Extremely Dilute Acid and Alkali Solutions. 13 
these experiments are much more uncertain on account of the 
danger of adsorption by the glass vessels used. For obtaining any 
‘trustworthy results, it would be necessary to employ vessels of a 
- ke that does not behave in this way. For the present, 
therefore, we shall mention merely general results. 
Alkali solutions show a drop in the &/m curves similar to that 
found for acids, and we should therefore expect a similar explana- 
tion. Hence the suggestion that the kation we are dealing with 
is ammonium, since unionised NH,.OH would be produced on the 
addition of a strong alkali, just as onicatsed H,CO, is produced on 
‘the addition of a strong acid. If there be an excess of free 
-earbonic acid in the distilled water, the ‘loss of conductivity’ is 
still more obvious. The fact that both NH, and CO, are normal 
le constituents of the atmosphere lends a good deal of support to 
the suggestion. 
Again, if a simple conductivity curve be drawn with Kohl- 
| Bich: s figures for KOH, the result is very nearly a straight line, 
but not as accurate a one as the corresponding H,SO, curve. It 
appears to be approaching concavity at greater concentrations 
‘than in the case for the acid. On the ammonium carbonate 
hypothesis, this is indeed what we should expect; for though 
: NH,OH is a weak base, it is more easily lonised than H »CO,; so 
| that we cannot expect the bending away from the straight line at 
the foot of the conductivity curve to be so sudden; it should begin 
at greater concentrations and proceed more gradually. 
_ The hypothesis we have been developing demands that if we 
could work with water of the degree of purity attained by Kohl- 
-rausch, preventing the CO, and NH, of the air from reaching the 
water at any stage of the experiments, the drop in the k/m curve 
would be pushed into a region of much smailer concentration,— 
or, more exactly, the ‘loss of conductivity’ would be much 
reduced. In Kohlrausch’s experiments (Fig. 1), this loss is 
2:7 x 10-* reciprocal ohm; in our experiments (Table IJ) it is 
about 08x 10-*. It could never be made to disappear entirely 
however, on account of the real ionisation of the water itself, and 
of the recombination of the H and OH ions on the addition of an 
acid. / 
