If a quantity of air exists under a certain pressure and at a certain temperature, on reducing the pressure it will expand, and in the act of expanding it will become cooler. This may easily be illustrated with an air-pump. Let a damp sponge or a piece of wet blotting-paper stand under a glass receiver over an air-pump until the air has become damp. If the apparatus is in a darkened room, and a powerful beam of light from a lantern is sent through the receiver, the damp air will be seen to be quite clear; but a stroke or two of the pump removes some of the air, the remainder is chilled by its own expansion, and a dense cloud is precipitated. If this cloud be viewed closely, it will be seen to be composed of minute particles, which, on looking towards the light, glow with the colours of a corona. In a few minutes the cloud will disappear, but it can be recalled again and again by successive strokes of the pump, getting thinner and thinner as the air gets more and more rarefied; an illustration of a second reason why the high clouds are thinner than the lower. Some years ago Mr. John Aitken showed that if the damp air used in this experiment were carefully filtered, so as to remove all foreign particles, no cloud was produced, and the introduction of a puff of unfiltered air was attended by immediate condensation. The deduction was that vapour, even below its saturation temperature, cannot produce cloud unless nuclei of some sort are already present, presumably dust particles. Later on it was shown by Mr. Shelford Bidwell and others that gaseous particles, such as those produced by the burning of sulphur, would serve the purpose, and that the brush discharge from an electrified point was in some mysterious way particularly effective. It has recently been shown by Mr. C. T. R. Wilson that causes such as the radiations of radium, or the impact of ultra-violet rays, acting on the air itself, splits up some of its particles into the smaller bodies known as ions, and that these are efficient nuclei. These experiments open up many most interesting questions, but, unless it is to explain the extreme density and darkness of a thunder-cloud, they do not seem to play any important part in determining the forms to be assumed. Nuclei in sufficient abundance are probably always present at any height which can be reached by enough vapour to form a cloud.