SUEVEY OF THE EAST OF FRANCE. 
11 
We thus find that the probable errors of any single observation, or rather of the mean 
value at any single station, = + 0+745 ^/- +013538, whilst that of the mean 
from all the observations =±0 030274. 
The large error at Marseilles will probably be due to the difficulty experienced in 
finding a convenient site for the observations. 
If, now, we turn to the series of observations taken at some of the above stations by 
Dr. Lamont, and reduced to the epoch of Jan. 1st, 1858, and if we consider the epoch 
Sept. 1st, 1869 as common to all stations of our Survey (which we are able to do without 
sensible error), we arrive at the following Table for determining the secular variation of 
the Dip in the east of France : — 
Table IV. 
Station. 
Dip, Jan. 1, 1858. 
Dip, 
Sept. 1, 1869. 
Diff. of 
Epoch. 
Diff. of Dip. 
Yearly rate 
of decrease. 
Dip, Jan. 1, 1869. 
Clermont 
64-202 
63-571 
Hf 
-0-631 
-6-054 
63-607 
Dijon 
64-917 
64-373 
>> 
~ 0-544 
— 0-047 
64-409 
Marseilles 
61-675 
60-540 
- 1-135 
-0-097 
60-576 
Montpellier ... 
62-255 
61-578 
-0-677 
-0-058 
61-614 
Moulins 
64-723 
64-045 
>> 
-0-678 
-0-058 
64-081 
Paris 
66-442 
65-823 
-0-619 
-0-053 
65-859 
Mean (omitting Marseilles) 
-0-054 
Comparing this mean annual change with — 0°+45, the rate for 1858 as deduced by 
Lamont, we find the decrease to be accelerated annually by — 0°+0082, which agrees 
closely with the acceleration for the period from 1780 to 1830, which General Sabine 
gives as — 0+0085. 
In our previous discussion of the series of observations taken in 1868 in the west of 
France, the deduced yearly rate of decrease in the Dip was found to be 0+62 ; the Dip 
would therefore seem to be decreasing rather more rapidly in the west than in the east 
of France. 
In the Table of the Dip observations it will be noticed that at a few stations the 
readings differ very considerably from each other; but I have retained them all in 
forming the equations of condition, as I cannot see a sufficient reason for discarding 
any, since the same attention as to choice of position and accuracy of observation was 
maintained throughout. When at any station the readings of two of the needles agree 
fairly together, but differ much from the third, this could scarcely be considered conclusive 
against the correctness of the third, unless all three had been observed under precisely 
similar circumstances of time and place ; since it is not impossible that an iron tube or 
other disturbing cause, of which we could obtain no information, had affected the two 
first needles and not the third. But to test the correctness of this view, I have solved 
the equations after omitting the most striking irregularities, viz. the three at Moulins, 
c 2 
