U4 
ME. B. D. STEELE OX THE MEASUEEMEXT OF TOXIC YELOCTTIES 
noticed in every experiment wliicli lias been made witli copjier sulpliate as inter¬ 
mediate solution, and it is alAvays mncli nifire noticeable in tbe more concentrated 
S( tint ions. 
A few exjieriments bave been carried ont wbicb slnnv tliat complex ions of a 
different kind to those bitlierto considered do undoubtedly exist. The most striking 
of tliese is tlie measurement of the transport number of a solution of potassium 
ferric oxalate ; for this salt it was possible to obtain a very good margin with the 
acetic ion as anion indicator. At tlie other end of the tube, where the comjffex entered 
the cation indicator, thei'e is formed a solution of lithium ferric oxalate, and the ion 
penetrated the lithium chloride jelly for about I’O centim. on its way to the anode as 
a. conpilex anion. 
In another experiment, evidence of the constitution of the periodides was obtained. 
1’be apparatus shown in fig. 1 was filled with a O'oN solution of potassium iodide 
containing iodine, the indicators beino' sodium acetate and lithium chloride. 
iVc 
At the cathode end, whei'e the -- Na houndary was formed, a precipitation of 
iodine occurred, doubtless on account of the fact that the I, ion being slower than the 
I ion, some of the former would lie left behind in a portion of the system, where 
the ecpiilibrium between KI, I.,, and KIg was disturbed, and tlierefore nearly the 
whole of the iodine was thrown out of tlie solution. The existence of complex 
Ig ions is conclusively jiroved by the behaviour of the solution at the anode end; 
here a ^ I margin ti'avels away fi-om the gelatine solution of lithium chloride, and 
simidtaneously the 1 ions enter the latter, forming lithium iodide, but not only I ions 
enter the jelly, but Ig also, and the jelly for a length of about 1 centim. becomes 
coloured a deep red by the solution of Li Ig that is produced. The entrance of the 
iodine is far too rapid to be explained by diffusion, and a lilank experiment that was 
carried out showed that in the same time the distance covered b}" diffusion of the 
iodine amounted to only about 1 millim., and showed no such sharp line of 
demarcation between the coloured and colourless portions of tlie system. 
Previous evidence of the existence of tlie compound KIg is given by Le Blaxc 
and Noyes (‘ Zeitsclirift fur Phys. Chem.,’ 13, 359, 1894, and 20, 19, 189G), Jakowkix 
{ibid., 6, 385, 1890), and Noyes and Siedenstickee {ibid., 27, 357, 1898), whilst 
Day’sox (‘Chem. Soc. Jl.,’ 79, 238, 1900) discusses the dissociation of the compound 
KIg into the ions K and Ig. 
Yet another series of experiments may be described which tend to show in the 
same manner the presence of complexes in solutions of copper sulphate. When a 
solution of this salt is followed hy the three indicator ions Cd, Mg, or Li, a great 
difference in behaviour is to be noticed. Cd as indicator fulfils the condition that 
it is slower than the Cu, and consequently the solution behind the boundary remains 
quite colourless ; with the other two ions this condition is not fulfilled. When Mg 
