mented; the weight of a cubic foot of dry air at 60° Fahr. is 536.28 grains, and that of a cubic foot of vapor at 60° is 5.77 grains; the conjoint weights would be 542.05 grains at 60°, but, owing to the enlargement of the air, the actual weight of a cubic foot of saturated air at 60° is only 532.84 grains.

00

SECTION III.

BAROMETER.

A good mercurial barometer is supplied to many army stations; the scale is brass, graduated on the scale to 20ths or half-tenths, and is read to Toths by means of a vernier. There is a movable bottom to the cistern, which is worked up and down by a screw, so as to keep the mercury in the cistern at the same level. Correction for capacity is thus avoided.

To fix the Barometer.-Choose a place with a good light, yet protected from direct sunlight and rain; fix the frame sent with the barometer very carefully with a plumb-line, so as to have it exactly perpendicular; then hang the barometer on the hook, and adjust it gently by means of the three screws at the bottom, so that it hangs truly in the centre. Test this by the plumb-line (a 4 oz. weight tied to a string will do), and then unscrew the bottom of the cistern till the ivory point is seen.

Before fixing the barometer the bottom should be unscrewed till the mercury is two or three inches from the top; the barometer should be rather suddenly inclined, so as to let the mercury strike against the top; if there is no air it will do this with a sharp click; if there be air there is no click; in that case turn the barometer upside down, and tap the side forcibly till you see the globule of air passing up the tube through the mercury into the cistern. Do not be afraid of doing this; there is no danger of any damage to the instrument.

Reading of Barometer.-Read the attached thermometer first; then adjust the cistern, so that the ivory point, perceptible through the glass wall of the cistern, seems just to touch the point of the image in the mercury. Then adjust the vernier, so as to cut off the light from the top of the mercury. Then read the scale with the help of the vernier.

00

A little difficulty is sometimes experienced, by those who are not accustomed to such instruments, in understanding the vernier. It will be, probably, comprehended from a little description, read with the instrument before us. On the scale of the barometer itself, it will be seen that the smallest divisions correspond to half-tenths; that is, to ths of an inch (=.05). The height of the mercury can be read thus far on the scale itself. The vernier is intended to enable us to read the amount of space the top of the mercury is above or below one of these half-tenth lines. It will be observed that the vernier is divided into twenty-five lines; but on adjusting it, so that its lower line corresponds with a line indicating an inch, it will be seen that its twenty-five divisions only equal twenty-four half-tenth divisions on the scale. The result is, that each division on the vernier is th less than a half-tenth division on the scale. One th of a half-tenth is ths of an inch (.05÷25=.002 inch). This being understood, adjust the vernier so that its lowest line accurately corresponds to any line on the scale. It will then be seen that its lowest line but one is a little distance below (in fact, .002 inch) the next line on the fixed scale.

Raise now the vernier, so that its second line shall correspond to the line on the scale to which it was a little below; and of course the bottom of the vernier must be raised .002 inch above the line it first corresponded with. If the next line, the third on the vernier, be made to correspond with the line on the scale just above it, the bottom of the scale must be raised double this (.004 inch) above the line it was first level with; if the next line on the vernier be made to correspond with a line on the scale, the scale is raised .006, and so on. Each division on the vernier equals .002 inch, and each five divisions equals th, or .01 inch.

The barometer is read thus. The vernier being adjusted to the top of the mercury, read on the scale to the half-tenth; then look above, and see what line on the vernier corresponds exactly to a line on the scale. Then read the number on the vernier, counting from the bottom; multiply by .002, and the result is the number of thousandths of an inch the top of the mercury is above the half-tenth line next below it. Add this number to that already got by direct reading of the fixed scale, and the result is the height of the mercury in inches and decimals of an inch.

Corrections for the Barometer.-The barometer supplied to military stations requires no corrections for capacity. There are two constant corrections for all barometers, viz., capillarity and index error. The first depends on the size of the bore, and whether the mercury has been boiled in the tube or not. Index error is determined by comparison with a standard barometer. The index and capillarity errors are put together. The capillarity error is always additive; the index error may be subtractive or additive, but the two together form a constant quantity, and the certificates furnished by the Kew Observatory, for all barometers verified there, include both corrections above mentioned.

Corrections for Temperature.--The barometer readings are, to facilitate comparison, always reduced to what they would have been were both scale and mercury at 32° F. If the temperature of the mercury be above this, the metal expands, and reads higher than it would do at 32°. The amount of expansion of mercury is .0001001 of its bulk for each degree; but the linear expansion of the brass scale must be also considered.

Schumacher's formula is used for the correction-viz.,

h = observed height of barometer in inches.

t = temperature of attached thermometer (Fahr.).

m = expansion of mercury per degree-viz., .0001001 of its length at 32°.

8 linear expansion of scale-viz., .00001041; normal temperature being 62°.

To facilitate the correction for temperature, tables are given in Mr. R.

Instead of multiplying the number on the vernier by .002, a little practice will enable the calculation to be made at once. On the vernier will be seen the figures 1, 2, 3, 4, and 5; corresponding to the 5th, 10th, 15th, 20th, and 25th lines, and indicating .01, .02, .03, .04, or .05 inch. Each line between these numbered lines equals .002 inch.

