364 



Meteorologtj : 



[Oct. 



tome that the dates "middle of December" and "1st July" were precise- 

 ly those on w hich the earth arrived at its Perigee and Apogee with 

 respect to the Sun : we will now see how far a formula (1 — R) x —h' 

 — h will represent the observations ; where h represents the mean state 

 of the barometer, and h' any other measure corresponding to R the 

 radius vector of the Earth ; — forming the equations of conditions, x 

 conies out 9,17, inches with which we cRn now compute the height of the 

 barometer for any time, from the formula h' = h + (1— R) thus:— 



MEAN HEIGHT OF THE BAROMETER, AT 10 A. M. 





From observation. 



From formula. 



Difference. 





Inches. 



Inches. 



Inches. 



January . . 



30.118 



30.095 



- ,023 



February. . 



30.060 



30.056 



—,004 





30.007 



29.989 



—.018 





29.935 



29.i>07 



—,028 





29.831 



29.840 



+ ,009 





29.798 



29.797 



—,001 





29.800 



29.796 



—,004 





29.847 



29.826 



— ,021 



September. 



29.877 



29.903 



4- ,026 





29.935 



29.980 



+ ,045 



November 



30.070 



30.052 



—,018 



December.. 



30.076 



30.094 



+ ,018 



Were it not that the month of October gives a difference so large in 

 itself, and so much at variance with the following one for Novem- 

 ber, there would be little fault to be found with the formula. Since 

 coming to this conclusion, however, I have examined all the barometri- 

 cal observations made at Madras, during the last forty years, when, 

 much to my surprize, the result for the month of October is nearly 

 confirmed, but that for November is different — thus 



From 40 years 1 observation. From formula. 



Inches. Inches. DifT. 



October 29.966 29.980 +.034 



November... 30.005 30.052 +.047 



If these two differences be substituted for those set down in the above 

 table, a much greater degree of consistency is produced, although the 

 differences are larger. That the mean state of the barometer, resulting 

 from four years' observation, should differ to the amount ,065 from the 

 general mean, would point out, that, during the month of November in 

 each of those four years, some considerable atmospheric derangement 

 must have existed j that such has actually been the case, is well 

 known ; for, on consulting the depth to which rain has fallen in 

 November, we find the average of 40 years = 14.13 inches 

 in 1832-33-34 and 35 = 7.54 „ 



