of Action between Magnets. 133° 
necessary to the value of H when Q and R are neglected,— 
observations being taken only at two distances,—are given in 
Table XIV. 
The effect of an increase in p would be to raise the 
corrections. For instance, if we assumed p=°86, then for 
H=:'185, and observations at 30 and 40 cms., we should 
raise the value of 6H to 7:6y for the Jones, 4°8y for the 
Cooke, and 4:7 for the Dover magnetometer. 
§ 20. In any individual case the values of 2)/I and 2)//I’ 
may one or both differ from the mean, and if they do differ, 
they may differ in opposite directions. Table XV. gives the 
changes in P, Q, and R due to change of p from ‘80 to °81. 
A change from °80 to*79 has a nearly equal effect, but in the 
opposite direction. 
Taste XV. 
Hffect of change in p from °80 to ‘81. 
Magnet in which P | Q R P’(30 & 40 cms.). 
p increased. Collr. Mirror Collr. Mirror Colir. |Mirror| Collr. | Mirror 
Rimes esd, | ache. | oa ai 
Jones (or Camb. "704 | -697 | 18°9 | 24:3 | 1444 | 750 668 737 
Inst. Co.). . | 
Cooke (and practi 692 | 487 | 46 | 21:5 | 1050; 60 682 524 
cally Elliott). | | 
‘67 
Increments in p in the collimator and mirror magnets 
always influence P in opposite directions, and a given 
increment in p has a very much greater effect when occurring 
in one than when occurring in both of the magnets. Small 
uncertainties in the pole-distance have comparatively little 
influence on Q, but materially affect P and P’. 
§ 21. It has been usual when testing at Kew to observe at 
least twice with each magnetometer, so as to check the 
values of the constants and the general behaviour. Deflexions 
are made at 30 and 40 ems., and the value of P’ is thence 
determined. In the case of Dover magnetometers the follow- 
ing are the results :— 
mean value of P’ from all instruments examined + 6°20 
2 4. c fast 3", \ nA + 6°43. 
Applying the correction —'12 from Table X.to the values 
of P’ (A, X’) in Table XII., we have for the average Dover 
magnetometer : y 80) 8] 89 
P 
Calculated) value of P’=  +6:13 +6°26 +6:°40 
Thus the mean from all the Dover magnetometers would 
