354 
THE REV. STEPHEN J. PERRY ON THE MAGNETIC 
The figures in column 3, which are obtained directly from the unifilar, furnish the 
first members of the equations, whose solution is 
x = 0-0119696, 
y =- 0-0012255, 
D= 17°-3881. 
Thus the position of the isogonics will be N. 5° 50' 46"-2 E. to S. 5° 50' 46"‘2 W., 
and the distance between lines, differing by 30', 41-56 miles. 
If, on the other hand, we use the corrected values of column 4 in the formation of our 
equations, the solution becomes 
x = 0-0115492, 
y = -0-0022202, 
D= 17°-3368; 
and then the isogonics lie at an angle of 10 c 52' 54"- 7 to the astronomical meridian, 
and their distance is 42-51 miles. The corrections have therefore altered very consider- 
ably the direction of the isogonics without affecting much their distance apart. We 
may judge from the following Table how far these corrections have diminished the pro- 
bable error in the results. 
Table XYI. 
Station. 
Uncorrected Declination. 
Error. 
Corrected Declination. 
Error. 
Observed. 
Computed. 
Observed. 
Computed. 
Aix-la-Chapelle 
0544 
0-483 
+0061 
0-508 
0-457 
+0051 
Alost 
1-420 
1-567 
-0147 
1-384 
1-516 
-0132 
Antwerp 
1-509 
1-402 
+0-107 
1-540 
1-377 
+0-163 
Arlon. 
0-530 
0-528 
+0-002 
0-440 
0-424 
+0016 
Bruges 
2-021 
2014 
+0-007 
1-975 
1-967 
+0-008 
Courtray 
1912 
1-954 
-0042 
1-795 
1-880 
-0-0S5 
Ghent 
1-983 
1-740 
+0-243 
1-862 
1-691 
+0-171 
Li6ge 
0-277 
0-732 
-0-445 
0-277 
0-690 
-0-413 
Louvain 
0-797 
1-220 
-0-423 
0-873 
1-178 
-0-305 
Mons 
1-373 
1-576 
-0-203 
1-256 
1-490 
-0-234 
Namur 
1-677 
1-100 
+0-577 
1-582 
1032 
+0-550 
j Ostend 
2121 
2-173 
-0-052 
2135 
2-120 
+0 015 
j Spa 
0-733 
0-577 
+0-156 
0-670 
0-528 
+0142 
! Tournay 
* Tronchiennes 
1-835 
1-882 
-0 047 
1-731 
1-796 
-0-065 
1-932 
1-773 
+0-159 
1-903 
1-723 
+0-180 
Turnhout 
1-205 
1-144 
+0-061 
1-072 
1-135 
-0-063 
The probable errors at any station will therefore be +0-16585 for the uncorrected 
and +0-15234 for the corrected values, whilst those of the mean are respectively 
+ 0-04146 and +0-03809. 
The direction of the isogonics, when the corrected numbers are employed, approaches 
nearer to that found for France than when the uncorrected observations are used, but 
it is still considerably in defect. There is scarcely any difference between the closeness 
of the lines in Belgium and in the West of France. 
For the calculation of the secular variation of the Declination, the only available 
observations made out of Brussels are those of Lamont in 1858, since the two observa- 
tions of last century mentioned at the beginning of this paper would not help us to 
