of the Barometer near Edinburgh. 161 



ever been kept in any part of Britain *. In the comparisons I 

 am about to institute with the observations made in other parts 

 of the world, I shall adopt the oscillation between 10 and 4, or 

 .0106 inch, since this agrees nearly with the usual hours of 

 observation, as it has rarely happened that meteorologists have 

 taken the trouble to observe the barometer at two periods near 

 one another. The amount just named exceeds somewhat the 

 oscillation observed by SOMMER at Kdnigsberg, in Latitude 

 54 42', which has hitherto been referred to as the solitary result 

 from northern Europe f. We shall afterwards see that the 

 amount I have assigned agrees much more closely with the strict 

 analogy of other observations in more southern latitudes. Besides, 

 the hour at which the afternoon minimum was observed by 

 SOMMER was much too early in all probability, being part of the 

 time at 2 p. M., and the remainder at 3 p. M. The hour in tem- 

 perate climates is probably later than is generally assigned : nor 

 does it add much to our confidence in the results, that BESSEL 

 informs us, that " notwithstanding the minuteness of the amount, 

 it is recognised in each annual mean :{:," when in my tables it is 

 eliminated in 28 out of 38 monthly means, notwithstanding its 

 extreme minuteness in winter. In London, where the amount 

 is more than twice as great as here, and the observations are made 

 only twice a-day, but with such regularity as to avoid the neces- 

 sity of any artifice for correcting omissions, I find the oscillation 

 detected in 46 monthly means out of 48 ; and at Paris, in every 

 one of 132 months ||, the amount there being nearly three times 

 that at Edinburgh. 



* I do not except Mr DANIELL'S Observations for reasons assigned below, . 9. 

 f HUMBOLDT, Relation Historique, 4to edition, torn. iii. p. 306. 

 J Astronomische Nachrichten, 1823, No. 26. 

 Philosophical Transactions for 1827, &c. 



i| See M. BOUVABD'S Memoir in the Memoires de 1'Institut for 1824. 



3 



