( 9 ) 

 discussion as that employed by General sabine and Captain 



ELLIOT. 



Th e hours of observation, for which the differences b — b\ are 

 found, do not coincide, in most cases, with the lunar hours, 

 but partly precede and partly follow them. The distance be- 

 tween the liours of observation and the lunar hours sometimes 

 amounts to half an hour. The effects of tliis want of syn • 

 clironism are coinpensated at the lunar hours where the in- 

 fluence of the moon increases as much to the following hour 

 as it has decreased from the preceding, and at the lunar hours 

 where the inverse takes place. But this is neither the case 

 at the hours of the maxima and the minima, nor at the hours 

 preceding and following them. This has the effect that the 

 range of the Lunar Atmospheric Tide is found less than it 

 really is, and that the phenomenon can never be known in its 

 true state, for at least at 12 of the 24 lunar hours. General 

 sabine made notice of this defect of his method in his paper 

 on the Lunar Atmospheric Tide at St. Helena. 



Por this reason I have employed an other method of dis- 

 cussion than General sabine's. 



The mean times of the superior culminations of the moon 

 at Batavia were calculated, and from these mean times the mean 

 times at the commencement of the several lunar hours of each 

 lunar day were deduced. Then the barometric pressures at the 

 several lunar hours of each lunar day were calculated, by in- 

 terpolation between the observed values at the solar hours. 

 The observations at the solar hours had been corrected before- 

 hand for the solar diurnal variation ; this was necessary in 

 order to be able to examine the variation of the Lunar At- 

 mospheric Tide at different periods of the synodical revolution 

 of the moon. 



The lunar days were arranged in periods of 28 or 29 days, 

 beginning with the lunar days of New Moon, except when 

 these days were partly Sundays, in which case the prece- 

 ding or the following day was taken as the first of the 

 period. These periods were united by 12 or 13 in epochs, 

 each nearly corresponding with a year of observation, The 

 means of the barometric pressures at the several lunar 



