In a paper, read for me to the Society in January last (ante, p. 245), I pointed out the difficulties I had met with in trying to reconcile Van der Waals' equation with Amagat's experimental data for common liquids, and I promised to recur to the question when the state of my health should permit. I now find that, as I had then only surmised, the constants in Van der Waals' equation necessarily become non-real when we try to adjust it to Amagat's data.