212 W. MAKRIOTT — METEOROLOGICAL OBSERVATIONS. 



reduce it to a fixed temperature, viz. the freezing-point, 32°. The 

 mode of taking the observation is this : — First note the reading of 

 the attached thermometer to the nearest degree ; then (if the 

 barometer is a Fortin) adjust the mercury in the cistern by turning 

 the screw at the bottom so that the ivory point is just brought 

 into contact with the surface of the mercury, but does not depress 

 it ; the ivory point and its reflected image in the mercury should 

 appear to touch each other and form a double cone. The next 

 thing is to adjust the vernier so that the lower edge shall form a 

 tangent to the convex surface of the mercury — iu fact, the front 

 and back edges of the vernier and the top of the mercury must be 

 in the same straight line. The scale on the tube is divided to 

 inches, tenths, and half-tenths, or five-hundredths of an inch, and 

 the vernier is made equal to 24 divisions of the scale, and is di^-ided 

 into 25 equal parts. Each division of the vernier is therefore 

 smaller than each division of the scale, by the 25th part of -05 ; 

 which is -002 inch. First read ofi^ the division on the scale at, or 

 below which the lower edge of the vernier stands. Suppose it is 

 between 29-70 inches and 29-75 inches ; we should call this 29-70 

 inches. Next look along the vernier until one of its lines is found 

 to agree with a line on the scale. Suppose this is at the second 

 division above the figure 2. As each of the figures marked on 

 the vernier count as hundredths, and each intermediate di\'ision as 

 two thousandths, our reading of the vernier will be -024 inch. 

 The reading of the barometer is therefore 29-70 + -024 = 29-724 

 inches. 



Having obtained the actual reading of the barometer, it now 

 requires corrections for (1) index error, (2) temperature, and (3) 

 height above sea-level. 



1. The correction for index error is ascertained by comparing 

 the barometer with a standard instrument, such as that at the 

 Kew Observatory ; the correction is always given on the certificate 

 of comparison. 



2. As the mercury in the barometer expands by heat, the 

 height of the column is aifected by temperature ; it is therefore 

 necessary to reduce the readings to a uniform temperature : 32"" is 

 the point now adopted. Tables of corrections have been prepared 

 for this purpose, and will be found in most treatises on Meteorology. 



3. In comparing barometric observations made at ditt'ereut 

 places, account must be taken of their respective heights above sea- 

 level ; for, the higher the station is, the lower -will the reading of 

 the barometer be. The heiglit of the cistern of the barouieter 

 above sea-level must therefore be accurately ascertained, which 

 can be done by levelling, etc., from the nearest bench-mark. The 

 proper corrections for altitude can then be calculated from the 

 tables given in most treatises. 



A form of table has been devised for combining all the above 

 corrections, and is the means of saving many troublesome cal- 

 culations, besides greatly reducing the liability to error. (See 

 ' Quarterly Journal of the Meteorological Society,' vol. iii., p. 1.) 



