197 



27. — Reduction of Meteorological Observatioxs. 



By William Marriott, F.M.S. 



Communicated by J. IIopkinson, Hon. Sec. 



[Read 13lh May, 1879.] 



Having on a former occasion read a paper before this Society 

 entitled "Instructions for taking Meteorological Observations,"* I 

 have now the pleasure of supplementing it, at the request of your 

 Secretary, by another on the "Reduction of Meteorological Obser- 

 vations." 



The readings of nearly all meteorological instruments require 

 some correction to give true results. The object of the "reduction" 

 is to eliminate the instrumental and other errors, so that the 

 readings may be comparable with those made at other places, and 

 also to deduce further results from the observations. 



As it is hardly possible to manufacture the instruments abso- 

 lutely perfect, it is necessary that they should be compared with 

 a recognised standard. The Kew Observatory possesses such 

 standards, and has a special department for the verification of 

 meteorological instruments. A moderate charge is made for the 

 comparison and a certificate given of the corrections to be applied. 

 It is essential that all the instruments employed be sent to Kew 

 for verification before being used, so that the proper corrections to 

 be applied to their readings may be known. As mercurial thermo- 

 meters are liable through age to read too high, it is desirable that 

 they be re-verified after the space of two or three years. 



These corrections may, no doubt, appear to some to be very 

 trivial matters and not worth taking notice of ; this may be so 

 for very rough, purposes : but if the observations are made for 

 scientific purposes, and are to have any pretence to accuracy, so as 

 to be comparable with those at other places, then it is important 

 that the corrections should be duly applied, otherwise the deduc- 

 tions drawn from the readings may be misleading and erroneous. 

 For instance, suppose the dry-bulb thermometer to have a correc- 

 tion of -j-0'^-5, and the wet-bulb — 0^*5, and the readings to be 

 50°-0 and 46°-0; the calculated dew-point for the latter would be 

 41'^-8, and the relative humidity 74 ; but if the proper corrections 

 were applied, the readings would then be : dry-bulb 50°'5, wet- 

 bulb 45''-5, dew-point 40''"2, relative humidity 68. 



Barometer. — There are three corrections to be applied to the 

 readings of the barometer, viz. for (1) index error, (2) temperature, 

 and (3) height above sea-level. 



(1). The correction for index error is that given on the Kew 

 certificate ; if the barometer be a Fortin, the correction will be the 

 same throughout the scale : but if of the Kew pattern, it will most 

 likely be variable. As the corrections are only given for every 



* ' Transactions,' Vol. I, p. 211. 



VOL. II. — I'T. VI. 16 



