IX AQUEOUS SOLUTIOX, AXU THE EXISTEXCE Of COMPLEX lOXS. 
145 
follows Cii the latter falls behind the margin, and the colorn- does not entirely 
disappear until a point about 2 centiins. behind the boundary is reached ; no copper, 
however, is found to enter the MgSO^ jelly. 
AYith Li as indicator the copper also lags Ijehind the boundary, but as has been 
previously mentioned, the colour shows no sign of entirely fading out, and when 
there exists a column 6 centims. long of Li^S(4^, solution, the blue colour of the 
copper left Ijehind is perfectly uniform, hut lighter in colour than the measured 
solution. On 'examining the lithium chloiide jelly after the experiment, it was 
found that copper had entered it for a distance of between 3-3‘5 millims., thus 
indicating the passage of the copper to the anode as complex anion. This behaviour 
is possibly associated with the fact that the douljle sulphates of copper belong to 
the class MhCu (SOjjn, and hence a complex Cu (SO Jo will be staljle in the presence 
of a monad cation such as Li, but not if the cation is of the type of Mg. 
The method described in the preceding pages gives a simple and accurate means 
of determining the transport number of the simplest type of salts ; in its })resent 
form it is applicable only to more concentrated solutions. It is hoped to iiKjdify 
it in such a manner as to permit of the measurement of more dilut;" solutions. 
For salts of the dyad metals, more doubt attaches to the accuracy of tlie method, 
in consequence of our want of exact knowledge as to the mechanism of the changes 
that may take place at the margin, on account of tlie presence of com})lexes. 
The existence of these may he considered to he established v itli a |)rohal)ility 
amounting almost to certainty, hut the evidence is, as yet, qualitative only ; by the 
solution f)f the problem, as to exactly how the margin velocity is inliuenced by the 
presence of complexes, and their dissociation outside the margin, it is possible that a 
means of determining quantitatively the proportion of complex t() sinqjle ions in a 
given solution may he indicated. 
In conclusion, it is the author’s very pleasant duty to express his indebtedness 
to Professor Abegg for the help he has received from him during the course of 
the work. 
VOL. CX(WI1(.—A. 
V 
