Page images



54. This instrument could not have been long observed before the discovery that its fluctuations had some unknown connections with the changes of the weather in temperate climates, especially as regards wind and rain; a high state of the mercurial column generally occurring in the finest or calmest weather, and a depression of it during rain and storms. Hence, by a too hasty generalization, it was supposed that the fineness of the weather was exactly proportional to the atmospheric pressure, and, accordingly, such words as "fair," "changeable," "rain," &c., were engraved on the scale, which words have only served the purpose of bringing a really invaluable instrument into disrepute, by making it promise that which it is incapable of foretelling.

The reason why the atmospheric pressure is generally greater in dry situations, and in dry states of the weather than in moist ones, is still very obscurely known. We cannot even touch upon it in this place, because its consideration would require on the part of the reader a knowledge of the laws respecting vapour, and we prefer that he should gain from this little book a tolerably complete knowledge of the barometer, than an imperfect idea of the barometer and the hygrometer; but we may observe that a due attention to both these instruments will lead to more accurate predictions respecting the weather than can be obtained by the use of either of them singly. Indeed, the barometer used alone has, as we have endeavoured to show, a far more direct application to the theory of winds than to that of rain. Its application to the latter is only indirect, and far from being understood theoretically; still, however, the average of a large number of observations made at different times, and in different E 3

places, has furnished rules which deserve some degree of reliance.

The most important fact to be remembered is, that the state of the weather to be expected is not so much connected with the absolute height of the column as with its motion, whether rising or falling. In order to observe this most important fact an upright barometer is necessary, since the upper surface of the mercury cannot be seen in a wheel barometer. If the mercury have a convex surface the column is rising; if it is concave it is falling; when it is flat it is generally changing from one of these states into the other (24).

A fall in the mercury generally indicates approaching rain, high winds, or a thunder-storm; but it is remarkable that snow is more frequently preceded by a rise than by a fall. With this exception, however, a rising state of the mercury commonly indicates the approach of fine weather. A very high wind, especially from the S.W., whether accompanied by rain or not, is perhaps connected with the lowest state of the barometer. In England a N.E. wind is more conducive to a high state of the mercury than any other.

When the mercury rises or falls steadily for two or three days together, it is generally found that rather a long continuance of settled weather will follow; rainy in the latter case, and fine and dry in the former. By the same rule frequent fluctuations in the height of the column are found to coincide with unsettled weather.

55. Many persons are fond of entering the height of their barometers in a register once or twice a day for years together, and make no further use of these registers than to exhibit them to their friends as curiosities, and point out a remarkably low state of the



barometer at one period, and contrast this with a remarkably high state at another period. It may be thought a harsh word, but it is a fact that, as far as science is concerned, these registers are no better than waste paper; whereas they might be made of inestimable value by taking out the monthly and annual means, and sending them for publication to some local newspaper, or to any scientific journal of repute. Persons who have a tolerably good instrument, and the leisure and inclination for these observations, should make their entries at the proper hours of the day, and these are indicated by the instrument itself (49, 51). The maximum height of the column is about 9 o'clock A.M., the mean at 12, and the minimum at 3 P.M. If a person can afford time to make three observations every day, he should select these hours. If he can only make two observations, the proper periods are the very convenient hours of 9 A.M. and 9 P.M. he can make only one observation, noon is the time. Professor Daniell remarks, that those who merely consult the barometer as a weather-glass would find it an advantage to attend to the three above-mentioned periods, for he has noticed that by much the safest prognostications for this instrument may be formed from observing when the mercury is inclined to move contrary to its periodical course. If the column rise between 9 A.M. and 3 P.M. it indicates fine weather, if it fall from 3 to 9 rain may be expected.


56. The measurement of heights was the first useful purpose to which the barometer was proposed to be applied, preceding even its application as a weatherglass; and, in this respect, it is certainly more to be depended on than any predictions as to the weather, made from it even for only a few hours in advance. This application was suggested by the results of ex

periments performed by Torricelli and Pascal, that the mercurial column diminished in height on ascending above the level of the sea. But, at the outset of this inquiry, Pascal fell into a great error by supposing the atmosphere to be of equal density throughout, and that as the whole atmospheric column supported about 30 inches of mercury, all that was necessary was to observe the depression of the mercury on ascending a mountain, and then, by comparing the relative weights of mercury and air*, to ascertain the height of the mountain. The error, however, was soon discovered. Halley showed that the density of air decreases in a geometrical progression, while the elevation increases in an arithmetical progression; that is, if at a certain height the density was half that at the earth's surface, it would be one-fourth at twice that height; one-eighth at thrice that height, and so on; and Mariotte, about the same time, having determined that the pressure of aerial fluids is exactly proportional to their density when the temperature is equal (9), it was clearly proved by Halley that the ratio of decrease in the pressure was different from that of the increase in the heights. Indeed, if the upward diminution of the temperature be equal for equal ascents (34), it may be shown, for heights which are in arithmetical progression, that the elasticity diminishes in geometrical progression like the density, but rather more rapidly.

Now, the relation between an arithmetical and a geometrical progression, is the same as that between a series of logarithms and their natural numbers; and it occurred to Halley to apply a common table of logarithms to the solution of these questions. It was necessary, however, to fix the unit of his two series,

* This has been determined by Biot and Arago, to be 10.466 to 1.



which he did by calculating that the height at which the atmospheric pressure is exactly half that at the earth's surface, must be about 3 inches. That is to say, although the atmosphere may extend to the height of 45 miles, yet its lower half is so compressed as to occupy only 3 miles, so greatly do the upper portions expand when relieved from pressure.

3 miles, 7 miles, 10 miles, 14 miles, &c.,

Hence at the height of J the elasticity of the atmosphere is

[ocr errors]


16th, &c. Halley was induced, by certain mathematical considerations, to fix upon the number 62,170 as a constant multiplier, and the rule for the measurement of heights may be stated as follows:-Observe the height of the barometer at the earth's surface, and then at the top of the mountain, or other elevated station; take the logarithms of those numbers, and subtract the smaller from the greater; multiply the difference by 62,170, and the result is the height in English feet. This process gives a very near approximation, especially in temperate climates.

But the progress of science soon rendered it evident that a correction for temperature was necessary in barometrical measurements, and a formula has been contrived to meet most of the difficulties of the question. The following rule will be found of easy application :-Multiply the difference of the logarithms of the two heights by the barometer, by 63,946; the result is the elevation in English feet. Then, in order to correct for temperature, take the mean of the temperature at the two elevations; if that be 69.68° Fahr. no correction is necessary; if above that quantity, add 4th to the whole height found for each degree above 69.68; if below, subtract the same quantity. For


« PreviousContinue »