ANDESITES OF DEVONSHIEE. 497 



1865. W. Vicary, F.G.S. On the Peldspathic Traps of Devon- 

 shire. Trans. Devon. Assoc, vol. i. pt. iv. pp. 43-49 (gives additional 

 localities and fuller details of the macroscopical characters of the 

 rocks). 



1890. E. N. Worth, F.G.S. The Igneous Constituents of the 

 Triassic Breccias and Conglomerates of South Devon. Quart. Journ. 

 Geol. Soc. vol. xlvi. pp. 69-83. 



1892. W. A. E. Ussher, F.G.S. Permian in Devonshire. Geol. 

 Mag. pp. 247-250. 



The present paper is the result of a visit to the localities in July 

 1890. 



The most satisfactory accounts of the Devon ' felspathic traps ' are 

 contained in the above-cited report of De la Beche, in Mr. Yicary's 

 paper, and in Dr. Hatch's notes to Mr. Ussher's paper. As is well 

 known, they are a series of volcanic rocks, of post-Carboniferous 

 age, which are exposed at numerous localities in the south of the 

 county. All those microscopically described in the present paper, 

 except the rock of Killerton, N.E. of Exeter, are olivine-basalts. 



II. Evidence of the Co:jttempoeak:eous (n-ox-Intrusive) 

 Chaeactee of the Rocks. 



Sir Henry De la Beche held them to be " sections of erupted 

 igneous matter contemporaneously intermingled with the red sand- 

 stones and conglomerates with which they are associated." ^ Mr. 

 Yicary, on the contrary, states that they " commonly appear as 

 dikes, filling fissures in the earlier rocks."' The following considera- 

 tions lead me to take De la Beche's view rather than Yicary's. 



1. At Washfield the rock is, as Mr. Yicary himself states, 

 " vesicular throughout." I observed vesicles over 2 inches long, 

 and the whole rock is as full of cavities as a sponge. At " Beere, 

 Thorverton, Rew, and the neighbourhood of Silverton " the rock is, 

 as Mr. Yicary says, " more or less vesicular from top to bottom of 

 the present workings." In almost every other case the top or 

 bottom of the lava or both are vesicular. 



2. In some cases, as at Pocombe, and, according to De la Beche,^ 

 between Ide and Dunchideock, the basalt intervenes between the 

 underlying Carboniferous and the overlying Permian (or Triassic) 

 rocks. This is most naturally explained by supposing it to have 

 been erupted to the (probably sub-aqueous) surface in the interval 

 between the deposit of the under- and the overlying rocks. 



3. The lavas usually form horizontal or nearly horizontal beds, 

 not vertical or highly inclined dykes. At Killerton, Dunchideock, 

 and Posbury these beds are in each case many acres in extent, and 

 at Western Town Mr. Yicary " caused a pit to be dug in the trappean 

 floor of the quarry and beneath it found a layer of hard sandstone." 



4. According to De la Beche,* " along the range of the igneous 

 rocks, particularly on the north, a sand occurs, here and there 



^ Eeport, p. 201. 2 Qp^ supra cit. p. 46. 



3 Ibid. p. 203. * Report, p. 200. 



