172 Mr J. D. FORBES on the Horary Oscillations 



to. in. in. in. 



same ratio, are 0.766 and 0.685, or as 1.1 : 1, this ratio being less 

 than one-fourth of the other. Yet the Paris observations for the 

 barometric tide are among the best we possess ; and these num- 

 bers are a mean from the registers of eleven years. Again, at the 

 very important station of the Grand St Bernard (to which M. 

 BOUVARD strongly appeals in support of his formula), where the 

 mean result is negative with regard to the oscillation at the level 

 of the sea, the mean temperature of the year being below cent., 

 therefore rightly expressed in the preceding Table, we should 

 have a very striking difference of result in summer and winter. 

 The temperature of the three summer months being about + 6 

 cent., and of the three winter months about 6 cent., the first 

 should give a considerable positive oscillation, the latter a consi- 

 derable negative one. Let us see, therefore, how the case stands. 

 For this purpose, I shall classify under the seasons the whole of 

 the monthly results which have been published, and of which 

 the last five years are complete. The sign of -f- in the following 

 Table indicates a rise between 9 and 3, that of a fall : the 

 former, therefore, is to be considered negative with regard to the 

 ordinary course, since the barometer rises when it usually falls at 

 the level of the sea. I may add, that, by taking the mean tem- 

 perature of the day in place of that of the particular period in 

 computing the oscillation by M. BOUVARD'S formula, here and at 

 Paris, we are doing it no injustice, as, though the summer and 

 winter temperature would both be somewhat higher *, the ratio 

 would be almost the same ; and it is about no evanescent diffe- 

 rences that we are disputing. 



* Indeed there seems much reason to doubt whether the period of the day al- 

 luded to has a mean temperature below cent. If it has not, M. BOUVARD'S for- 

 mula would totally fail, as the equatorial oscillation would come out with a wrong 

 sign. 3 



