
judged by reference to the results obtained by Mr. T. W. Baker 
on the first occasion when it was tried. Independent mea- 
surements were made in this instance by Mr. F. E. Smith, 
who employed the microscope apparatus previously in use. 
Magnetic Method at Kew. . 

Weights used each 658 grms. Direct Measurement 
Using at Bushy House. 
one two Weights used eacl: 
magnet deflecting magnets. 493 grms. > 
Distance where bending | eee 
effect observed ... cms. { ee av i 
—06u (minutes of are) ...... 292 ato o2i01 > 2D 
Caleulated 07, ......... em. ‘0088 -0098 ‘0098 -0156 
Calculated EIX10—° ...... 063 351 368 265 309 370 
Se ~~ ee +, ——_—_ 
VLEANS x necks o7ah EA - 361 364 
Here E is Young’s modulus for the bar, and I the moment 
of inertia of the cross-section about a perpendicular to the 
plane of bending through the c.¢. What either method 
really gives directly is HI/h, where h is the height of the 
magnet’s centre above that of the cross-section. Inthe magnetic 
method / represented of course the exact height at which the 
magnet is found in actualuse. In Mr. Smith’s measurements 
h had a slightly different value, so the value of EI formed the 
most convenient basis for comparison. 
Excluding the observations at 40 cms. in the magnetic 
method—to which we shall return presently—we have an 
exceedingly close agreement between the results from the two 
