to 100ths (.01) and 1,000ths (.001), and the number of feet corresponding to these amounts calculated from the table, which is easy enough.

TABLE II.-Correction due to Mean Temperatures of the Air; the Temperature of the Upper and Lower Stations being added and divided by 2.

TABLE III.— Correction due to Difference of Gravitation in different

Example.-At two stations the barometer read respectively 29.9 and 21.2, the temperatures of the air being 60° and 40°.

Final corrected height of upper station above sea-level. 9,511

A very simple rule for approximative determinations has been given by Mr. R. Strahan.' Read the aneroid to the nearest hundredth of an inch; subtract the upper reading from the lower, leaving out or neglecting the decimal point; multiply the difference by 9; the product is the elevation in feet.

If the barometer at the upper station is below 26 inches, or the temperature above 70°, the multiplier should be 10.

Weight of the Air.-The barometer expresses the weight of the air in inches of mercury. The actual weight can be determined if the reading of the barometer, temperature, and humidity are all known.

The weight of a cubic foot of dry air, at 32° Fahr. and normal pressure, is 566.85 grains. For any other temperature the weight can be calculated. Multiply the coefficient of the expansion of air (viz., .0020361 for 1° Fahr.) by the number of degrees above 32, the sum added to unity will give the volume of a cubic foot of dry air at that temperature. Divide 566.85 by the number so obtained. The result is the weight of the dry air at the given temperature.

SECTION IV.

RAIN,

Rain is estimated in inches; that is, the fall of an inch of rain implies that on any given area, say a square yard of surface, rain has fallen equal to one inch in depth. The amount of rain is determined by a rain-gauge.

1 Pocket Altitude Tables, by G. J. Symons, F.R.S., 3d ed., 1880, p. 5.

Two gauges are supplied for military stations; one to be placed on the ground, one 20 feet above it; in all parts of the world the latter indicates less rain than the lower placed gauge; this is due to wind.'

Several kinds of gauges are in use. The one used by the Army Medical Department is a cylindrical tin box with a rim or groove at the top; a circular top with a funnel inside fits on to this groove, which, when filled with water, forms a water-valve. The opening above is circular (the circle being made very carefully, and a rim being carried round it to prevent the rain-drops from being whirled by wind out of the mouth), and descends funnel-shaped, the small end of the funnel being turned up to prevent evaporation. But leaves, dust, or insects sometimes choke this tube, so that it is now generally straightened, the loss by evaporation being insignificant, compared with that caused by obstruction. The best size for the open top, or, in other words, the area of the receiving surface, is from 50 to 100 square inches. The lower part of the box is sunk in the ground nearly to the groove; the upper part is then put on, and a glass vessel is placed below the funnel to receive the water. At stated times (usually at 9 A.M. daily) the top is taken off, the glass vessel taken out, and the water measured in a glass vessel, graduated to hundredths of an inch, which is sent with the gauge.

3

2

If snow falls instead of rain, it must be melted and the resulting water measured. This may be easily done by adding a measured quantity of warm water, and then subtracting the amount from the total bulk of water.

From a table of the weight of vapor it will be seen that the amount of vapor which can be rendered insensible, increases with the temperature, but not regularly; more, comparatively, is taken up by the high temperatures; thus, at 40°, 2.86 grains are supported; at 50°, 4.10 grains, or 1.24 grain more; at 60°, 5.77 grains, or 1.67 grain more than at 50°. Therefore, if two currents of air of unequal temperatures, but equally saturated with moisture, meet in equal volume, the temperature will be the mean of the two, but the amount of vapor which will be kept invisible is less than

1 See British Rainfall (G. J. Symons, F.R.S.), 1872, p. 33, and 1881, p. 41. ? A glass vessel should not be used in winter, for fear of breakage in frost.

3 If this glass is broken it can be replaced by the following rule, or a rain-gauge can be made. It need not be round, though this is now thought the best form, but may be a square box of metal or wood, and may be of any size between 3 and 24 inches in diameter, but 5 to 8 is the most convenient range.

Determine the area, in square inches, of the receiving surface, or top of the gauge, by careful measurement. This area, if covered with water to the height of one inch, would give us a corresponding amount of cubic inches. This number of cubic inches is the measure for that gauge of one inch, because when the rain equals that quantity it shows that one inch of rain has fallen over the whole surface.

A

Let us say the area of the receiving surface is 100 square inches. Take 100 cubic inches of water and put it into a glass, put a mark at the height of the fluid, and divide the glass below it into 100 equal parts. If the rainfall comes up to the mark, one inch of rain has fallen on each square inch of surface; if it only comes up to a mark below, some amount less than an inch (which is so expressed in ths and oths) has fallen. To get the requisite number of cubic inches of water we can weigh or measure. cubic inch of water at 62° weighs 252.458 grains, consequently 100 cubic inches will be (252.458 × 100) = 25245.8 grains, or 57.7 ounces avoir. But an easier way still is to measure the water,- -an ounce avoir. is equal to 1.733 cubic inches, therefore divide 100 by 1.733, and we obtain the number of ounces avoir. which corresponds to 100 cubic inches. It is always best, however, to use a gauge made by a regular maker, if possible, as inaccurate records are worse than none.

