| 
| 
| 
243 
nutes; thus leaving a total decrease, from year to year, of 
five minutes. 
When we compare the course of these changes with those 
observed by Cassini, we find that the directions of the move- 
ments are precisely opposite,—as, indeed, we might have been 
led a priori to expect from the opposition in the directions of 
the secular changes ; and we are led to generalize the law as 
follows : 
‘‘ From the beginning of the year to the beginning or 
middle of April, the mean declination undergoes little or no 
change. Thence, to the beginning of August, its movement 
is retrograde (or opposite in direction to the secular change) ; 
and from the beginning of August to the end of the year, it 
is direct.” 
The phenomena just described are, it is manifest, the 
resultants of two distinct changes, namely, the annual varia- 
tion properly so called, and the progressive or secular change. 
If the latter be subtracted, at the rate of 5./0 per annum, or 
0./4 per month, the remaining numbers give the true annual 
period. When thus considered, the annual variation exhibits 
an increase of the declination during the first seven months of 
the year, and a decrease during the remaining five months ; 
the apparently stationary condition of the declination, during 
the first three months, arising from the mutual compensation 
of the periodical and the progressive changes. 
It appears, then, that the annual variation (unlike the diur- 
nal in this respect) is a single oscillation. The minimum 
occurs near the end of January, and the maximum in the 
beginning of August ; and the whole range of the change is 
6.6 minutes. 
The laws of the annual variations of the declination and 
of the temperature, present the most complete accordance 
in the epochs of maxima and minima, as well as those of the 
mean values. The maximum of temperature occurs about 
August 1,—that of declination about August 8. The mini- 
mum of temperature takes place about January 15,—the mini- 
