104 MR BROUN ON THE RELATION OF THE VARIATIONS OF THE 
the morning maximum, and extends the other forward at all other hours of the 
day, making the greatest strides at the time of the afternoon maximum. During 
the second half of the year, it rests the previously advancing foot about an hour 
before the evening maximum, and brings the lagging foot forward at the other 
hours, but with the greatest rapidity at the time of the principal minimum. 
13. There is perhaps nothing more difficult in groping for the laws which regu- 
late certain phenomena than the separation of the effects due to different causes ; 
but it is quite obvious that, before we can arrive at any sound conclusion as to 
simple laws, this must be done. In the determination of the diurnal period all 
the observations at each hour for a calendar month or year are summed, and the 
means taken; in these summations are included several irregularities named dis- 
turbances ; if the disturbances occurred equally positive and negative at the same 
hour, or were equally distributed over the twenty-four hours, a large enough series 
of observations would serve to eliminate them ; neither of these suppositions seems 
to hold, and accordingly, certain hours in some months are more affected by dis- 
turbances than the same hour in other months, or than the other hours of the 
same month; the diurnal curve, therefore, is complex. There are other causes, 
as will be seen afterwards, which render it more so. 
14. In the attempt to determine whether the horizontal intensity varies with the 
moon’s declination, the days were numbered from the day of the moon’s greatest N. 
declination, counting that day 0 till it returned to the greatest N. declination again ; 
and, as 13 of the moon’s revolutions, with regard to node, are equivalent to 12 
lunations, and nearly toa year, the 15 revolutions, with regard to declination, were 
selected for summation; as, by this means, any effect due to varying phase, or to 
annual period, would be eliminated. The mean intensity for each of the 13 days on 
which the moon had its greatest N. declination were then summed together ; the 
means for the 15 days numbered 1, in which the moon was moving south, and so 
on. For the purpose of verifying the result thus obtained, similar summations 
of the observations for 1845 were made; in this case, however, only 12 revolutions. 
with respect to declination, were obtained, so that any effect of phase will not be 
perfectly eliminated. No attempt has in either case been made to eliminate dis- 
turbances. The results of these summations were projected, having previously 
eliminated the effect of secular change. (See Curves, No. 6, Plate IV.) The 
curve, from the observations of 1844, indicates a maximum about 2 days after 
the moon has attained its greatest S. declination, and a maximum about a day 
after it has attained its greatest N. declination—the maxima have nearly equal 
values, so also have the minima. The branches ascending to and descending 
from the period of greatest S. declination are greatest; so that the periods of mi- 
nima are nearer, the greatest N. declination being about 5 days before and after it. 
The curve deduced from the observations of 1845 shews the maxima nearly at 
