SOLAR-DIURNAL VARIATION OF THE DECLINATION. 513 



which it has not been eliminated, and in still higher 

 latitudes (more particularly when we approach the 

 localities from whence the disturbances appear to ema- 

 nate), so great as to mask altogether the diurnal vari- 

 ation proceeding from other sources. 



It has been stated that the principal deflections 

 caused by the diurnal variation in the course of the 

 twenty-four hours, are those which take place during 

 the hours of the day. We have now to consider the 

 minor phenomena that occur during the hours of 

 the night. The progressive march of the declination- 

 magnet from the extreme elongation which it reaches in 

 the early hours of the afternoon to its opposite extreme 

 reached in the forenoon hours of the following day, (a 

 movement of the north end of the magnet from west to 

 east in the extra-tropical parts of the northern magnetic 

 hemisphere, and from east to west in the same parts of 

 the southern magnetic hemisphere,) is found to be 

 interrupted in both hemispheres by a minor retrogres- 

 sion (of variable duration); after which the previous 

 direction is resumed. The phenomenon of this " noc- 

 turnal episode," as it has been not inappositely called, 

 appears to have been first observed and noticed by 

 M. de Humboldt himself, in conjunction with M. Gray 

 Lussac, at Eome, in 1805 (p. 127); but although a very 

 important feature, it seems to have been unaccountably 

 lost sight of, for it is wholly unnoticed byM. Arago in 1836 

 (Annuaire, p. 283), when stating in detail the diurnal 

 march of the declination in the northern and southern 

 hemispheres, On the reception at Woolwich of the 

 earliest observations of the British colonial observato- 

 ries, it was at once perceived, from the examination of 



IV. L L 



