The experiments referred to in my paper of March 6th (anté, p. 65) have been completed, but the results are by no means so exact as I hoped to make them. There was great difficulty in procuring the small bore tubes for the piezometers, and thus I had to employ them without previous calibration, as the solutions to be experimented on had already been prepared, and their densities determined at definite temperatures. Delay might have led to evaporation. When I proceeded to the calibration, after completing a large series of experiments, I was greatly annoyed to find that the bores of many of the tubes were by no means uniform. This accounts for the fact that my experiments, though fairly concordant, are not sufficiently so to afford more than a very strong probability in favour of the general result of the inquiry. For this reason I have described my paper as a Preliminary Note.