_Frost._—The appearance of frost in the form of hoar-frost (the way in which atmospheric water-vapour is deposited in air below the freezing-point), or of thin ice formed on exposed water, should always be carefully looked for and noted. In hot, dry countries the intense radiation from the ground at night often reduces the temperature below the freezing-point, although, during the day, the ground may be very hot. The appearance of frost at sunrise is a valuable check on the readings of a minimum thermometer, and in most cases is a more trustworthy datum. Similarly in cold countries, where snow is lying on the ground or ice covering the rivers, the appearance of thaw, especially in cloudy weather, is a delicate test of the rise of the air temperature to the freezing-point. The traveller should never fail to record cases of melting and solidifying of any substances due to changes of temperature. The softening of candles and the freezing of mercury or of spirits give information regarding temperature at least as valuable as the readings of thermometers. _Other Observations._—Any peculiar atmospheric phenomena, such as the appearance of the zodiacal light after sunset, the aurora, the electrical lights seen on pointed objects, and known as St. Elmo’s fire, rainbows, especially lunar rainbows, haloes, the appearance of mock-suns or moons, meteors or shooting stars, should be noted on their occurrence, as many of them are valuable weather prognostics. Attention should also be given to any appearances of mirage, or other effects of irregular distribution of atmospheric density. A mirage is only rarely so perfect as to show ships inverted in the air, palm-grown islands in the sea, or distant oases in the desert. The common form is an unusual intensification of refraction, raising land below the horizon into sight, or apparently cutting off the edges of headlands or islands at sea or on large lakes. It is worth while observing the temperature of the air and of the water or ground when an unusually clear mirage effect is visible. Another interesting series of observations may be made on the colours of the sky and clouds at sunrise and sunset. A phenomenon often observed at sunset, but the existence of which is still sometimes denied, should be looked for. This is the appearance of a gleam of coloured light at the moment when the upper edge of the sun dips below a cloudless horizon. A note should be made of the nature of the horizon, whether land or sea, and of the colour of the light if it should be observed. When opportunity offers, the first ray of the rising sun might be similarly observed.