86 
MR. A. W. RtrCKER AND DR. T, E. THORPE ON A MAGNETIC 
The distribution of these stations affords, at all events, 'prima facie evidence that 
the differences between them are not accidental. That they cannot be due to errors 
of measurement is evident when we remember that an error of O'd in the annual rate 
of change determined from experiments separated by an interval of fifty years would 
require an error of 5' in one or other of the observations, or of 2'’5 in both. We 
therefore decided to regard the figures in fig. 1 as giving the true rate of secular 
change of Dip between 1857 and 1887 at the stations indicated. 
The determination of tlie mean rate of decrease of Dip between the years 1857 and 
1887 does not, however, give, the value of that quantity during the period of our 
survey. As the rate is diminishing it would be less in recent than in earlier years. 
The evidence we are able to collect on this head is confined to data from Greenwich, 
Kew, and Stonyhurst, and to a few stations at which we observed both near the 
beginning and near the end of our survey. In the following Table we have entered 
such facts as were known to us at the time when the progress of the reductions made 
it necessary to come to a decision on this point. 
Our own observations and those taken at Stonyhurst are corrected for diurnal 
variation and disturbance. The Table illustrates the importance of these corrections 
wlien conclusions are to be dra.wn from observations between which the interval of 
time is small. 
Table III. 
Statiou. 
Date. 
Dip. 
Date. 
Dip. 
Interval 
in years. 
Secular 
rate of 
change. 
Assumed 
rate. 
Loch Aylort . 
Sept., 1884 
O 
71 
24-0 
Aug., 1888 
0 
71 
22-3 
4 
- 0-4 
- To 
Kerrera . 
Aug., 1884 
70 
51T 
Aug., 1885 
70 
48-7 
1 
- 2-4 
- 1-2 
Strauraer . 
Sept., 1884 
70 
141 
Aug., 1888 
70 
111 
4 
- 0-8 
- 1-4 
Stonyhurst 
1884 
69 
16-2 
1887 
69 
9-0 
3 
- 2-4 
- 1-6 
Kew .... 
1884 
67 
39-5 
1888 
67 
36-4 
4 
- 0-8 
- 1-5 
Greenwich . 
1884 
67 
30-2 
1886 
67 
27-0 
2 
- 1-6 
- 1-5 
Leeds 
Sept., 1886 
69 
9-7 
Dec., 1888 
69 
6-9 
2-2 
- 1-3 
- 1-5 
Mean . 
- 1-4 
- 1-4 
Note. —Stornoway is omitted because tbe observations were made in 1884 in the Castle grounds and 
in 1888 on Ard Point. 
These figures illustrate the fact that there may be a considerable discrepancy at 
any particular station between the actual secular change at a given time and that 
deduced from the mean taken over a long series of years. On the wdiole, we think it 
probable that the present rate of change of Dip is somewhat less than its mean value 
between 1857-87, and we have, therefore, reduced the numbers for that period by 
one tenth, neglecting fractions of a tenth of a minute. This assumed rate gives a 
correct mean rate for the stations referred to above in Table III., though it is con¬ 
siderably too high for some and too low for others. It is probable that in the case of 
Kerrera the observed rate of change is not quite correct. The site is subject to local 