H. Scott's "Instructions in the Use of Meteorological Instruments," which is distributed to medical officers.

Correction for Altitude above Sea-level. As the mercury falls about Too (.001 inch)' for every foot of ascent, this amount multiplied by the number of feet must be added to the height, if the place be above sea-level." The temperature of the air has, however, also to be taken into account if great accuracy is required. Tables for correcting for small altitudes are given in Scott's "Instructions."

When all these corrections have been made, the exact height of the mercury represents the conjoint weights of the oxygen, nitrogen, carbon dioxide, and watery vapor of the atmosphere. It is difficult to separate these several weights, and late observations, which show that the humidity existing at any place is merely local, and that vapor is most unequally diffused through the air, render it quite uncertain what amount of the mercury is supported by the watery vapor. Yet that this has a considerable effect in altering the barometric height, particularly in the tropics, seems certain (Herschel).

The height of the barometer at sea-level differs at different parts of the earth's surface; being less at the equator (29.974) than on either side of 30° N. and S. lat., and lessening again toward the poles, especially toward the south, from 63° to 74° S. lat., where the depression is upward of an inch. It also differs in different places according to their geographical position. Like the thermometer, it is subjected to diurnal and annual periodic changes and to non-periodic undulations.

In the tropics the diurnal changes are very steady; there are two maxima and two minima; the first maximum is about 9 a.m.; the first minimum about 3 to 4 P.M.; the second maximum at 10 P.M.; the second minimum at 4 A.M. These changes are, perhaps, chiefly dependent on the watery vapor (Herschel). In this country the diurnal range is less, but occurs at about the same hours. The undulations depend on the constantly shifting currents of air, rendering the total amount of air over a place heavier or lighter. The wind tends to pass toward the locality of least barometric pressure. In this country the barometer falls with the southwest winds; rises with the north and east; the former are moist and warm, the latter dry and cold winds.

Isobarometric lines are lines connecting places with the same barometric pressure.

Measurement of Heights.-The barometer falls when heights are ascended, as a certain weight of air is left below it. The diminution is not uniform, for the higher the ascent the less weighty the air, and a greater and greater height must be ascended to depress the barometer one inch. This is illustrated by the following table: "

1 The exact amount is a little below this, but varies with altitude; at sea-level the amount is .000886 for every foot of ascent.

2 For the British Isles, the mean sea-level at Liverpool has been selected by the Ordnance Survey as their datum.

3 The height can be taken readily from this table, by calculating the number of feet which must have been ascended to cause the observed fall, and then making a correction for temperature, by multiplying the number obtained from the table, which may be called A, by the formula (t is the temperature of the lower, and t' of the upper station)

The measurements of heights in this way is of great use to medical officers; aneroid barometers can be used, and are very delicate instruments. The new pocket aneroids will measure up to 12,000 or 14,000 feet.

A great number of methods are in use for calculating heights. It can be done readily by logarithms, but then a medical officer may not possess a table of logarithms.

The simplest rule of all is one derived from Laplace's formula. Mr. Ellis' has stated this formula as follows:-Multiply the difference of the barometric readings by 52,400, and divide by the sum of the barometric readings. If the result be 1,000, 2,000, 3,000, 4,000, or 5,000, add 0, 0, 2, 6, 14, respectively. Subtract 24 times the difference of the temperatures of the mercury. Multiply the remainder by a number obtained by adding 836 to the sum of the temperatures of the air, and dividing by 900. A correction must also be made for latitude, which can be done by Table III., p. 111.

Tables such as, those given by Delcros and Oltmanns are very convenient for estimating heights by the barometer. A table less long than these, but based on the same principle, has been given by Negretti & Zambra in their useful work,' and is copied here.

A good mercurial barometer, with an attached thermometer, or an aneroid compensated for temperature, and a thermometer to ascertain the temperature of the air, are required. Two barometers and two thermometers, which can be observed at the same moment at the upper and lower stations, are desirable.

Supposing, however, there is but one barometer, take the height at the lower station, and correct for temperature to 32°. Take the temperature of the air. Ascend as rapidly as possible to the upper station, and take the height of the barometer (correcting it to 32°) and the temperature of the air; then use the accompanying tables, taken from Negretti & Zambra's work. If the height is less than 300 feet, Tables II., III., and IV. need not be used.

=

"Table I. is calculated from the formula, height in feet 60,200 (log. 29.922-log. B)+925; where 29.922 is the mean atmospheric pressure at 32° Fahr., and at the mean sea-level in latitude 45°; and B is any other barometric pressure; the 925 being added to avoid minus signs in the table.

1 Proceedings of the Royal Society, 1865, No. 75, p. 283.

A Treatise on Meteorological Instruments, by Negretti & Zambra, 1864.

TABLE I-Approximate Height due to Barometric Pressure.

"Table II. contains the correction necessary for the mean temperature of the stratum of air between the stations of observation; and is computed from Regnault's co-efficient for the expansion of air, which is .002036 of its volume at 32° for each degree above that temperature.

'Table III. is the correction due to the difference of gravitation in any other latitude, and is found from the formula, x = 1+.00265 cos. 2 lat. "Table IV. is to correct for the diminution of gravity in ascending from the sea-level.

To use these tables: The barometer readings at the upper and lower stations having been corrected and reduced to temperature 32° Fahr., take out from Table I. the numbers opposite the corrected readings of the two barometers, and subtract the lower from the upper. Multiply this difference successively by the factors found in Tables II. and III. The factor from Table III. may be neglected unless great precision is desired. Finally, add the correction taken from Table IV." (Negretti & Zambra.)

In the table the barometer is only read to 10ths, but it should be read