768 REPORT— 1885. 
sodium and mercury, gold and mercury, cast-iron, with negative results 
so far... But none of these alloys except sodium and mercury are well 
chosen. It would seem advisable to choose two metals far apart in the 
voltaic series, or in the thermo-electric series, so that one might cling more 
to positive electricity, and the other more to negative, if possible. Or 
they might be chosen with their so-called ‘ specific heats of electricity’ 
very different. Or, again, it may be well to choose two metals differing 
considerably in conducting power, so that one might be rubbed along with 
thé current, so to speak, more than the other. 
I might suggest gold and lead, or copper and tin,” or copper and zinc, 
or bismuth and antimony ; beside, of course, mercury and zinc, or mercury 
and sodium, or, again, copper and zinc dissolved in mercury &. But 
it is to be remembered, from the experiments of Dr. Guthrie, Prof. 
Chandler Roberts, and others, that diffusion in all such cases is exces- 
sively rapid, and that precaution must be taken to diminish its effect in 
restoring uniformity of composition. 
A U tube with a constriction at its bend might be used, the whole 
filled with homogeneous alloy and kept just above the fusing-point. Pass 
a very strong current through the U, and gradually cool down till the 
whole is solid, keeping the current on all the time. Then analyse the 
two legs, and see if they show any difference in composition. 
4, Is there any relation between Optical Opacity and 
Electrolytic Conductivity ? 
Metallic conductors are opaque, but electrolytic conductors are, 
ordinarily speaking, transparent. At the same time, ordinary water is by 
no means so transparent as bisulphide of carbon; it absorbs a great deal 
of some kinds of radiation. Does this opacity increase as conductivity 
increases, say by addition of acid, and diminish as the water becomes 
purer; or is there no relation between conductivity and opacity in 
liquids ? 
Vision through long tubes is the simplest mode of attacking the ques- 
tion, but to get anything like a complete or satisfactory answer, the 
absorption spectrum of liquids in the less refrangible ultra-red region 
would have to be observed. 
There are two obvious ways of explaining the transparency of a con- 
ducting liquid on the electro-magnetic theory. One is to assume that the 
minute and rapidly reversed electromotive forces which constitute light 
are incompetent to effect anything like alternate decomposition and re- 
combination, so that the liquid to such forces behaves exactly like a 
dielectric ; a supposition which would be equivalent to admitting a slight 
disobedience to Ohm’s law, and which is a very reasonable one. The 
other is to assume that electrolytic conduction only occurs among dis- 
sociated atoms, and that these free atoms constitute so small a percentage 
of the total number as not appreciably to affect the transparency or 
dielectric character of the bulk of the liquid. On this second hypothesis, 
however, some slight absorption of radiation ought to be expected, and it 
might increase as the conductivity, 1.e. percentage of dissociated atoms, 
increased. 
1 In Mr. Shaw’s very useful article ‘ Electrolysis’ in the Hncy. Brit., the slip is 
made of saying that these alloys have been electrolysed. 
2 Prof. Roberts’ curious alloy Sn Cu, is well worth trying; see Phil. Mag., Dec. 
1879, p. 558. 
