116 Messrs. Wright and Thompson on the Determination of 
VI. Magnesium. 
185. A series of cells was set up containing bright mag- 
nesium (wire) opposed to amalgamated zinc and electro- 
cadmium in their respective chloride-solutions, the strength 
being uniformly °25MCl, 100H,0O. The values exhibited 
much greater fluctuations as time elapsed than were observed 
with most of the other cells examined ; whilst the mean read- 
ings during the first half-hour after setting up also showed 
much less concordance, as the following figures show. As with 
the sulphate-cells (¢ 177), magnesium acquired the lower 
potential when opposed to zinc, and « fortiori when opposed 
to cadmium. 
Magnesium-Zince. | Magnesium-Oadmium. 
Maximum ..... ... T71 1101 
Minimum ......... 636 966 
JEN ISV ENO Sosphodsoe “702 1-030 
Probable error ... +:012 + ‘012 
Hence the following valuations of the voltaic constant result, 
these values necessarily being negative, as with the sulphate- 
cells :— 
Zine-magnesium cell .0t 70.8). ae 
Cadmium-magnesium — Zinc-cadmium { ee —-700 
Mean . . . —-701 
Julius Thomsen finds Mg, Cl, aq.=186930 for solution of 
strength °25 MgCl, 100 H, O; whence Hy=—1:634. Conse- 
quently the thermovoltaic constant for bright magnesium in 
contact with chloride solution of this strength is 
—*701— (—1°634) = + :983, 
or, approximately, the same large positive quantity as that 
found with sulphate solutions. 
VIL. Aluminium. 
186. In precisely the same way were two series of observa- 
tions made with cells containing plates of bright aluminium 
opposed to amalgamated zine and electro-cadmium immersed 
in solutions of their chlorides respectively, the solution- 
strength being °25 Alz Cl, 100 H, O*, and equivalent amounts 
* The aluminium-chloride solution was prepared by saturating dilute 
hydrochloric acid with recently precipitated well-washed aluminium 
hydroxide. 
