on a Permanent Hlectric Current. 181 
site to that in iron. The action in nickel, though not really 
measured, was seen to be very decided, and may possibly prove 
to be as strong as that in iron. 
TIN. 
The action in this metal is very small and has not been 
measured with any accuracy. Its magnitude may be 3 that 
of the action in gold. 
o other conductors have been tested in such a manner as to 
warrant an expectation of detecting an action. 
In the following table the results obtained with the different 
metals are brought together. Those obtained with very thin 
strips will be mirked thus (?) for reasons which must be evi- 
dent to any one who has read the preceding pages: 
MXV 
Metal Plate. M. Cc. ae 
Gold, No. 6 [‘ hard” ] 152 x 101° } 
nos @ a 6400 kxtan 23°44" 150x10" (1.1. og 
aot 6400 “ “ 42°14% 150x 10% i 
oo BS 6400° * “ 49°98” 1] g1 
“No. 5 [soft or semi-cohes.]6400 “ “ 42° 267 161 x10! | 
13558 6330 * “26° 2% = 163x10 | Oo. 99 
Sea 6440 “ “ 99°48" 162x101 
ey ae 6440 “ “ 43° 0% 164x100 
“ No. 4 [" soft] 6480 “ “ 99°91% 155x 10 
oe” 6420 “ 96°25’ = 155x110 | ae ao 
aka toe 6480 “ “ 42°16 154x10” 
vs a 6480 “ “ 28°43’ 154x101! 
‘* No. 30 (A) [semi-cohes.?] 6520 “ “ 48°38’ 123x10" 
e eT ve 6600 “ * 31°30% 124x 10 11250 x 10° 
‘ Bg alt 6600 “ * 40° 39% 128x101 
“* . .{B)fsemi-cohes.7] 6760: ** © 68° 0%. . 139 10” 9 
a we MOE 6760 “ 39° 96% 141x101 t 1400 aan 
ilver, No. 10 6580 “ “ 49°17% 114x10” ‘ 
o a 6580“ * 39°907 118x190 ¢ 1160x10 
. eposited] (A 7120 “ 43°33” 134x109! : 
“ a 7120 “ “ 19°39 137% 1910 ¢ 1855 x 10°? 
; : o $ (B) 6640 “ 47° 397 1350 x 10°? 
ron ( 6680 “ © 38°37” —127x 109 4 
eS 6680 “ “ 46°13’ —130 {1285 x10 
gaan .« «@ gg? ey 4170 x 10° 
Platinum 
Nickel—effect large, possibly as strong as in iron. 
Tin—effect probably much smaller than in platinum. 
list is placed a number representative of this magnitude. In 
the case of gold this number is a quantity inversely propor- 
tional to the mean of the results obtained with the five differ- 
ent plates named above. In finding the corresponding number 
for silver, I have, for obvious reasons, used only the result ob- 
tamed with the plate of No. 10. The representative number 
given for tin has been very roughly estimated and may be one 
