Clouds, balloons, and kites naturally supplement one another. While clouds indicate the direction and velocity of the air at different heights, yet the lower clouds often conceal the upper strata, or there may be no clouds at all, in which case balloons or kites will aid us to determine the drift of the currents. When there is little wind at the ground, or to reach heights of several miles, we must employ balloons, but otherwise kites are preferable in most cases. The thermal and hygrometric conditions of the free air can be ascertained only by personal observations in balloons, or by raising self-recording instruments with balloons and kites, and this latter method it is predicted will be the path of greatest progress. CHAPTER II CLOUDS--FORMATION AND CLASSIFICATION--MEASUREMENTS AT BLUE HILL--THE INTERNATIONAL OBSERVATIONS Clouds must have been among the earliest observed natural phenomena, and they were used from time immemorial as weather signs. Yet their every-day occurrence was very likely the reason why their origin was not studied until about a century ago. Father Cotte, in his classic work on meteorology, published in 1774, devotes only a couple of paragraphs to clouds, but AbbÈ Richard, in his contemporary _Histoire Naturelle de l'Air_, discusses the appearance and theories of clouds in ten chapters. The cause of evaporation was unknown in the last century, and it was not until its close that Dalton, the English chemist, proved that water-vapour exists independently in the air, and Hutton explained that precipitation was produced by the contact of a current of saturated air with a colder one. Although there remains much to be learned about cloud formation, yet it is now pretty well established that its most effective cause is the ascent, and consequent cooling by expansion of the air, rather than the mixture of masses of air having different temperatures. The ascent of the air may result from its being forced up a mountain slope by its horizontal movement, or from its being drawn up in a vortex, but most commonly the air rises from its lessened specific gravity when warmed. If the temperature of the quiescent air decreases faster than 1∞ for each 183 feet of height, which is the adiabatic rate of cooling for dry air, as explained in the last chapter, air warmed locally will rise and cool at this rate until the dew-point is reached. Then the vapour in the air will be condensed upon particles of dust, which Aitken found to be more numerous in clouds than outside them. The most conspicuous of the clouds formed by rising currents is the cumulus, or rounded summer cloud, which has been aptly termed "the visible capital of an invisible column of air." Saturated air cools as it rises more slowly than dry air, consequently the upward motion is maintained through the cloud mass, causing the swelling up of the tops of the cumulus clouds, which reach their highest development in the thunder clouds, or