of the Barometer near Edinburgh. 171 



11. I have now some more particular observations to add 

 upon the theoretical part of the Table. \st, I conceive that it 

 is little to be expected, or perhaps desired, that one formula 

 should represent the decrement both in latitude and height, and 

 the change at different hours and seasons, which, even in the 

 much better known facts of terrestrial temperature, has never 

 been, nor satisfactorily could be, executed. I conceive that, in 

 the purely physical bearings of the question, the existence of 

 such a common law would be surprising, because such a coinci- 

 dence must have been derived from the concurrence of indepen- 

 dent agencies. 2d, The observations at considerable heights, in 

 different latitudes, are too few, too brief, and too inconsistent, to 

 afford us any thing like satisfactory results. For example, the 

 observations at Caracas, at a height of 3000 feet, give a greater 

 amount of oscillation than those at Cumana, at the level of the 

 sea, at the same latitude. 3d, M. BOUVARD proposes his formula 

 as capable of expressing the variation observed at any period of 

 the day or year, and has shewn that the ratio of the different 

 mean temperatures of the periods of forenoon and evening oscil- 

 lations, correspond tolerably nearly with that of the observed 

 amount of these oscillations, at least in Europe ; but I would 

 ask, Why do not the great differences of temperature, corres- 

 ponding to the seasons, give the marked results which his formula 

 would suppose, producing a tropical amount in summer, and an 

 arctic one in winter ? At Paris, for instance, where the mean of 

 the summer temperatures, that is, of the months of June, July, 

 and August, is, according to HUMBOLDT*, 18.l Cent.; and of 

 the winter months (December, January, and February), 3.7 

 Cent., the ratio is 4.9 : 1 ; whilst the corresponding oscillations, 

 which, according to M. BOUVARD'S formula, ought to have the 



* Des Lignes Isothermes. Mem. d'Arceuil, torn. iii. Table. 



