Vol. 65.] PETROGRAPHY OF THE NEW RED SANDSTONE. 241 



extinctions; blue acicnlar tourmaline becomes less frequent, but tbe 

 brown stumpy prisms of this mineral are more common ; brookite 

 is scarce, and iron-ores are less abundant. In South Devon garnets 

 first appear on this horizon (Table II). 



The Pebble-Bed in Devon and Somerset presents two more or 

 less distinct types : one existing from the south coast to the neigh- 

 bourhood of Uffculm, the other obtaining from Uffculm northwards. 

 It will be convenient to refer to these respectively as the southern 

 and the northern type. Both types are bound together by the 

 prevalence of staurolite, and they may be compared and contrasted 

 as follows : — 



Southern type. 



Staurolite abundant. 

 Blue tourmaline rare. 

 Brookite rare. 

 Garnets absent. 

 Fluorspar present. 

 Ferruginous cement. 



Northern type. 



Staurolite abundant. 



Blue tourmaline more common. 



Brookite more common. 



Garnets present. 



Fluorspar absent. 



Ferruginous and calcareous cement. 



All along its course the Pebble-Bed contains many minute frag- 

 ments of schistose material ; and much of the quartz, especially in 

 the south, has undulose extinction. 



The Upper Marls and Sandstones, from south to north, are 

 characterized by an abundance of pink garnets and the general 

 occurrence of brookite ; and in the south more especially by the 

 occasional presence of actinolite. Staurolite is not so abundant in 

 these rocks as iu the Pebble-Bed, and, in addition, it is in these 

 upper beds that we most usually meet with such cementing-materials 

 as barytes, celestine, and gypsum in quantity. Barytes, however, 

 occurs less plentifully at certain localities on other horizons, both in 

 the Bed rocks of this district and in those of the Midlands. 



V. Conclusions. 



A most important factor in the study of these sediments is the 

 relation that they bear to the pre-existing rocks of the West of 

 England. In dealing with this subject several questions arise, the 

 most prominent of which is — What rocks were capable of yielding 

 the material which entered into the composition of the New Bed 

 Sandstone? There is no doubt that most of the older rocks, now 

 exposed at the surface, supplied detritus ; but it does not appear 

 possible, as pointed out in my previous paper, for them to have 

 furnished all the mineral species detected in the heavy residues. 



In the Lower Breccias and Sandstones staurolite occurs most 

 plentifully in the southernmost exposures ; and, taking into con- 

 sideration the evident local origin of these deposits, as proved by 

 the larger rock-fragments, it seems safe to assume that they were 

 laid clown not far from a series of staurolite-bearing rocks. No 

 such rocks, however, exist in this part of England, but, as previously 

 suggested, probably lay to the south of the present coast-line. 



