]\gt ; , where represents the equivalent angle ( at per hour ) to the time elapsed since Greenwich midnight , the component of the horizontal difference vector inclined at any angle to the magnetic meridian ( or direction of ) is 2 nt Here is supposed to be expressed in terms of as unit , which means , for the place and epoch considered , writing as the equivalent of 1 ' .
The mean declination at Kew for the eleven years 1890-1900 was W. , so that , to get the component called above , we must write for , and similarly for X. The results finally obtained for X , , and the vertical component of the vector , in terms of as unit , were as follows:\mdash ; , ( 1 ) , ( 2 ) .
( 3 ) The fact that is insignificant was , of course , implied by fig. 2 , but it is interesting to find that the 24-hour term in it is especially insignificant , while the 24-hour term in X dominates all the others .
The 24-hour term in is also largely dominant , so that the difference vector is essentially a phenomenon of a regular kind with a 24-hour period .
between the angles and those deduced rom the complete period .
The only oonspicuous difference between the sunspot maximum and minimum results il in the amplitude of the difference vector .
The excess for the sunspot mmnm group is in much the same proportion as in the case of the quiet day vector itself . .
The N. quiet day inequalities were given in my previous paper Uy for March .
, and December , in addition to the year as a whole .

