162 



all the main experimental points of the investigation ; and, upon the 

 whole, I can perceive nothing substantiated which is positively irre- 

 concileable with the principle of interference ; while the new modifi- 

 cations of the phienomena here presented, so far as general consi- 

 derations can be relied on, seem sufficiently conformable to the 

 undulatory theory ; but as to their more exact quantitative explana- 

 tion, no definitive opinion can be pronounced until certain analytical 

 investigations of great length and complexity shall have been gone 

 through ; by which alone theory can be brought into exact and 

 satisfactory comparison with experiment. 



2. " On the Lunar Atmospheric Tide at Singapore." By Cap- 

 tain C. M. Elliot, M.E., F.R.S. Received Dec. 18, 1851. 



The discussion of the barometric observations at St. Helena by 

 Colonel Sabine having clearly and decidedly shown the moon's in- 

 fluence on the atmosphere, the author determined to discuss in a 

 similar manner the barometric observations at Singapore. The re- 

 sults of this discussion are given in the present communication. 



In order that a comparison might be made between the results at 

 Singapore and at St. Helena, he copied to a considerable extent the 

 form of the different lunar tables drawn up by Colonel Sabine in his 

 paper published in the Philosophical Transactions. 



The observatory at Singapore was in latitude 1° 18' 32" N. and 

 longitude 103° 56' 30" E. of Greenwich. The cistern of the baro- 

 meter, one of Newman's, having a tube 0*532 inch in diameter, was 

 a few feet above high-water mark. The observations, during the 

 whole of 1841 and the early part of 1842 and that of 1843, were 

 made at every two hours ; during the remainder of the time, to the 

 close of 1845, at every hour. 



The diurnal variation of the barometer having been eliminated, by 

 deducting the mean monthly height at each hour, from the height 

 given by observation, the residual quantities were arranged in tables ; 

 and the observation corresponding the nearest in time to the moon's 

 superior culmination for each day being marked as hour of lunar 

 time, the whole were again rearranged in tables according to lunar 

 hours. The variation or range of the mean of the sums of the dif- 

 ferences thus arranged is exhibited in a table, in the last column of 

 which are given the means of all the hours for each period of six 

 months. In a second table are given the differences between these 

 mean results in the last column of the preceding table and the num- 

 bers corresponding to the several hours in the other columns. 



The means of the complete years of observation, 1841, 1844, 

 1845, are shown in a third table, in which are also given the means 

 of the first six months of 1842 and 1843, during w^hich two-hourly 

 observations w^ere made, and the means of the latter halves of these 

 years, during which the observations were made hourly. 



The means of the twenty-four months of the two-hourly observa- 

 tions, and of the thirty- six months of the hourly observations, are 

 given in Table IV. Finally, Table V. exhibits the results of the 

 observations of three years, so combined as to show the efi^ect on 



