66 
MR. A. W. RUCKER AND DR. T. E. THORPE ON A MAGNETIC 
The following are the best stations for the application of a similar test to Scotland. 
The two sets of observations at Scarnish and Portree were taken on two different 
days. Those at Stornoway were on the same day, but at different stations. The observa¬ 
tions at Bunnahabhain and Stranraer, which have been already discussed, are here 
arranged in order of time without reference to the station or instrument :— 
O 
Station. 
A. 
Corrected 
Declination. 
i 
Scarnisli. 
+ 2-4 
' 
O 
24 
49-8 
- 3-6 
+ 3-0 
24 
51-8 
Portree (2) . 
+ 3-3 
+ 2-0 
22 
21-6 
+ 3-7 
- 2-0 
22 
227 
Stornoway, Ard Point 
+ 2-0 
- 2-0 
23 
507 
(1) 
+ 6-3 
- 1-0 
23 
48-4 
+ 4-4 
23 
507 
Ard Point (2) . 
+ 2-4 
- 3-0 
24 
87 
+ 6-4 
- PO 
24 
7-9 
+ 4-4 
24 
8-3 
Bunnahabhain . 
- OT 
22 
46'0 
+ 1-7 
22 
44-0 
+ 4-3 
- 2-0 
22 
46'0 
+ 6-3 
- 1-0 
22 
47-3 
+ 5-7 
22 
47-9 
+ 4-6 
22 
47-3 
Stranraer . 
+ P9 
- 2-0 
21 
12-5 
+ 4-6 
- 3'0 
21 
137 
+ 6-4 
- 1-0 
21 
13-0 
In some cases where there is an appearance of a regular change, such as would be 
caused if the diurnal variation was not properly estimated, it can be otherwise 
accounted for. Thus the first four observations at Bunnahabhain wmre independent 
in every particular, but the last two depend upon the same Sun observation as the 
fourth. .This pi’obably accounts for the somewhat closer agreement in these cases. 
The magnitude of the errors which would be introduced if the disturbance corrections 
were omitted can be estimated by noting that at Cavan the first and last observations 
when uncorrected give 22° 27'‘6 and 22° 22''4, those at Portree 22° 23''6 and 22° 20''7, 
and those at Scarnish 24° 49'‘8 and 24° 54''8 respectively, instead of the much more 
accordant numbers entered in the Table. 
These results when corrected are quite as good as those we obtained when observing 
at Kew, and furnish a complete a ■posteriori justification of the method of correction. 
The values of the diurnal variation employed for the year 1884-5 were furnished 
to us by Mr. Whipple. For 1885-6 and 1886-7 we used the Tables given in the 
reports of the Kew Committee for those years (‘Roy. Soc. Proc.,’ vol. 41, p. 415, and 
vol. 43, p. 226). 
These differed rather markedly as to the magnitude of the maximum variation from 
the curve employed for 1884-5, the difference amounting to about 1''5. As the report 
