Magnesia dust has small affinity for water-vapour; accordingly, it condenses at almost exactly the same temperature as the glass. But gunpowder has great condensing power. All have noticed that the smoke from exploded gunpowder is far more dense in damp than in dry weather. In the experiment it will be found that the dust from gunpowder smoke begins to show signs of condensing the vapour at a temperature of 9 deg. Fahr. above the dew-point. In the case of sodium dust, the vapour is condensed from the air at a temperature of 30 deg. above the dew-point. Dust collected in a smoking-room shows a decidedly greater condensing power than that from the outer air. We can now understand why the glass in picture frames and other places sometimes appears damp when the air is not saturated. When in winter the windows are not often cleaned, a damp deposit may be frequently seen on the glass. Any one can try the experiment. Clean one half of a dusty pane of glass in cold weather, and the clean part will remain undewed and clear, while the dusty part is damp to the eye and greasy to the touch. These observations indicate that moisture is deposited on the dust-particles from air, which is not saturated, and that the condensation takes place while the air is comparatively dry, _before_ the temperature is lowered to the dew-point. There is, then, no definite demarcation between what seems to us clear air and thick haze. The clearest air has some haze, and, as the humidity increases, the thickness of the air increases. In all haze the temperature is above the dew-point. The dust-particles have only condensed a very small amount of the moisture so as to form haze, before the fuller condensation takes place at the dew-point. At the Italian lakes, on many occasions when the air is damp and still, every stage of condensation may be observed in close proximity, not separated by a hard and fast line, but when no one could determine where the clear air ended and the cloud began. Sometimes in the sky overhead a gradual change can be observed from perfect clearness to thick air, and then the cloud. A thick haze may be occasioned by an increased number of dust-particles with little moisture, or of a diminished number of dust-particles with much moisture, above the point of saturation. The haze is cleared by this temperature rising, so as to allow the moisture to evaporate from the dust-particles. Whenever the air is dry and hazy, much dust is found in it; as the dust decreases the haze also decreases. For example, Dr. Aitken, at Kingairloch, in one of the clearest districts of Argyleshire, on a clear July afternoon, counted 4000 dust-particles in a cubic inch of the air; whereas, two days before, in thick haze, he counted no fewer than 64,000 in the cubic inch. At Dumfries the number counted on a very hazy day in October increased twenty-fold over the number counted the day before, when it was clear.