Height correction. As the pressure of the air is reduced as we rise above sea level it follows that there will always be a correction to be applied whenever the cistern is above sea level. The correction from this cause amounts to about .001 inch for each foot above sea level and is always additive. Gravity correction.—The earth being a spheroid flattened at the poles, the polar radius is less than the equatorial radius, and as the force of gravitation varies inversely as the square of the distance from the centre of mass and also as centrifugal force is directly opposed to gravity at the equator and vanishes at the poles, it follows that a column of mercury is shorter at the poles than it would be at the equator; and as it is necessary for the sake of precision in comparison that some latitude shall be agreed upon as a standard for the measurement of weight, the latitude chosen is 45°. From what has been said it is clear that a column of mercury between the equator and 45° N. or S. is longer than it would be at the standard latitude, and in order to reduce it to standard” a minus correction would have to be applied, whereas a column of mercury between 45° N. or S. and the pole is shorter than at the standard latitude, and the correction to reduce it to “standard ” would be plus. C 5 SETTING AND READING A BAROMETER When setting the vernier and reading the barometer great care should be taken to have the eye on the same level as the top of the mercury, and the tube must be allowed to hang vertically in the gimbals and not held A in the hand while setting the vernier, because any inclination of the tube causes 1394 vernier should be set for reading when the с mercury has completed its downward D In Fig. 2, the bottom of the vernier (D) having been brought into coincidence with the top of the mercury, the scale 8 line 29:50 exactly coincides with the zero Fig. 2. of the vernier, and whenever this is the case the uppermost graduation (C) on the vernier should always coincide with a scale graduation as shown. In this example the reading is 29.5, or 29:50, or 29.500 inches. In Fig. 3 the bottom of the vernier (D) is above 30.00 B Fig. 3. and below 30:05; it is therefore obvious that the reading lies between 30.00 inches and 30.05 inches. Looking carefully at the graduations on the vernier it is seen that the figure 3 nearly coincides with a scale gradua. tion; the figure 3 indicates .03; on a closer examination it is seen that the graduation on the vernier half-way between 3 and 4 is in exact agreement with a scale reading; this graduation represents .005, so that D is 035, or thirty-five thousandths of an inch above the scale reading next below D. The reading can be put down in the following form Reading on scale in inches ... 30.000 30.035 Fig. 4 shows the graduations of a Kew pattern barometer; it is graduated on one side in centibars with millibar divisions, and on the other in inches with twentieths of an inch divisions. This drawing is an exact reproduction of a barometer of the latest pattern, and shows the graduations exactly as they appear on the instrument; the equivalent number of millibars corresponding to any reading of inches and tenths being the same as shown by the instrument, but it should be observed that this does not agree with the Table of Equivalents of inches and millibars as shown on page 9 of the Seaman's Handbook of Meteorology. The barometer in this figure is reading at 30-5 inches, which corresponds to 1,030-8 millibars. The Standard Temperaturestamped on the instrument is the temperature at which a reading of 1,000 millibars of pressure would be correct at sea level in lat. 45° N. or S. It is, therefore, the “ Fiducial Temperature” for that instrument in lat. 45°, and the cistern of the barometer at sea level. Should the barometer not be at sea level a new " Fiducial Temperature” will have to be found by adding 1° for every five feet, or I'5 meters, that the cistern is above sea level, to the temperature stamped on the instrument, the result will be the Fiducial Temperature for that height in lat. 45° N. or S., when the pressure is 1,000 millibars. The Fiducial Temperature is different in different latitudes, and it is recommended to compile, by means of Table I., a table of "Fiducial Temperatures ” from the Equator to 75o N. or S., making due allowance for the height of the cistern above sea level. To correct a reading the observer has then only to consider the difference between the “ Actual Temperature,” as shown by the attached thermometer (Fig. 5), and the “Fiducial Temperature” shown by the table for that particular latitude The correction for this difference of temperature amounts to 'I millibar for every 6 of a degree of temperature that the Actual differs from the Fiducial, or i millibar for every 6o of difference; when the attached thermometer reads higher than the Fiducial Temperature the correction is subtractive, and when the attached thermometer reads lower than the Fiducial Temperature, the correction is additive. Fahrenheit Centibars, th millibar divisions. Inches, with zb-inch divisions. Centigrade from the Absolute Zero. Fig. 4. Fig. 5. SUPPLEMENTARY CORRECTIONS FOR SPECIAL ACCURACY The correction in Table II. is in reality a fractional part of the pressure, and ought to be adjusted proportionally for different points in the range of atmospheric pressure. The adjustment is made by adding 1 per cent. to the quantity taken from Table II. for each centibar above too, and subtracting i per cent. for each centibar the barometer reading is below 100 centibars. One per cent. only begins to be appreciable when the correction from Table II. is about 10 millibars and, except on rare occasions, may be neglected. CORRECTION FOR SCALE ERROR This can be provided for by the table of Kew corrections, which gives the Standard Temperature at different points of the scale. A properly graduated scale should have the same Standard Temperature throughout its range. If correction for Standard Temperature in different parts of the scale be necessary, it can be found from Table II. Example of scale correction. Barometer M. has Standard Temperature 286° A. at 1,000 millibars, and 280° A. at 900 millibars. Find the scale correction when the reading is 920 millibars. The Standard Temperature at 920 millibars would be 281° A. or 5° below standard conditions, and is equivalent to reducing the “Fiducial Temperature,” by 5°, which, in Table II. A., gives •8 millibar to be subtracted from the reading. (A.) ACTUAL TEMPERATURE ABOVE THE FIDUCIAL TEMPERATURE. (B.) ACTUAL TEMPERATURE BELOW THE FIDUCIAL TEMPERATURE. Relation of millibars to inches. (See Fig. 4). --The units on the absolute scale are related to one another as follows: IO millibars = I centibar. = I bar. The millibar is adopted as the working unit in the Daily Weather Service. The scale of millibars is related to the conventional scale of mercury inches as follow: Thus one-tenth of a millibar corresponds with '003 mercury inch, which may be taken as the limit of accuracy to which it is possible to read a barometer under favourable conditions. A few examples will now be given of the method of correcting barometer readings recommended by the Meteorological Committee. Barometer M. 12 metres (40 feet) above sea level in lat. 52° N. reads 1013'1 millibars; attached thermometer 285° A. Find the Fiducial Temperature in lat. 52o and correct the reading, the Fiducial Temperature for that instrument in lat. 45° at sea level being 286° A. Fiducial Temperature in lat. 45° at sea level 286° A. Correction for 12 metres + 8° Fiducial Temperature in lat. 45° at a height of 12 metres 294° A. Latitude correction, Table I. + 4° Fiducial Temperature in lat. 52° at a height of 12 metres Actual Temperature Attached Thermometer 298° A. 285° A. Diff. 13° Uncorrected reading 1013.1 millibars. Correction for 13° diff. between Actualand Fiducial Tempera tures from Table II. B. + 2.2 millibars 101503 Proportional adjustment i per cent. of 22 negligible Scale Error-nil Barometer MA. 50 feet above sea level in lat. 60° S. reads 1,028 millibars; attached thermometer 278° A. Find the Fiducial Temperature for that latitude, and correct the reading, the Fiducial Temperature for that instrument in lat. 45° at sea level being 285° A. |