Bernard Hohson — Permian Breccias of 8. Devon Coast. 313 



light, and contains porphyritic brown mica. It closely resembles 

 a spherulitio liparite I collected in the Grand Canon of the Yellow- 

 stone. The most interesting volcanic fragments, however, are those 

 which closely resemble the Permian felspathic traps from the Exeter 

 district, which I described in 1892.' One of these volcanic fragments 

 shows felspar laths in an opaque matrix consisting partly of iron-ore 

 and partly of red ferruginous matter. It can be paralleled by my 

 section 897 of olivine basalt from Eaddon Quarry, west of Thorverton, 

 six miles north of Exeter, although I must confess that none of the 

 fragments in the conglomerate matrix show the red pseudomorphs 

 after olivine so characteristic of most of the Permian felspathic 

 traps. Another type of volcanic fragment very abundant in the 

 matrix shows felspar laths, iron-ore, brown mica, and abundant 

 faintly greenish short prisms, which appear to be apatite. This 

 •rock resembles the one I termed mica-augite-andesite (Mr. TealP 

 terms it biotite-trachyte), especially ray section 900 from Killerton 

 Park, 5| miles N. 33° E. of Exeter, except that the Killerton rock 

 <!0ntains augite, which I have not found in the Watcombe fragments. 



The late Mr. K. N. Worth ^ also records felspathic traps from the 

 Permian breccias of Torbay and Teignmouth. On the other hand, 

 Mr. W. A. E. Ussher ^ says : " The absence of signs of con- 

 temporaneous vulcanicity in the New Eed Rocks of the area [Geol. 

 Survey Map, new series, sheet 350, Torquay] is accounted for by 

 their evident attenuation through a conformable overlap, as they are 

 traced northward towards Exeter, the volcanic horizon being higher 

 -in the series than the strata represented on the map." 



Three alternatives occur to me — either Mr. R. N. Worth and I are 

 mistaken in our identification of Permian felspathic traps in the 

 breccias, or Permian felspathic traps may occur in situ beneath and 

 hidden by the breccias, or Mr. Ussher may perhaps be mistaken 

 in considering that they lie at a higher horizon. 



5. Cove ^ mile south of Bundle Head. 



Here, again, the Devonian limestone fragments in the breccio- 

 conglomerate still form a majority : the largest (subangular) lime- 

 stone block observed measured 1ft. 4 in. by 1 ft. 1 in. by 8 in.; an 

 average fragment would be about 3 inches across. Fragments of 

 red and purple sandstone, quartz-porphyry and so-called lydian stone ^ 

 occur. A subangular block of quartz-porphyry measuring 1 ft. by 

 6 in. by 11 in. + (partially embedded) was observed, and about 

 a dozen smaller quartz-porphyry blocks occurred within three yards 

 of it. The rock fragments here are on the whole smaller than at 

 Petit Tor Crags and Watcombe Cove. Seams of breccia alternate 



1 Quart. Joum. Geol. Soc, p. 496. 



' Ussher & Teall, Geol. of Exeter : Geol. Survey Mem., 1902, p. 78. 



3 Quart. Journ. Geol. Soc, 1890, p. 76. 



* Geol. of Torquay: Geol. Survey Mem., 1903, p. 112. 



* Rev. J. J. Conybeare, "On the Red Rock Marie, or Newer Red Sandstone" : 

 Annals of Philosophy, x.s., vol. i (1821), p. 257; Conybeare & Phillips, Geol. 

 England and "Wales, pt. i (1822), p. 294 ; also Ussher, Geol. of Torquay, 1903, 

 p. Ill, line 4. 



