40S Graphical Representation of the daily Rate of the [Dfic. 



it asserts nothing more than can he proved by experiments, and 

 we cannot yet prove hy experiments, that all the phenomena in 

 nature can be solved by mere mechanical causes ? Certainly 

 these things deserve to be better considered." * 



Article II. 



A Graphical Representation of the daily Rate of the Barometer 

 during a Year in London, Paris, and Geneva. By J. P. 

 Pictet, Assistant Professor of Natural Philosophy in Geneva.f 



The representation in Plate XIV. was made out by the author 

 upon a very small scale from a great collection of observations of 

 the rate of the barometer at London, Paris, Geneva, and Ma- 

 drid, laid before the Natural History Society of Geneva. It 

 contains a complete year of observations, made in the three first 

 named cities. The upper half of the copper-plate reaches from 

 the autumnal equinox of 1S06 to the vernal equinox of 1807, 

 the lower half extends from this equinox to the autumnal 

 equinox of the same year. " We see here very evidently," says 

 M. Pictet, " how much more the barometer rises and falls in 

 winter than in summer. The complete harmony which exists 

 between the three barometrical curves, and the very regular 

 parallelism of their inflections is surprising, when we consider 

 the great distance of the three places from each other. % 



In January and February we find the greatest, in July and 

 August the smallest, alterations in the height of the barometer. 

 Very high and very low positions correspond in all the three 

 places within a day. The figures upon the plate mark the day 

 of the month on which these maxima and minima were ob- 

 served. An exception is observable in October, 1806. The 

 barometer was highest in London on the 24th, at Paris on the 

 25th, and at Geneva on the 26th. Likewise in January, 1807, 

 it was highest at London on the 5th, in Paris on the 6th, and at 



* Phil. Trans. 1714, vol, xxix. p. 173. 



+ This curious paper was published in the Bibliotheque Britannique for 

 Jan. 1811. 1 have translated it from Gilbert's Anualen for May, 1812. It is 

 evident that the German editor has made considerable alterations in the ori- 

 ginal.— T. 



t The same thing holds good in another set of barometrical curves con- 

 structed by M. Pictet upon a much larger scale, embracing a longer period of 

 time, and including likewise a curve for Madrid. 



When these curves are compared, it appears that all the principal inflexions 

 agree very nearly with each other in time and amount : only, the farther north 

 a place lies, the sooner docs the change begin at it. 



