20 SOLAR-DIURNAL VARIATION 



In the above table + signifies westerly and — easterly deflection ; it may be 

 compared with similar tables constructed for Toronto, 1 Dublin, 2 and Prague. 3 It 

 will be observed that the preceding table, which gives the observed variation, refers 

 to an epoch l'JI minutes later than the exact local hour (that is, to an exact 

 Gottingen hour), whereas the computed table refers to the exact Philadelphia hours. 



From the computed tabular values, aided by the diagrams, we can now deduce 

 some of the general features of the diurnal variation and its annual inequality. 



The general character of the diurnal motion (see type-curves) is nearly the same 

 throughout the year; the most eastern deflection is reached a quarter before 8 

 o'clock in the morning (about a quarter of an hour earlier in summer, and half an 

 hour later in winter) ; near this hour the declination is a minimum ; the north 

 end of the magnet then begins to move westward, and reaches its western elonga- 

 tion about a quarter after one o'clock in the afternoon (a few minutes earlier in 

 summer). At this time the declination attains its maximum value. The diurnal 

 curve presents but a single wave, slightly interrupted by a deviation occurring 

 during the hours near midnight (from about 10 P. M. to 1 A. M.), when the 

 magnet has a direct or westerly motion; shortly after 1 A. M. the magnet again 

 assumes a retrograde motion, and completes the cycle by arriving at its eastern 

 elongation shortly before 8 o'clock in the morning. This nocturnal deflection is 

 well-marked in winter, vanishes in the summer months, and is hardly perceptible 

 in the annual curve. According to the investigations of General Sabine, it is 

 probable that, if we had the means of entirely obliterating the effect of disturb- 

 ances, this small oscillation would almost disappear. In summer, when it has no 

 existence, the magnet remains nearly stationary between the hours of 8 P. M. and 

 3 A. M., a feature which is also shown by the annual type-curve. 



The two preceding plates show a close general resemblance in the diurnal curves 

 for the six months when the sun has north declination, and a similar resemblance 

 in the other six months when it has south declination. 



The analytical expressions give the epoch and amount of variation with greater 

 precision. The hours of minimum and maximum deflection are obtained from the 



equation ( = o; and the hours of the mean declination, when the curves cross 



1 </ n 



the axis of abscissae, from the condition A d = o. The following table contains 



these results for each month and the two principal seasons of the year, also the 



critical interval between the two adjacent hours of the mean position'. 



1 Vol. III., Tabic LXVI; compare also with Table VII. of Vol. II. 



- Trans. Royal Irish Academy, Vol. XXII., Part I., Table III. 



■ Academy of Science at Vienna, Vol. VIII. of .Math. Section, Table II. 



