360 ARRAN. GEOLOGY. SCHIST. 



finable; a considerable series of alternations, or else a 

 perfect gradation, taking place at the conterminous 

 boundary. It must also be remarked, as will hereafter 

 be more fully detailed, that the series of argillaceous 

 schist is formed of many different substances ; including 

 graywacke of various aspects, chlorite schist, micaceous 

 schist, and common argillaceous slate. In different 

 parts of the tract which is constituted by this series, 

 these substances also vary in their relative proportions ; 

 some of them being occasionally absent, while another 

 becomes predominant. Thus, where it is even most 

 regular in position, its limits can rarely, if ever, be 

 ascertained by mere observation ; but must be deduced 

 by a general comparison of bearings, and by the pro- 

 longation of some ascertained boundary to another 

 point where the limit may be accessible. But in Arran, 

 no such assistance can be obtained, as the linear position 

 of both the series is destroyed or perverted ; while the 

 parts, which from analogy must be judged to have 

 been originally conterminous, now occupy positions 

 wide of each other, or are obscured by the changes 

 of character they have undergone with change of 

 position, or else are inaccessible by reason of their 

 situations in the interior country. The granite will 

 however be found in a great measure to separate the 

 argillaceous series from the micaceous ; while those 

 parts where the two must be conjectured to meet, 

 are either invisible to the degree required for ascer- 

 taining their mutual boundary, or undergo a confused 

 transition at the probable place of junction, of such 

 a nature arid extent as to render that undefinable. 



Under this uncertainty, the map will show that the 

 limit in question must exist between two points, the 

 one of which lies between Loch Ransa and Catcol 

 bay, and the other on the southern declivities of 

 Goatfell and of Ben huish. In neither of these places 

 however can that limit be ascertained, and accord- 



