Geological Society. 143 



British Isles.' His researches have embraeed the Western Islands of 

 Scotland, St. Kilda, and the Faroe Islands. 



1. In an account of the rocks of the basalt-plateaux attention 

 s particularly directed in this paper to a type of banded basic lavas 

 which play an important part in the structure of the volcanic dis- 

 tricts both of the Inner Hebrides and of the Faroes. The banding 

 sometimes consists in layers of more highly vesicular structure, 

 sometimes in alternations of distinctly different lithological cha- 

 racter, such as close-grained basalt and more coarsely-crystalline 

 dolerite. These banded rocks are more particularly developed in 

 the lower portions of the volcanic series. At a distance they might 

 be mistaken for tuffs or other stratified deposits. They occupy a 

 conspicuous place in the great precipices of the west and north of 

 the Faroe Islands. 



Numerous examples are cited of the ending-off of basalt-sheets in 

 different directions, indicative of many local vents from which 

 the lavas issued. An account is also given of tuffs and other 

 stratified intercalations which occupy a subordinate place in the 

 structure of the plateaux. 



2. A number of examples are adduced of the volcanic vents which 

 form a characteristic feature of the basalt-plateaux. A remarkable 

 row of five such vents was met with by the author at the base of 

 the great cliffs on the west side of Stromo, in the Faroe Islands. 

 They are occupied with agglomerate, and their saucer- shaped craters 

 have been filled in by successive streams of lava from neighbouring 

 vents, the whole being buried under the great pile of basalt-sheets 

 forming the island of Stromo. 



An instance of similar structure is described from Portree Bay, 

 the agglomerate in this case being connected with a thick and wide- 

 spread sheet of tuff intercalated among the basalts. Another 

 example is cited from the eastern end of the island of Canna, 

 where the ejected volcanic blocks are associated with records of 

 contemporaneous river-action. 



3. The paper describes in some detail the evidence for the flow 

 of a large river across the lava-fields during the time when volcanic 

 activity was still vigorous. Thick sheets of well-rolled grave! are 

 intercalated among the basalts of the islands of Canna and Sanday. 

 These masses of detritus consist mainly of volcanic material, but 

 they include also abundant pieces of Torridon Sandstone and rocks 

 from the Western Highlands. The current of water which. trans- 

 ported them certainly came from the east. That it flowed while the 

 volcanic vents continued in eruption is shown by the bands of tuff 

 and the large blocks of slag contained in the conglomerates, as well 

 as by sheets of vesicular basalt interstratified in the same deposits. 

 -From the terrestrial vegetation whereof the macerated remains are 



enclosed in the tuffs and shales, and from the entire absence of 

 marine organisms, it maybe confidently inferred that the water was 

 that of a river. The large size of many of the rounded blocks that 



