T1IK 1'KKMIAX SYSTKM 333 



out by Herr von Keinach. 5 The Devonshire Series is much 

 thicker than that of the Midlands, but thins northward and 

 appears to have been formed in an isolated lake basin which 

 persisted throughout both Permian and Triassic times till it was 

 merged into a larger inland sea (see Fig. 64, p. 201). 



The succession, where most fully developed in South Devon near 

 Teigumouth, Dawlish, and Exeter, is shown in Fig. 108, and has 

 been described by Mr. Ussher 6 as follows : 



Feet. 

 Red marls with occasional sandstones (? Permian) about 500 



Red sandstones 250 to 400 



Red conglomerate and breccia .... 900 to 1000 

 Dark-red Watcombe clays (local) .... 100 to 150 



Maximum about 2000 



The Watcombe clays occupy some little space north of Torquay 

 by Watcombe, Barton, and Daccombe, and they reappear between 

 Teignmouth and Bishopsteignton. They consist of fine dark-red 

 clay which has long been used for the manufacture of terra-cotta 

 ware, and also for the making of tiles and hard bricks. The 

 deposit seems to fill a broad hollow in the Palaeozoic rocks, for it is 

 overlapped by the breccias both to the south and north as well as 

 westward. 



The overlapping breccias consist in the lower part of con- 

 glomerate and breccia, in which fragments of Devonian limestone 

 are abundant. These beds are well exposed in the fine cliffs 

 between Babbacombe and the estuary of the Teign, and their thick- 

 ness is estimated at from 400 to 500 feet, but the frequent faults 

 make the estimate uncertain. The higher breccias differ in con- 

 taining a larger proportion of derived volcanic rocks, chiefly of a 

 peculiar red quartz-porphyry, and some of the boulders are of large 

 size. This portion has a thickness of from 500 to 600 feet, so that 

 the total thickness of breccia is about 1000. These breccias stretch 

 northward beneath the Haldon Hills to Dunchideock and Ide. 



South-west of Exeter there are sheets of trachytic and andesitic 

 lava at their base, indicating the contemporaneous existence of 

 volcanoes, but the material of these lavas is quite different from 

 that of the blocks in the breccia, which must have belonged to 

 another set of volcanic eruptions, probably of rather earlier date. 



A similar breccia is found in the Crediton valley forming a long 

 tongue which runs westward across the valley of the Tawe, and has 

 an outlier at Hatherleigh (see Fig. 64, p. 201). North of Silverton 

 the breccias pass into loose gravels which contain a different assem- 

 blage of stones, evidently derived from rocks to the northward, 

 and the whole series becomes thinner as it is traced into Somerset 



