Physical Structure and older stratified Deposits of Devonshire. 645 



however,almost entirely destroyed by cleavage-planes cutting obliquely through them. On the 

 banks of the tributary of the Lyne, south from Waters'-meet, the slates become so highly calca- 

 reous, that an attempt was made to burn them for lime ; and we have been informed, that along 

 the same strike of country, similar beds have been found in Exmoor Forest. In the higher parts 

 of this group, the hard gritty and siliceous bands, generally of a dull gray colour, continue to 

 alternate with the slaty and chloritic beds, and are in some places very fossiliferous, as may be 

 seen in the Valley of Rocks, and at the great Castle-Rock on the coast ; near which the series 

 ends — being carried by a southern dip under the second group. A passage, indeed, seems to be 

 made into the red upper group, through the intervention of some bands of red, earthy schist, alter- 

 nating with micaceous and siliceous flagstone. In the absence of any visible base, it is impos- 

 sible to estimate the entire thickness of the formation, we have been describing ; but from the 

 lowest beds in the gorge of the Lyne to the crests of the hills overhanging the Valley of Rocks, 

 and still more to the beds of passage above indicated, there is a thickness of many hundred feet. 

 Among the fossils obtained from this group (unfortunately in the form only of casts, and gene- 

 rally ill-preserved) we have many corals, an Orthis, and two Spirifers ; but as we shall return 

 to this subject, we will not, for the present, attempt to describe our imperfect list. 



2. The most striking characters of the second group are its coarse texture, and red or purple 

 colour. These colours are partly derived from the red oxide of iron diffused through the beds, 

 and partly from the veins of the same mineral which traverse the cliffs, and are washed far and 

 wide by the waters. Indeed, some of the beds themselves (where no veins intersect them) are so 

 impregnated with an earthy oxide of iron, that they are dug up entire, and the richest parts of 

 them are extracted for the furnaces*. Cases like these are, however, to be regarded as rare 

 exceptions. The coarser beds are generally thick, irregularly jointed, and break into rude, shape- 

 less, angular fragments ; and their colours vary through white, gray, greenish gray, variegated, 

 and purple, to deep red. The coarsest varieties, at North Hill and in some parts of the Quan- 

 tocks, pass into a conglomerate of red sandstone, with quartz pebbles, not distinguishable from the 

 conglomerates of the old red sandstone. Some of the hills west of Porlock are covered with a 

 white sandstone of this formation, fragments of which lie bleaching on the moors, and might be 

 mistaken, at first sight, for grits of the coal formation. The fine sandstones contain, however, 

 fragments of clay-slate, and not unusually pass (through the intervention of purple, variegated, 

 micaceous, flaggy grits, exactly like the flaggy beds of old red sandstone) into red slate, grey- 

 wacke, and greywacke slate ; among which, bands of fine, glossy clay-slate are here and there 

 interpolated. Considered as a whole, and from mineral characters only, we might compare some 

 parts of this group with the most characteristic portions of the old red sandstone. Other parts we 

 might compare with the red arenaceous beds of St. Abb's Head and the Lammermuir chain ; with 

 the red sandstones and schists of the Longmynd, Lyth, and Haughmond Hills of Shropshire ; and 

 with some of the coarser portions of tlie Upper Cambrian System, as developed in South Wales. 



The whole thickness is exposed only in the coast section ; extending, as before stated, from the 

 west side of the Castle Rock to Combe Martin. The dip is steadily towards the south ; and there 

 are, we believe, few breaks or dislocations by which the same beds are repeated over again. 

 The breadth of the formation, measured transversely to the strike, is nearly three miles ; the 

 thickness of the group must, therefore, be very great. Considered as a whole, it is neither cal- 



* There is a work of this kind on the south side of the red sandstone plain, about two miles 

 from Timberscombe, in Somersetshire ; and similar works were once carried on in the hills north 

 of Combe Martin. (See Map.) 



4o2 



