244 F REPORT—1874., 
the resistance was altered until the balance was got as accurately as possible. 
It was found that in this way changes of resistance amounting to 0-001 ohm, 
or to about + 4, of the quantity to be measured, could generally be detected 
without much difficulty. When measurements were made at atmospheric 
temperatures, the pyrometers were always immersed in water nearly up to 
the junction of the sheath surrounding the coil with the stem, and were placed 
in the water at least an hour before the measurement was made. Owing, 
however, to the conductivity of the mass of iron forming the stem, it was 
impossible to insure that the pyrometer-wire was even nearly at the same 
temperature as the water, if this differed much from the temperature of the 
air in the laboratory ; and consequently it was thought better not to attempt 
to get measurements at a fixed temperature, but to make the determinations 
as nearly as might be at the temperature of the laboratory, and to reduce the 
results to a fixed temperature by calculation. As 10°C. was approximately 
the mean temperature at which the measures were actually taken, this was 
adopted as the temperature of reference. 
The formula employed for correcting the observations for temperature was 
obtained as follows :—A piece of the same German-silver wire as that of which 
the 10-ohm coil was made in the set of coils used was immersed in water 
and its resistance measured, in one experiment at 17°-6 and 100° C., and in 
another, with a slightly different arrangement of the apparatus, at 18° and 
100°. These determinations gave for the value of a in the formula 
Ye Yio {1 +a (0 ta 10)], 
where @ is the temperature of the wire, y, the observed resistance at 6°, and 
Y,> the resistance at 10°, the values 0°000306 and 0:000312 respectively. The 
mean value 0:000309 was adopted in subsequent calculations. On the other 
hand, the following determinations were made with one of the pyrometers, 
No. 404 :— 
(1) Temperature of water surrounding pyrometer =18°, temperature of 
e ” 11x 400 ‘ 
- = oO. ] = -—_——____ = 7 
German-silver coils =19°%1. Resistance 114400 10-7056 in terms of 
German-silver standard at 19° 1. . 
(2) Temperature of water =91°-6, temperature of standard =19°-8, Re- 
13 x 1040 
sistance in terms of standard at 19°-8 =734 1040 712840. 
(3) Temperature of water =93°-3, temperature of standard =20°. Re- 
é z 13 x 1500 
eR > Tet . 
sistance in terms of standard at 20 =B +1500 = 1? 888. 
(4) Temperature of water =18°, temperature of standard =18°8. Re- 
3 ; 11 x 401 
°°, = aS 7 
sistance in terms of standard at 18°-8 =T14- 401719 70638. 
The conditions of experiments 1 and 4 and 2 and 3 respectively were so 
nearly identical, that each of these pairs was combined to give a single mean ; 
this gave— 
Resistance at 18° in terms of standard at 18°-95 =10-7060, whence re- 
sistance at 18° in terms of standard at 10°= 
10-706 (148-95 x 0:000309)=10:7356, . . . . (1) 
: Resistance at 92°45 in terms of standard at 18°-9 =12:864, whence re- 
sistance at 92°45 in terms of standard at 10°= 
12-864 (149-9 x 0-000809)=128994 . . . + (2) 
