The Weather. 



107 



The daily oscillation of temperature, at Green wi en, from 

 the average of five years (bi-hourly observations, 1841 — 1845), 

 is represented by the following table, above and below the 

 mean temperature for the period. 



Table III. — Diurnal curve of temperature at Greenwich 

 for five years (1841 — 1845) : — 



(F.) 





Deg. 







Deg. 



1 h. a.m. 



— 3-4 



(F.) 



1 h. p.m. 



+ 5-1 



3h. „ 



— 4-2 





3h. „ 



+ 5-8 



5L „ 



— 4-8 





5 h. „ 



+ 4-4 



7h. „ 



— 3-9 





7h. „ 



+ 2-0 



9h. „ 



— 1-2 





9h. „ 



— 0-6 



11 h. „ 



+ 2-3 





11 h. „ 



— 2-3 



About an hour before sunrise (5 h. a.m.) is the coldest time, 

 and three o } clock in the afternoon is the hottest hour of the 

 day. The average daily range is 10 o, 6 (F.). These hours and 

 the daily range vary very considerably in the different months. 

 Wherever observations are continued at a particular hour of 

 the day, special tables are therefore required for every month, 

 to deduce the ' c mean monthly temperature " from the mean 

 temperature observed at any hour. Where, however, self- 

 registering maximum and minimum thermometers are used, a 

 simpler method is to ascertain the average highest day tem- 

 perature, and the average lowest night temperature for the 

 month. The arithmetical mean between these quantities is 

 always higher than the true mean monthly temperature of the 

 air by the amount contained in the following table. 



Table IY. — Correction to be applied subtractively to the 

 arithmetical mean between the average highest day tempera- 

 ture and the average lowest night temperature for any month, 

 observed with self-registering thermometers, to obtain the 

 true mean monthly temperature : — 



Deg. 





Deg. 





Deg. 



Jan. . — 0-2 (F.) 



May 



— 1-7 (F.) 



Sep. 



— 1-3 (F.) 



Feb. —0-4 



June 



— 1-8 



Oct. 



— 1-0 



Mar. — 1-0 



July 



— 1-9 



Nov. 



— 0-4 



April — 1*5 



Aug. 



-1-7 



Dec. 



— o-o 



Twelve mean monthly temperatures having been obtained by 

 either mode, the average of all is the mean temperature for the 

 year. This is found to be a very constant quantity, although 

 the annual range of temperature is subject to very considerable 

 variations. 



The mean yearly temperature of the air having been deter- 

 mined by observation at different places of the globe, lines of 



