and dale. Water barometers have been made on the same principle as the mercury barometer; but as water is 13½ times lighter, bulk for bulk, than mercury, the height of the column must be 13½ times 30 inches, or, allowing for variations, not less than 34 feet. This, of course, is very cumbrous; the evaporation of the water presents another considerable difficulty,[21] still such a barometer is a very interesting instrument, as it shows the atmospheric fluctuations on 13½ times the scale of the ordinary barometer. A range of about five feet is thus obtained; and not only the great waves, but even the comparatively small ripples of the atmospheric ocean are displayed by it. In stormy weather it may be seen to rise and fall and pulsate like a living creature, so sensitively does it respond to every atmospheric fluctuation. But why should the height of the barometer vary while it remains in the same place? If the quantity of air surrounding the earth remains the same, and if the barometer measures its weight correctly, why should the barometer vary? Does the atmosphere grow bigger and smaller, lighter and heavier, from time to time? These are fair questions, and they bring us at once to some of the chief uses of the barometer. The atmosphere is a great gaseous ocean surrounding the earth, and we are creeping about on the bottom of this ocean. It has its tides and billows and whirling eddies, but all these are vastly greater than those of the watery ocean. At one time we are under the crest or rounded portion of a mighty atmospheric wave, at another the hollow between two such waves is over our heads, and thus the depth of atmosphere, or quantity of air, above us is variable. This variation is the combined result of many co-operating causes. In the first place, there are great atmospheric tides, caused, like those of the sea, by the attraction of the sun and moon; but these do not _directly_ affect the barometer, because the attracting body supports whatever it lifts. Variations of temperature also produce important fluctuations in the height and density of the atmosphere, some of which are indicated by the barometer—others are not. Thus a mere expansion or contraction of _dry_ air, increasing the depth or the density of the atmospheric ocean, would not affect the barometer, as mere expansion and contraction only alter the _bulk_ without affecting the _weight_ of the air. But our atmosphere consists not only of the permanent gases, nitrogen and oxygen; it contains besides these and carbonic acid, a considerable quantity of gaseous matter, which is not permanent, but which may be a gas at one moment—contributing its whole weight to that of the general atmosphere—and at another moment some of it may be condensed into liquid particles that fall through it more or less rapidly, and thus contribute nothing to its weight.