218 PROF. BOSWELL ON THE CEETACEOUS AND [vol. lxxix. 



Tourmaline is exceedingly abundant. It is present in all 

 tints, and usually in large grains (02 to 03 mm.), mostly irregular 

 in form. One flat hexagonal blue crystal, resting on the well- 

 developed basal plane, was observed. In certain ' pay-streaks ', 

 as noted above, the mineral forms as much as 25 per cent, of the 

 sand or gravel. Equidimensional grains as well as needles of 

 blue tourmaline occur frequently in these sediments, as in the 

 Cretaceous and Eocene deposits described above. 



Staurolite is, on the whole, not common, and the grains are 

 always small. They may* have been derived locally from other 

 staurolite-bearing sediments, such as the Greensand. 



Grlauconite is uncommon, and is certainly detrital, its source 

 being doubtless the marine Cretaceous deposits. 



Andalusite occasionally occurs, the grains being either clear 

 and pleochroic, or full of inclusions. In the latter case, the 

 variety should probably be termed chiastolite. The record, 

 however, of 'chiastolite' in the table on p. 226 refers to the 

 discovery of several excellent crystals showing the well-known 

 cross due to inclusions. Two of these are illustrated in figs. 6 a 

 & 6 6, p. 220. 



Topaz rarely occurs. Of tinstone several records have been 

 made. Hornblende, epidote, and garnet are recorded from 

 single grains. 



Anatase, as might be expected, is not uncommon in these 

 clayey beds, indigo-blue, yellow, and grey crystals having been 

 noted. The ilmenite present is frequently altered to leucoxene. 



Very large quantities of limonite occur. Most of the grains 

 result from the decomposition of the schorl, and many are doubtless 

 schorl-grains coated with limonite. The blue grains appear to be 

 more easily attacked. If this is the case, it may help to explain 

 the preponderance of brown crystals of tourmaline in sediments of 

 various ages. 



Kyanite has not been observed. 



The mineral assemblage described above is thus far from rich, 

 and could have been derived practically entirely from local deposits. 

 The small staurolite- and glauconite-grains may have been obtained 

 from the Cretaceous deposits of the Haldons, or from beds of the 

 same age now removed by denudation. 



Some of the andalusite and the topaz may have passed through 

 a cjcle of erosion and sedimentation, and thus have been derived 

 from older sediments, such as the Eocene and Cretaceous ; Dart- 

 moor would supply the rest. Of the ability of these minerals to 

 withstand such treatment, however, we have at present no proof. 



The occurrence of chiastolite showing the t3 r pical ' made ' 

 character is unusual in British sediments, and is confined, so 

 far as present records go, to the Bovey deposits and neighbouring 

 gravels (Riddaford). It is noteworthy that this variety of the 

 mineral occurs commonly in the Dartmoor aureole. 1 



1 It occurs also, of course, in the country of theRohans in Britanny, hence 

 their armorial bearing-s. 



