Sir D. Brewster on the Law of Daily Temperature. 345 



11 hours 11 minutes; at Inverness, 11 hours 13 minutes*; 

 at Kingussie, 10 hours 44 minutes ; at Belleville, 11 hours 

 14 minutes; at Tweedsmuir, 11 hours 15 minutes; at Ply- 

 mouth, 11 hours; at Philadelphia, 11 hours 20 minutes; at 

 Trincomalee, 11 hours 5 minutes; at Colombo, 10 hours 55 

 minutes; at Kandy, 11 hours; and at Madras, 10 hours. 



Another general law which we obtained from the hourly 

 observations at Leith is, that the half sum of the mean tempe- 

 ratures of any two hours of the same name, differs at an ave- 

 rage only a quarter of a degree of Fahrenheit from the mean 

 temperature of the year, the maximum difference being 0, i21, 

 and the minimum o, 059, and the differences following a re- 

 gular law. The following are the results of Jour years' hourly 

 observations : — 



Differences between the half sum and the 



Mean difference 0-264 0369 



The law of the differences is here very interesting. At Leith 

 the mean temperature of 4 h. a.m. and 4 h. p.m., and of 10 h. 

 a.m. and 10 h. p.m., approaches nearest to that of the year; 

 and at Inverness the hours of 9 h. a.m. and 9 h. p.m., and of 

 3 h. a.m. and 3 h. p.m. On this subject Humboldt remarks : — 



" We are surprised, at the first glance, by the generality of 

 this law. The homonymous hours are very unequally distant 

 from the hour of the maximum of the daily temperature, and 

 the hours of equal temperature, (we may say by analogy with 

 the praclidt- of astronomers in the determination of the true 

 time, the corresponding thermometrical altitudes,) give for each 

 place an epoch very different from that of the maximum. It 

 it is a thing truly remarkable, that from the mean of two or- 



* Deduced from three years' hourly observations, made at Inverness by 

 Mr. Thomas Mackenzie of Rayning's School, and at the expense of the 

 British Association. The hours of mean temperature are 8 h. 28 m. a.m., 

 and 7 h. 41 m. 



