be a less quantity of air above the former, than above the latter instrument, and consequently, the barometer in the attic will read lower than that on the ground floor.

Precisely the same reasoning will apply if we are considering two barometers at different heights on the side of a mountain, or one at an inland station, and the other at the level of the sea; in every instance the barometer at the higher station will read lower than that at the lower.

Hence we find that all readings must be reduced to their equivalent readings at a standard elevation, as well as at a standard temperature; this standard elevation is the mean level of the sea, and the reduction is carried out by means of tables.

A slight consideration of the foregoing remarks, and of the last-named principle in particular, will show how it comes about that the reading given in a Daily Weather Report for an inland station, like Oxford, or Nottingham, can never agree with the actual reading taken by an observer at either of those stations at the same time, unless the latter reading be corrected and reduced to 32°, and to sea level.

Temperature. This does not require much explanation. In addition to the actual readings of the thermometers, 'in the shade' (i.e., suspended in a properly constructed louvre-boarded screen at a height of four feet above the ground), which are taken at certain definite hours, the Daily Weather Reports give also two columns, showing respectively, the maximum and minimum read

ings taken during the twenty-four hours, ending at 8 o'clock each morning, by thermometers so constructed as to register, one the highest, and the other the lowest, temperature reached during the interval which has elapsed since the last setting.

The last-named readings may also be turned to very useful purpose. The mean temperature of the day is often mentioned in connection with sanitary statistics, as for instance in the Registrar-General's Reports. This mean temperature is theoretically the average of twentyfour thermometer readings taken at hourly intervals during the day. There are well-known rules for determining this mean temperature out of various combinations of hours of observation. The simplest method, and, for practical purposes, nearly the most exact, is to take the average of the maximum and minimum. readings given in the Daily Weather Report, on any day, and put that down as the mean temperature of the preceding day.

Humidity. The amount of moisture in the air is measured by the reading of a thermometer with its bulb encased in muslin, and kept damp (a wet bulb), compared with that of an ordinary (dry bulb) thermometer, without any such mounting, taken at the same moment.

This is not the place to enter into the theory of these observations. Suffice it to say, the greater the difference between the readings of the two thermometers the drier is the air, and when the two thermometers read

alike, the atmosphere is exceedingly damp. The difference between the two thermometers ranges in this country from 0° to 10° or even 15°, and sometimes a difference of upwards of 20° has been noticed.

The chance of rain depends to a great extent on the degree of humidity of the air from time to time, and if we are dealing with reports from an extensive tract of country, as North America, or the Continent of Europe, the distribution of the moisture, or of the vapour tension will afford great assistance in tracing out the probable motion of storms. As regards these islands, the fact that most of the telegraphic stations are on the sea-coast detracts from the value of the reports of humidity, inasmuch as the amount of moisture in the air is seriously affected by proximity to the sea; and so we are unable to use reports of vapour tension in weather study as much as our neighbours.

Rain is measured by means of a rain-gauge. The values given represent the depth of water which would have accumulated on a level piece of ground, if none of the rain which fell could have escaped by drainage, etc. Too much faith must not be placed on the rain returns in the daily reports, for two reasons. Firstly, the gauges are often necessarily placed in towns, where a good exposure is not to be had, so that the amount measured is not quite the same as might be yielded by a gauge in a more open situation; and secondly, the stations are so sparsely distributed over the kingdom that it is impossible that they should give a precise account

of the rain-fall in every county. What they do show is whether or not the rain is general and heavy.

Wind. As a general rule, for the purposes of the Daily Weather Report, this is not measured by an instrument, but is simply estimated according to the Beaufort scale, which is so named after the late Admiral Sir Francis Beaufort. The following is the scale,' with the approximate equivalent velocity of the wind in miles per hour, as determined in the Meteorological Office.

10. Whole gale

Close-reefed topsails

and courses

56

Or, that with which she could scarcely bear closereefed main-top-sail and reefed foresail

65

Or, that which would reduce her to storm-stay-sails. 75
Or, that which no canvas could withstand

90

1 Since Admiral Beaufort's time there has been a great change in the rig of merchant ships by the introduction of double topsail yards. It seems therefore advisable to add to Beaufort's scale the amount of sail which his ship would have been able to carry had she been rigged with double topsail yards, but under all other circumstances the same. The change would only affect forces 6 to 10.

6. Topgallant sails.

7. Topsails, jib, &c.

8. Reefed upper topsails and courses.

9. Lower topsails and courses.

10. Lower main-top-sail and reefed foresail.

It is obvious that as this scale refers to the rate of sailing of, or to the amount of sail carried by, a ship, it is at first sight not well suited for use at land stations; but experience has shown that the reporters' estimates are never very far from the truth, when once they have gained some practice in observing. There are objections to supplying instruments for the measurement of wind to telegraphic stations, owing to the great difficulties which would be experienced in erecting them in suitable positions. Anemometrical indications taken in the middle of a town are almost worthless, as the buildings produce such eddies that the true movement of the air cannot be ascertained from the instrument.

As regards the numbers of the Beaufort scale, it is those from 6 upwards which present the most interest. 6 is the lowest number which is taken in the Meteorological Office to justify the issue of a warning to the coast, and 9 is the lowest figure which by the regulations of the Board of Trade can be pleaded by a captain as 'stress of weather,' in case of casualty to his craft. These velocities are not uniform, like that of an express train. The statement of 56 miles an hour, means 56 miles in the hour, but during that hour the wind may have been gusty, and at times have had a velocity of near 100 miles an hour, while at other times its hourly speed may have scarcely reached 30 or 40 miles.

It may be remarked that if we employed the pressure, instead of the velocity of the wind, to measure the violence of a gale, we should be able to obtain a record