Harmonic Analysis of Tidal Observations. 563 



Major Baird has now caused to be sent the mean sea levels with 

 reference to the zeros of the several tide-gauges. The reference of 

 the zero of any gauge to a bench-mark ashore has principally a local 

 interest. Full statements on this head are given in the prefaces to 

 the Indian Tide Tables, but these are not reproduced. 



The table of mean sea levels which follows immediately comprises 

 all the stations in which more than a single year of observation has 

 been reduced. The day of the month, prefixed to each series of 

 results, denotes the first day of the year for which the mean sea level 

 is given. 



In the Fourth Report to the British Association on ' Harmonic 

 Analysis' (1886), it is shown that the oscillations of mean sea level are 

 far too large to be explained by the known astronomical inequality 

 with a period of nearly nineteen years. 



This is not a convenient occasion for the discussion of the present 

 series of values, but I remark that 1882 was a year in which the 

 whole Indian Ocean stood low, whilst 1885 was one in which it stood 

 high. 



If variation in the Sun's temperature is the cause of variation of 

 sea level, we might expect to find a periodicity with a period of ten or 

 eleven years. It is then worth noticing that at Karachi there is a 

 minimum in 1872 and again in 1882.* The observations are clearly 

 insufficient to do more than to raise the question. 



[Captain Wharton has been good enough to give me Mr. Russell's 

 results for mean sea level at Sydney, and it is interesting to note the 

 very large oscillation of level, with a minimum simultaneous with 

 that at Karachi. ]t 



* Sporer gives 1878'8 as the time of minimum sun-spots, 

 t May 8, 1889. 



