350 
PROFESSOR 0. MASSON ON IONIC VELOCITIES. 
certainly seen to exist. Moreover, one would naturally expect the author’s results to 
show better agreement with Kohlrausch’s 1879 values than with his later ones, as 
the former were deduced from the conductivities of normal and half-normal solutions. 
This is notably the case, and is conspicuous in the lithium values. The only striking 
difference is that seen when the relative values of cation and anion are compared, 
wliich would he still more pronounced were one to include SO 4 in the Kohlrausch 
columns ; l)ut he himself excludes it. This difference is, however, but a repetition of 
what has been already noticed in the comparison of the results obtained Avith 
Hittorf’s transport numbers ; for Kohlrausch bases his calculations on Hittorf’s 
aMuo of^J in KCl. 
The character of the observations on which the author’s values are based Avill be 
best judged by inspection of the specimen curA^es shoAvn in the accompanying plates. 
In fig. 3 are shoAvn the curves for half-normal chlorides of potassium, sodium, 
and lithium ; and in fig. 4 the curA^es for the three strengths of lithium sulphate. 
In all of these figures the right-hand curA^e represents the progress of the yelloAv 
boundary, i.e., of the anions, Avhile the left-hand cuiwe represents the simultaneous 
progress of the blue boundary, ?.c., of the cations. The gradual narroAving of the 
figure enclosed by the right-hand and left-hand curves represents the colourless jeUy 
originally filling the Avhole tube, and becoming curtailed till it A'anishes at the apex. 
The curvature, Avhich corresponds to diminishing A’elocity (and current), is seen to 
depend on the nature of tlie salt used—to be most marked in the case of the best 
conductor (KCl), and to nearly disappear in that of the Avorst (Li 2 S 04 ). Fig. 4 
shoAvs the regular result of increased concentration, Avhlch decreases tlie ionization (cc), 
and therefore the A^alues of the Avorking A^elocities (U and V). 
Tlie displaced point at the apex of the half-normal KCl cuiwe, already described, 
is shoAvn in fig. 3 ; and in fig. 4 it may l)e noticed that the later readings of the 
yelloAV boundary Avere slightly inaccurate, this being due to the fact that the 
experiment, Avhich AAns a long one, had to be finished by gas light. 
In conclusion, the author desires to express his great obligation to his colleague, 
Professor T. R. Lyle, for much kindly suggestion and practical aid. 
