CHAPTER XV. 



METEOROLOGICAL REPORT — BY A. DE C. SOWERBY. 



IVAETEOROLOGICAL observations were commenced at T'ai-yiian Fu, in 



Shansi, on May i6th, igo8, and were kept up — with as much care and 



regularity as circumstances would permit — throughout the whole duration of 



the expedition, i.e., till September 12th, 1909. The instruments used were : — 



One portable mercurial barometer. 



Three aneroid barometers. 



One hypsometer, with spare thermometer. 



One small psychrometer (wet and dry bulb thermometers). 



Three swing psychrometers. 



Six swing thermometers. 



One maximum and one minimum thermometer. 



One six-inch diameter rain gauge. 

 All thermometers were graduated on the Fahrenheit system, while the 

 barometers, boiling point thermometers, and swing thermometers were 

 provided with Kew certificates. 



The barometer observations were made either with the mercurial 

 barometer or with an aneroid. In the former case the readings have been 

 corrected for temperature, so that they represent the true air pressure at 

 station level ; and in the latter case, comparisons between the mercury 

 barometer and the aneroids were carried out at intervals, in order to guard 

 against errors due to changes of zero in the aneroids. 



The readings of dry and wet bulbs printed in the diary were made with 

 swing thermometers. This was necessary on account of the inevitable absence 

 of a suitable screen for the exposure of the instruments. The reduction of 

 the readings in order to determine the humidity of the air will therefore 

 require tables appropriate to the well-ventilated psychrometer. The tables 

 commonly used in England, which are based on Glaisher's Factors, are not 

 suitable for the reductions, and would give values of the relative humidity 

 lower than the true values. The tables of the U.S. Weather Bureau, which 

 were prepared for use with sling psychrometers, are more appropriate. 



The readings of the maximum and minimum thermometers are probably 

 less reliable than those of the dry and wet bulbs, because it was necessary to 

 improvise some shelter for them, and some of the readings may be affected by 



135 



