No. 18. [From Nature, August 17, 1899, Vol. LX, 

 PP- 3 6 4-3 6 5 ] 



THERMOMETRIC SCALES FOR 

 METEOROLOGICAL USE 



IN the course of some recent work on the meteorology of 

 Ben Nevis, which involved extensive extracting and computing 

 work, I have again had forcibly impressed on me the great 

 advantage which Fahrenheit's thermometer has over that of 

 Celsius for meteorological use, especially in temperate regions. 



In chemistry and physics the range of temperature 

 covered is so great that Celsius' scale, which is now uni- 

 versally used, adequately meets every case. The size of the 

 degree and the change of sign at the melting-point of ice 

 do not cause any inconvenience in the laboratory. It is 

 otherwise in the meteorological observatory. There the range 

 of temperature dealt with is very restricted, and the Celsius 

 degree is too large, while the change of sign in the middle 

 of the working part of the scale is simply intolerable. The 

 latter peculiarity is the fruitful introducer of error into both 

 the observations and the reductions, and besides it greatly 

 increases the fatigue of both classes of work. 



In view of the agitation to abolish the use of Fahrenheit's 

 scale, and to replace it universally by that of Celsius, it may 

 not be inopportune to direct attention to some of the ad- 

 vantages in securing accuracy and in relieving labour which 

 Fahrenheit's scale offers over that of Celsius when used for 

 meteorological purposes. 



In tropical countries it matters little whether one scale 

 or the other is used, except that the size of Fahrenheit's 

 degree is much the more convenient, as the first decimal place 

 is always sufficient. But in Europe and in North America, 

 where the greater number of meteorological observatories 



