TRANSACTIONS 
Berwick-on-T weed (W.). | 
Hawick. 
Dumfries (W.). 
Stranraer (W.). 
Ayr (W.). 
Carstairs (W.). 
Edinburgh (W.). 
Glaszow (W.). 
Fairlie. 
OF SECTION A. 
Crianlarich. 
Oban (W.). 
Kerrara. 
Tona. 
Coll. 
Tiree. 
Loch Aylort. 
Bannavie (W.). 
Dalwhinnie (W.). 
Soa. 
Canna. 
Loch Boisdale. 
Loch Maddy. 
Stornoway (W.). 
Callernish (W.). 
Gairloch. 
Loch Inver (W.). 
Loch Eriboil. 
927 
Campbelltown (W.). Ballater. Thurso (W.). 
Port Askeg (Islay) (W.). Aberdeen (W.). Wick (W.). 
Tarbert (Loch Fyne). Banff (W.). Golspie (W.). 
Strachur. Elgin (W.). Lairg. 
Stirling. Boat of Garten. Stromness (W.). 
Dundee. Inverness (W.). Kirkwall (W.). 
Pitlochrie (W.). Fort Augustus (W.). Lerwick (W.). 
Crieff. Kyle Alin (W.). 
W. signifies a station also observed at by Mr. Welsh in 1857 or 1858. The 
stations are distributed as uniformly as possible over the whole area of Scotland, 
and they have been chosen with the view of repeating the measurements at as 
many as possible of the positions selected by Mr. Welsh, and, at the same time, of 
avoiding as far as might be regions of great local disturbance. At stations where 
Mr. Welsh’s notes gave the requisite details we have observed when possible on 
the exact site he occupied. At new stations we have been careful to select positions 
which are likely to remain open, so that future observations can be made there. 
Some of the declinations determined by Mr. Welsh on the west coast were dis- 
carded when his results were published by Professor Stewart, as the mirror of 
the magnetometer was found to have been out of adjustment on several occasions. 
Owing to the improvements made in the Kew Unifilar since the time of the last 
survey, the liability to error from this source is practically obviated: it is now 
possible to ascertain readily whether the mirror is properly set and, if necessary, to 
readjust it, before making the solar azimuth observations. With the view, how- 
over, of still further minimising the errors due to the axis of the mirror being either 
not parallel to its plane, or not perpendicular to the line of collimation of the tele- 
scope, our observations of the transits have been made by alternate reversals of the 
axles in its bearing and by back and front observations of the sun. 
The accuracy of thedeterminations of solar azimuth, and hence of the declination, 
has been greatly increased by the circumstance that we were able to make frequent 
comparisons of our chronometers with the time signals sent daily from Greenwich 
at 10 a.m. and | p.m. along the telegraphic system of England and Scotland. We 
have to thank Mr. Cunynghame, the Director-General of Telegraphs at Edinburgh, 
and Messrs. Tansley and Redford, Superintendents of the Glasgow and Ayr dis- 
tricts respectively, for the facilities afforded us in this part of our work. We are 
also much indebted to the postmasters at the various places visited for the readiness 
with which they arranged for the transmission or reception of the signals. 
Our Magnetometer (No. 60 Elliott) and Dip Circle (No. 74 Dover) were com- 
pared with the Kew instruments by observations taken by ourselves and by Mr. 
T.-W, Baker at the Kew Observatory. Our thanks are due to the Kew Committee 
of the Royal Society and to Mr. G. W. Whipple for the assistance thus rendered. 
The actual work of the survey was prefaced by an inquiry as to the nature and 
extent of the influence exerted by the great centres of local disturbance at various 
parts along the west coast of Scotland. With a view of obtaining information on 
this matter we selected the island of Mull, the geological characters of which have 
been very fully described by Professor Judd, and which, as Mr. Welsh’s determina- 
tions show, exerts a highly disturbing influence. The observations were made in 
the summer of 1883, and are published in the ‘ Proceedings of the Royal Society’ 
for 1883-4, They serve to indicate within what limits such centres of local 
disturbance may be approached without affecting the general direction of the 
5 
