HOURLY CORRECTIONS FOR PERIODIC VARIATIONS, 



CORRECTIONS TO BE APPLIED TO THE MEANS OF THE HOURS OF OBSERVATION, OK 



SETS OF HOURS, IN ORDER TO OBTAIN THE TRUE MEAN TEMPERATURES 



OF THE RESPECTIVE DAYS, MONTHS, AND OF THE YEAR. 



The following set contains all the tables for correcting the means of observations 

 on atmospheric temperature for the effect of diurnal variation which have been pub- 

 lished by Dove, together with a few others of the same description. Dove's tables 

 are found in two papers, published in the Memoirs of the Royal Academy of Berlin 

 for 1846 and for 1856, and in the first Report on the Observations of the Meteoro- 

 logical Institute of Prussia, Berlin, 1851. 



In the first paper are twenty-nine tables, in Reaumur's scale, nine of which have 

 been republished, in Fahrenheit's scale, in the Proceedings of the British Associa- 

 tion for 1847, and will also be found below. In that series the corrections have 

 been formed by finding first the differences between the hourly and the true means, 

 and then computing the observations by Bessel's formula, in order to eliminate the 

 accidental irregularities due to the shortness of the period during which the observa- 

 tions were taken. Calling x the horary angle reckoned from noon, Bessel's for- 

 mula is 



tx = u-{-u' sin {x + U') + u" sin (2 x + U") + u'" sin (3a; + U'"). 



The stations at which hourly observations were made are Trevandrum, Madras, 

 Bombay, Salzuflen, Prague, St. Petersburg, Catharinenburg, Barnaul, Nertchinsk, 

 Matoschkin-Schar, Strait of Kara, and Boothia Felix. Bi-hourly observations were 

 taken at Brussels, Greenwich, and Toronto ; in all others the night observations are 

 wanting, and were obtained by interpolation. Moreover, in several stations the num- 

 ber of observations was small, at Madras even only thirty-six days. The tables of 

 that series may be readily distinguished from those belonging to the same stations in 

 the second, by their containing the corrections for several sets of hours, which are 

 not found in the tables of the other. 



In Dove's second series, and in all other tables, the corrections given are simply 

 the differences, with reverse signs, between the hourly and the true means, excepting, 

 however, the stations of Toronto, in which the corrections were computed, by Bes- 

 sel's formula, by Colonel Sabine ; of Prague, by Jelineck ; of Salzburg, and those of 

 Geneva and St. Bernard, by Plantamour. 



The observations from which these tables are derived were made hourly at Hobar- 

 ton during 8 years ; at the Cape of Good Hope, for 5|- years ; St. Helena, 5 years ; 

 Madras, 5 years ; Bombay, 4 years ; Calcutta, IJ- years ; Toronto, 6 years ; Phila- 

 delphia, 3 years ; Makerstoun, 3 years ; Utrecht, If years ; Prague, 104- years ; 

 Munich, 7 years ; Salzburg, 6 years ; St. Petersburg, 10 years ; Catherinenburg, 6 

 years ; Barnaul, 5 years ; Tiflis, 4 years ; Nertchinsk, 6 years ; Peking, 4 years ; 

 Sitka, 5 years. In the following stations the observations were bi-hourly : — Wash- 

 ington, for 14- years ; Greenwich, 7 years ; Dublin, 4 years ; Brussels, 9 years ; Ge- 

 neva and St. Bernard, 4 years ; Schwerin, 3 years. 



The observations made in England, and in her colonies, are found in the various 

 government publications. Those of the Russian stations are taken from the Annuaire 

 Meieorologique et Magnetique des Jngtnieurs des Mines, and in the Annales de 



E 11 



