many Point* in the Frith of 'Clyde. 251 



The difference of the extremes of latitude vary from 27" to 

 & 27", and those of longitude vary from 12" to 3'. 



The following are a few of the positions of Ailsa Craig : 



None, 1835, Lat. 55 17' 0" N. Long. 5 8 0" W. 



Lynn, 1825, 55 1CJ 500 



Mackay, 1804, 55 15 550 



True, 1836, 55 15 13 5 7 



Here the differences of latitude vary from about 1' to 2', and 

 the difference of longitude from 1' to 7'. It is clear, too, that 

 the true distance of Pladda Light, due north from Ailsa Craig, 

 is 10' 20", while some of these positions give 15' N. and 3' of 

 longitude E., &c. From these few comparisons, therefore, it is 

 manifest, that, in thick weather during the day, vessels making for 

 the Clyde are in no little danger, especially towards the evening, 

 in the winter season, when the days are short, and just before 

 the light-houses are lit up ; because ships may be close in shore 

 while they expect to be five or six miles from it. Before they 

 discover their mistake by the light- houses, especially in blowing 

 weather, it may be too late often to avoid the danger. New 

 surveys of the coasts are therefore more necessary than additional 

 light- houses. 



The large map of Scotland by Arrowsmith, from Dr MacCul- 

 loch's observations, coloured to shew the geological character of 

 Arran, along with the rest of the country, and sold at L. 5, 5s. 

 partakes of the same errors here, and is therefore totally unfit 

 to trace the connexion of the geological features of Arran with 

 the opposite coast of Ayrshire, Bute, and Cantire. In the re- 

 cently published map of the basin of the Clyde by Mr Knox, 

 the same erroneous positions in the Island of Arran are obvious- 

 ly followed, while a dotted line, drawn directly south from about 

 Kilmory, instead of from Pladda Light, indicates the course and 

 distance Hi' to Ailsa Craig. In some of the recently published 

 maps, several of the more glaring errors are corrected, though 

 much is yet required to make them perfect. 



During the short time, previously alluded to, which I spent 

 in Arran> I made a few observations with a six inch astronomi- 

 cal circle of Kater^s construction, made by Robinson, with some 

 improvements suggested by myself. It has three verniers, each 



