78 



entry in the Table (exclusive of course of the column which exhibits 

 the mean of the five preceding columns) is derived from a wholly 

 independent body of observations which belong to itself alone, and 

 are not employed in the deduction of any of the other entries. 



TABLE II. Lunar-diurnal Variation at Hobarton in the several 

 years from October 1843 to September 1848 ; omitting dis- 

 turbed observations differing 2'' 13 from their final normals. 



It may operate as an encouragement to those who have not yet 

 subjected their observations to any process of examination or analysis, 

 to perceive, by this example, how substantially satisfactory are the 

 results which may be obtained from even a single year of hourly ob- 

 servations, after the larger disturbances and the solar-diurnal variation 

 have been eliminated. 



I have spoken in a recent paper of an unexceptionable test by 

 which we may satisfy ourselves as to the confidence which may be 

 reposed in a series of observations, whether obtained by the eye or 

 tabulated from instrumental traces. Such a test is furnished when 

 the entries at solar hours are rewritten according to the lunar hours 

 to which they most nearly approximate, and when consequently their 



