1876.] 



Variations of the Barometer in India. 



29 



The daily mean height of the barometer, derived from 24 hourly 

 observations for each civil day and for each station, are projected, p. 28. 

 The relative positions and approximate distances (in English miles) of 

 the three stations are given in the preceding figure. 



The first movements to be considered in the figures (p. 28) are those shown 

 by the dotted curves, which represent the mean barometric height for 27 

 days, including 13 days before and 13 days after each point. In these 

 means the variations due to the solar rotation and lunar revolution 

 periods may be considered approximately eliminated. It will be seen 

 that at all the stations the height diminished from the 9th or 10th of 

 January till the beginning of April : this movement is part of the annual 

 variation, which is most marked at Madras, amounting to 0*3 inch from 

 January to June, while it is only 0'06 inch at Singapore. At Simla in 

 1845 the range was about 0*25 inch, but less regular than at the other 

 stations*. 



The dotted curves show movements which appear independent of the 

 regular annual variation, and which are similar at all the stations ; these 

 will require a larger series for their consideration. 



"When we examine the change of daily mean barometric height, three 

 well-marked movements are to be seen at all the stations. The first ter- 

 minates about the end of January, the second in the beginning of 

 March, and the third towards the end of March ; these movements are, 

 I believe, due to the solar rotation, and perhaps the lunar re\'olution ; 

 they will be considered in another note t. The special object of this 

 discussion is the examination of the changes of pressure from day to 

 day. 



It will be seen from the curves, p. 28, that in general, on whatever day the 

 barometer attains a maximum (or minimum) at one station, a maximum 

 (or minimum) occurs at the same time at the other stations. The means, 

 however, for the civil days only are not fitted to show the exact hours 

 of occurrence of the maximum or minimum : to determine these epochs 

 with the greatest possible accuracy, the following method has been 

 adopted. In order to eliminate the diurnal oscillation, the means of 

 24 hourly observations are always taken ; but these means have been 

 obtained for the 24 hours, having for their middle hour each hour of the 



* The annual vaviation obeys local laws ; considerable differences of barometric 

 height exist for months on two sides of the southern Ghats, where the conditions of 

 temperature and humidity are far from being the same. The diurnal variation in 

 India, especially the descent from the forenoon maximum to the afternoon minimum, 

 seems to depend, though to a less extent, on the same conditions. Mr. Buchan has 

 lately shown by the difference of tlie amplitude of this oscillation for the Tolcanic region 

 of South Italy and for the Spanish peninsula, that other conditions may exist (" On the 

 Diurnal Oscillations of the Barometer," Trans. Eoy. Soc. Edin. vol. xxvii. p. 406). 



t It may be remarked here that the solar rotation appears to produce also, and 

 more especially in high latitudes, a IS-day period, that represented by the term 

 h sin 20 in the function of sines for the 26-day period. 



