Vol. 52.] EOCKS OF THE LIZABD DISTEICI. 19 



false or current-bedding. 1 This, as stated in our paper of 1891, 

 was difficult to explain on any other hypothesis than that of an 

 original stratification, which also accorded very well with the aspect 

 of the banded rock. Doubts, however, as mentioned in the paper 

 on Sark by Mr. Hill and myself, 2 had arisen as to the validity of these 

 difficulties, and proof had been obtained that distinct mineral 

 banding might occur in a true igneous rock of like composition, 3 

 so that the whole question was investigated de novo. Three locali- 

 ties appeared to supply evidence of special importance, which I shall 

 endeavour to describe. 



(a) Inclusions in the Hornblende-schist at Porthoustock 



Cove. 



The hornblende-schist on the northern side of this cove is some- 

 times very dark in colour, sometimes rather conspicuously streaked 

 or banded with felspar : the more homogeneous and the more banded 

 rock alternating twice or thrice, in a vertical space of about as 

 many yards. Sometimes fairly well-formed crystals of white felspar 

 occur porphyritically in the dark rock, up to about § inch loug^ 

 sometimes the rock becomes coarser and recalls the structure of a 

 hornblendic gabbro. Now and then 4 the latter rock forms a kind 

 of 'eye' in the normal banded schists and presents a structure 

 resembling current-bedding (fig. 1). But rather towards the western 



Pig. 1. — One of four 'eyes,' consisting of a coarse, gabbro-like 

 material in the normal hornblende-schist. {North of' Porthoustock 

 Cove.) 



end of the Cove we found an inclusion of nearly pure, rather 

 coarsely crystallized hornblende. It is about 6 or 7 inches thick in 

 the part drawn (fig. 2, p. 20): the ordinary hornblende-schist 

 bending round it, as indicated, but it ' trails out ' to the right and. 

 interruptedly (there are two layers), to the left. In this direction, at 

 about the same level, are two or three similar but smaller inclusions. 

 A slice from the largest inclusion 5 shows it to consist almost entirely 



1 Quart. Journ. Geol. Soc. vol. xxxix. (1883) p. 4 ; see pi. i. fig. 2. 



2 Ibid. vol. xlviii. (1892) p. 122. 



3 Op. cit. pp. 134-138, cf. p. 124. 



- 4 I saw about four instances — varieties of linear streaking were countless. 



5 The specific gravity of the specimen, as determined with a Walker's balance, 

 is 3'124 ; the mean of two specimens of fairly normal hornblende-schist is 2'748. 

 I am indebted to Miss E-aisin for these and other determinations in this paper. 

 T ensure accuracy, each of us read the balance. 



c2 



