^{page 15 note *} Weid. Ann., xxxii. pp. 1–19, 1887.

^{page 15 note †} Weid. Ann., xl. pp. 161–202, 1890.

^{page 16 note *} There is another way of considering the whole question, from which it would appear that there is no such thing as supersaturation in a vapour; that is, no strain in the vapour which either dust or “molecular shocks” can relieve. What is generally called a saturated vapour is one whose tension is equal to the tension of the vapour at a flat surface. Now this tension is not so high as the tension at a surface of extremely small convex curvature; and vapour that is in equilibrium with the vapour at a convex surface is supersaturated to the flat surface; so that saturation is a relative and not an absolute quantity, relative to the curvature of the condensing surface, and a vapour that is supersaturated to a flat surface is not necessarily saturated to a surface of very small curvature. It would thus appear that there is no strain in a vapour till a surface makes its appearance; but after it is formed the lower tension at its surface determines a movement of the vapour molecules towards it.

^{page 18 note *} It was necessary to be constantly adding cupric sulphate to the acid in A. If this was not done, considerable condensation took place in the products.

^{page 21 note *} Proc. Roy. Soc., Edin., vol. xx. pp. 66, 93; Trans. Roy. Soc., Edin., vol. xxxvi., Part III., pp. 621, 693.