In practice no great error will be made in assuming the relative humidity to be fifty per cent. For the moisture content never exceeds five per cent of the mass of the moist air, and hence in assuming a fifty per cent relative humidity, when there is actually a maximum or minimum humidity, the greatest possible error in estimating the moisture content is 2.5 per cent of the mass of moist air. Now if 2.5 per cent of a mass of air be assumed to be aqueous vapor when all is really dry air, or conversely if 2.5 per cent of the whole mass be assumed as dry air when it is really aqueous vapor, an error of much less than 2.5 per cent is made in estimating the true density. No error at all would ensue if both air and vapor were of the same density; but since one is ⅝ as heavy as the other, the possible error is ⅜ of 2.5 per cent, or 0.6 per cent. This is a negligible quantity in all mechanical considerations, except where great accuracy is required. When any gas changes density or volume it also changes temperature, unless there be transfer of heat between it and its environment. When change of volume occurs without such transfer of heat the expansion, or contraction, is called “adiabatic;” when it occurs at constant temperature, the expansion is called “isothermal,” the temperature being kept uniform by suitable transfer of heat; when it occurs at constant pressure it is called “isopiestic.” In either case work may be done by the enlarging gas, if it press against a moving piston, or yielding envelope of some kind; and conversely work may be spent on the gas in compressing it either isothermally, adiabatically or isopiestically. If, for example, a balloon rises rapidly its contents will expand adiabatically, pushing the envelope out in all directions against the static pressure of the embracing atmosphere. Thus it will do work and rapidly cool. But if it rapidly sinks, it will contract adiabatically and grow warm, owing to the work done by the surrounding air in compressing it. A like thing occurs when a great volume of air rises or sinks quickly in the free atmosphere. In this case the change of temperature is about 6° C. for each kilometer change of level, so long as the air remains unsaturated. A familiar example of this effect in Nature is manifested when an uprushing column of moist air chills, and precipitates moisture, forming a cloud toward its top. Thus a lone thundercloud in a clear sky may mark the upper part of such a column, or upward vortex in the air. And contrarywise, a descending column may absorb its visible moisture, causing it to become clear aqueous vapor, and thus vanish from view. CHAPTER XIV GENERAL DISTRIBUTION OF HEAT AND PRESSURE