690 J. W. JUDD 01T THE SECOISTDAKY K0CKS OP SCOTLAND. 



and the farmhouse of Udrigle, where the. upward succession is 

 abruptly terminated by the western fault. After making every pos- 

 sible reduction, and checking our estimates of the breadth of outcrop 

 of the several members by means of the large and accurate Admi- 

 ralty Chart, I feel justified in stating the following to be the succes- 

 sion and approximate thicknesses of the strata at this locality. The 

 section given proceeds from the base upwards. 



(1) Coarse breccia-conglomerates of subangular fragments of sand- 

 stone, quartzite, limestone, &c, alternating with red and mottled, 

 false-bedded, often argillaceous sandstones, with a few pebbles scat- 

 tered through their midst, the whole being traversed by many veins 

 of white calc-spar ; where in contact with the underlying Torridon 

 Sandstones the larger fragments are all evidently derived from that 

 formation, and are sometimes almost perfectly angular and of con- 

 siderable dimensions. In the lower part of this series the breccia- 

 conglomerates certainly preponderate over the sandstones ; but in the 

 higher parts the former are found gradually to become less and less 

 prominent, and towards the summit only occasional bands of conglo- 

 merate occur in the midst of thick masses of argillaceous strata. 

 This series is certainly of very great thickness, the lowest possible 

 estimate being 500 feet. 



(2) The argillaceous sandstones of the lowest member of the series 

 graduate upwards almost insensibly into thick masses of red and 

 variegated clays and marls, which in places are very sandy in cha- 

 racter. These very strikingly resemble portious of the English 

 Keuper, and only occasionally contain thin sandstone bands, which 

 are sometimes conglomeratic. Some beds of sandstone of a sea- 

 green colour, and irregular bands of impure concretionary limestone, 

 also occur in this part of the section. In its upper part this member 

 of the Poikilitic exhibits a preponderance of arenaceous over argil- 

 laceous strata, and conglomerates become more common ; its thick- 

 ness is certainly over 200 feet. 



(3) Soft reddish argillaceous and sandy beds in alternating succes- 

 sion, with some bands of white and greenish sandstone and grit, 

 which, on account of their hardness and power of resisting atmo- 

 spheric denudation, often stand out very prominently in the clifT- 

 section. Conglomeratic beds are exceedingly rare in this part of the 

 series, which certainly exceeds 200 feet in thickness up to the point 

 M^here the upward succession is abruptly terminated by the great 

 fault. 



These three divisions of the Poikilitic, which, however, are not in 

 any way sharply divided from one another, as seen at Gruinard Bay, 

 have each their analogues, which seem to follow one another in 

 somewhat corresponding order, at several other localities where the 

 same strata are exposed in the Western Highlands. 



The next point at which, in proceeding southwards from 

 Gruinard Bay, we detect traces of the Poikilitic beds, is in the 

 island of Raasay. Here there occur two very interesting sections. 

 At one of these, situated at the point called Ku-na-Leac, the highest 

 beds of the formation are seen unconformably underlying all that 