Usually a one-inch measure is so large a glass, that half an inch is considered more convenient.

VOL. II.-8

the mean, and some vapor therefore necessarily falls as fog or rain. Thus one saturated current being at 40°, and the other at 60°, the resultant temperature will be 50°, but the amount of invisible vapor will not be the mean, viz., 4.315, but 4.1; an amount equal to .215 will therefore be deposited.

Rain is therefore owing to the cooling of a saturated air, and rain is heaviest under the following conditions,-when, the temperature being high, and the amount of vapor large, the hot and moist air soon encounters a cold air. These conditions are chiefly met with in the tropics, when the hot air, saturated with vapor, impinges on a chain of lofty hills over which the air is cold. The fall may be 130 to 160 inches, as on the Malabar coast of India, or 180 to 220 in Southern Burmah, or 600 at Cherrapoonjee, in the Khasyah Hills. Even in our own country the hot air from the Gulf Stream impinging on the Cumberland Hills causes, in some districts, a fall of 80, 100, 200, and even more inches in the year.

The rainfall in different places is remarkably irregular from year to year; thus at Bombay the mean being 76, in 1822 no less than 112 inches, while in 1824 only 34 inches fell.

The amount of rain at the different foreign stations is given under the respective headings.

SECTION V.

EVAPORATION.

The amount of evaporation from a given moist surface is a problem of great interest, but it is not easy to determine it experimentally, and no instrument is issued by the Army Medical Department. A shallow vessel of known area, protected around the rim by wire to prevent birds from drinking, is filled with a known quantity of water, and then, weekly or monthly, the diminution of the water is determined, the amount added by rain as shown by the rain-gauge being of course allowed for.

Water has been placed under a cover, which may protect it from rain and dew, and yet permit evaporation, and the loss weighed daily; but it is impossible to insure that the evaporation shall be equal to that under the free heavens.

A third plan is calculating the rate of evaporation from the depression of the wet bulb thermometer, by deducting the elastic force of vapor at the dew-point temperature from the elastic force at the air temperature, and taking the difference as expressing the evaporation. This difference expresses the force of escape of vapor from the moist surface.

Instruments termed Atmometers have been used for this purpose; the first was invented by Leslie. A ball of porous earthenware was fixed to a glass tube, with divisions, each corresponding to an amount of water which would cover the surface of the ball with a film equal to the thickness of Tooth part of an inch. The evaporation from the surface of the ball was then read off. Dr. Babington has also invented an ingenious Atmidometer.1

00

The amount of evaporation is influenced by temperature, wind, humidity of the air, rarefaction of the air, degree of exposure or shading, and by the nature of the moist surface; it is greater from moist soil than from water.

1 See Negretti & Zambra's Treatise, p. 141, for details.

The amount of vapor annually rising from each square inch of water surface in this country has been estimated at from 20 to 24 inches; in the tropical seas it has been estimated at from 80 to 130, or even more inches. In the Indian Ocean it has been estimated at as much as an inch in twentyfour hours, or 365 in the year, an almost incredible amount. No doubt, however, the quantity is very great.

It requires an effort of imagination to realize the immense distillation which goes on from the tropical seas. Take merely 60 inches as the annual distillation, and reckon this in feet instead of inches, and then proceed to calculate the weight of the water rising annually from such a small space as the Bay of Bengal. The amount is almost incredible.

This distillation of water serves many great purposes; mixing with the air it is a vast motive power, for its specific gravity is very low (.6230, air being 1), and it causes an enlargement of the volume of air; the moist air is therefore much lighter, and ascends with great rapidity; the distillation also causes an immense transference of heat from the tropics, where the evaporation renders latent a great amount of heat, to the extra-tropical region where this vapor falls as rain, and consequently parts with its latent heat. The evaporation also has been supposed to be a great cause of the ocean currents (Maury), which play so important a part in the distribution of winds, moisture, and warmth.

1

SECTION VI.

WIND.

Direction. For determining the direction of the wind a vane is necessary. It should be placed in such a position as to be able to feel the influence of the wind on all sides, and not be subjected to eddies by the vicinity of buildings, trees, or hills. The points must be fixed by the compass; the magnetic declination being taken into account; the declination of the place must be obtained from the nearest observatory; in this country it is now about 21° (or two points) to the westward of true north.* The direction of the wind is registered twice daily in the army returns, but any unsual shifting should receive a special note. The course of the wind is not always parallel with the earth; it sometimes blows slightly downward; contrivances have been employed to measure this, but the matter does not seem important.

Various plans are resorted to for giving a complete summary of the winds, but this is not required from the medical officer.

Velocity.-A small Robinson's anemometer is now supplied to each station; it is read every twenty-four hours, and marks the horizontal movement in the preceding twenty-four hours.

3

This anemometer consists of four small cups, fixed on horizontal axes of such a length (1.12 foot between two cups), that the centre of a cup, in one revolution, passes over th of a mile, the circumference being 3.52 feet. These cups revolve with about a third of the wind's velocity; 500 revolutions of the cups are therefore supposed to indicate one mile, and

1 Or, better still, by the pole star.

*Thus N. magnetic will be N. N. W. true, S. magnetic S.S. E. true, and so on.

The current of air is opposed one-fourth more by a concave surface than by a convex one the same size.