It has long been known that, if air saturated with water-vapour be suddenly expanded, a cloud of small globules of water is formed. These drops are formed round the dust particles present in the gas, which act as nuclei for the condensation of water around them. The experiments of R. von Helmholtz and Richarz[65] had shown that chemical reactions, for example the combustion of flames, taking place in the neighbourhood, affected the condensation of a steam-jet. Lenard showed that a similar action was produced when ultra-violet light fell on a negatively charged zinc surface placed near the steam-jet. These results suggested that the presence of electric charges in the gas facilitated condensation. A very complete study of the conditions of condensation of water on nuclei has been made by C. T. R. Wilson[66]. An apparatus was constructed which allowed a very sudden expansion of the air over a wide range of pressure. The amount of condensation was observed in a small glass vessel. A beam of light was passed into the apparatus which allowed the drops formed to be readily observed by the eye. Preliminary small expansions caused a condensation of the water round the dust nuclei present in the air. These dust nuclei were removed by allowing the drops to settle. After a number of successive small expansions, the air was completely freed from dust, so that no condensation was produced. Let _v₁_ = initial volume of the gas in the vessel, _v₂_ = volume after expansion. If _v₂_/_v₁_ < 1·25 no condensation is produced in dust-free air. If however _v₂_/_v₁_ > 1·25 and < 1·38, a few drops appear. This number is roughly constant until _v₂_/_v₁_ = 1·38, when the number suddenly increases and a very dense cloud of fine drops is produced. If the radiation from an X ray tube or a radio-active substance is now passed into the condensation vessel, a new series of phenomena is observed. As before, if _v₂_/_v₁_ < 1·25 no drops are formed, but if _v₂_/_v₁_ = 1·25 there is a sudden production of a cloud. The water drops of which this cloud is formed are finer and more numerous the greater the intensity of the rays. The point at which condensation begins is very marked, and a slight variation of the amount of expansion causes either a dense cloud or no cloud at all.