ON COMPARISON AND REDUCTION OF MAGNETIC OBSERVATIONS. 217 



for the mean diurnal inequalities of winter, summer, and the whole 

 year. 



The a, h coefficients for the 'year' are the arithmetic means of those 

 for ' winter ' and ' summer ' ; but as a check on the accuracy of the work 

 they were calculated independently. 



Variation to the west in the declination is regarded as positive. 



Table V. 



The only change required in Table V. if time were measured from 

 noon, instead of as throughout the present paper from midnight, would 

 be the alteration of the signs of the entries referring to «,, 6,, a^, and 63. 



It should be explained that the calculations on which Table V. is based 

 proceeded in every case to at least one decimal place, usually two, beyond 

 that shown. 



The table shows that in the case both of the declination and the 

 horizontal force, the terms whose period is twenty-four hours are very 

 much greater in summer than in winter, and the same phenomenon shows 

 itself in the terms in the declination whose period is twelve hours. On 

 the other hand the terms in the declination whose period is six hours 

 appear relatively much larger in winter than in summer. It would hardly 

 be prudent to attach too much weight to this last conclusion, in view of 

 the extreme smallness of the quantities involved ; but it is in accordance 

 with Tables XV. and XVI. of the ' Greenwich Magnetical and Meteoro- 

 logical Observations' for the two years 1890 and 1891. A comparison of 

 the other features common to these two Greenwich tables and to Table V. 

 ■will be found of interest. The Greenwich tables, it should be noticed, are 

 not confined to ' quiet ' days, and in treating the horizontal force they 

 take as unit the (1/100000) of this force at Greenwich, or, say, lO"* x 182 

 C.G.S. unit. 



§ 10. The number of degrees in the angles 27r?ir„/24 is easily obtained 

 from Table V. by means of the formula (3). Instead of giving these 

 angles explicitly I have preferred to give the times t„, answering to the 

 earliest maxima in the day of the Westerly declination and the hori- 

 zontal force. These times are included in Table VI., along with the 

 earliest times in the day when the several terms of the types appearing 

 in (1^) vanish and have their minima. The interval between successive 

 maxima or successive minima is double that between successive zero- 

 points, and equals the periodic time, i.e., is 24, 12, 8, or 6 hours, as the case 

 may be. 



With one impoi-tant exception — that of the harmonic term in the 

 declination whose period is twenty-four hours — the first maximum of 

 every term appears earlier in the day in summer than in winter. The 

 difierence between the corresponding times in winter and summer is 



