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



the parabola, and to the form of the visible horizon. We have 

 therefore made the necessary calculations, and obtained the 

 following results : — 



It appears from this table that in three of the four parabolic 

 branches the deviation from the parabola is considerably less 

 than in the Leith observations ; whereas in the afternoon 

 branch and in that part of it between the hours of four and eight, 

 the deviation is nearly double of what it is in the other branches 

 and in the Leith observations. The observed temperatures 

 are less than they should be, and we can account for this 

 only by the interposition of some obstacle in the azimuth of 

 four, five, six, and seven o'clock p.m., or from the thermo- 

 meter being better protected from extraneous influences at 

 these particular hours*. 



As there are many meteorological registers which have been 

 kept only at two hours of the day, it became of importance to 

 deduce from them the approximate hours of mean temperature, 



* We find upon looking at the Register, that the thermometer is placed 

 in the shade in a window facing the north. The line of the building is 6° 

 15' to the north of due west, and hence the thermometer is better protected 

 from the sun in the afternoon than in the morning. 



