7. Though the immediate cause of the destruction or reproduction of great masses of air at certain times, when the wind changes from north to south, or from south to north can not yet be ascertained; yet as there appears greater difficulty in accounting for this change of wind for any other known causes, we may still suspect that there exists in the arctic and antarctic circles a BEAR or DRAGON yet unknown to philosophers, which at times suddenly drinks up, and as suddenly at other times vomits out one-fifteenth part of the atmosphere: and hope that this or some future age will learn how to govern and domesticate a monster which might be rendered of such important service to mankind. INSTRUMENTS. If along with the usual registers of the weather observations were made on the winds in many parts of the earth with the three following instruments, which might be constructed at no great expence, some useful information might be acquired. 1. To mark the hour when the wind changes from north-east to south-west, and the contrary. This might be managed by making a communication from the vane of a weathercock to a clock; in such a manner, that if the vane mould revolve quite round, a tooth on its revolving axis should stop the clock, or put back a small bolt on the edge of a wheel revolving once in twenty-four hours. 2. To discover whether in a year more air passed from north to south, or the contrary. This might be effected by placing a windmill-sail of copper about nine inches diameter in a hollow cylinder about six inches long, open at both ends, and fixed on an eminent situation exactly north and south. Thence only a part of the north-east and south-west currents would affect the sail so as to turn it; and if its revolutions were counted by an adapted machinery, as the sail would turn one way with the north currents of air, and the contrary one with the south currents, the advance of the counting finger either way would shew which wind had prevailed most at the end of the year. 3. To discover the rolling cylinders of air, the vane of a weathercock might be so suspended as to dip or rise vertically, as well as to have its horizontal rotation. RECAPITULATION. NORTH-EAST WINDS consist of air flowing from the north, where it seems to be occasionally produced; has an apparent direction from the east owing to its not having acquired in its journey the increasing velocity of the earth's surface; these winds are analogous to the trade-winds between the tropics, and frequently continue in the vernal months for four and six weeks together, with a high barometer, and fair or frosty weather. 2. They sometimes consist of south-west air, which had passed by us or over us, driven back by a new accumulation of air in the north, These continue but a day or two, and are attended with rain. See Note XXV.