22 
THE BEV. STEPHEN J. PEEBT ON THE MAGNETIC 
Table XYII. 
Station. 
Observed Declination. 
Computed Declination. 
Error. 
Rheims 
16-618 
16-908 
-0-290 
Metz 
15-871 
16-116 
— 0-245 
Strasburg 
15-473 
15-435 
+ 0-038 
Issenheim 
15-687 
15-499 
+ 0-188 
Dole 
15-976 
15-974 
+ 0-002 
Dijon 
16-503 
16-184 
+ 0-319 
Avignon 
15 935 
15-652 
+ 0-283 
Marseilles 
15-579 
15-404 
+ 0-175 
Monaco 
14-411 
14-567 
-0-156 
Montpellier 
16-432 
15-958 
+ 0-474 
Grenoble 
15-695 
15-518 
+ 0-177 
N. D. de Myans 
15-067 
15-583 
-0-516 
Mongre 
16-826 
16-064 
+ 0-762 
St. Etienne 
14-793 
16-483 
-1-690 
Clermont 
16-342 
16-649 
— 0-307 
Moul ins 
16-368 
16-686 
-0-318 
Douay 
17-870 
17-445 
+ 0-425 
Boulogne 
18-105 
18-022 
+ 0-083 
Paris 
17-140 
17-449 
-0-309 
The largeness of these errors is mainly, I think, due to the unsteadiness of the Jones 
altazimuth, which had been substituted, on account of its lightness, in lieu of the Cooke 
transit-altazimuth used during the survey of the west. This unfortunately dimi- 
nishes greatly the value of the results, and makes them scarcely comparable with those 
obtained for the west of France. The probable error for a single station is found to be 
±0-35884, and for the mean ±0-08458. Omitting the two worst results, viz. those 
for Mongre and St. Etienne, we obtain 
D=3-4989, +=0-0083462 y= -0-0021920, r=0-00863, w=14° 42' 57"-2 ; 
with probable errors of ±0-21389 and ±0*05347. 
The results, if we may judge of them by the amount of the probable error, will be 
still more improved if, besides casting out the two worst results, we correct each indi- 
vidual observation for the disturbance occurring at the time in this magnetic element. 
The correction to be applied may be obtained from measurements of the Stonyhurst 
photographic curves, as explained in my former paper on the Survey of the west of 
France. The almost identical occurrence of these disturbances in neighbouring countries, 
with regard, at least, to the element under discussion, is now so well established as to 
render unnecessary any justification of the appliance of such a mode of correction; but, 
unfortunately for its present efficacy, no disturbance happened during any of the obser- 
vations that will enable me to smooth very considerably the observed inequalities. 
From the solution of the equations formed with the corrected observations we obtain 
D = 3-4757, +=0-0082848, y= -0-0021781, '>-=0-00857, u=U° 43' 48"-6, 
with probable errors of ±0-19745 and ±0 - 04936. 
W e will now pass on to the consideration of the secular variation of the Declination. 
