234 ME. H. H. THOMAS ON THE [Aug. 1909, 



as to form crystals of this shape. It is probable, therefore, that in 

 these crystals the true basal plane is not present, but that some 

 external factor has determined a tabular habit. In all probability, 

 these crystals were formed between the cleavage-planes of mica, and 

 thus had their dimensions restricted in the direction of the vertical 

 axis. 



It appears that flaky crystals giving an interference-figure were 

 detected by Mr. T. H. Cope * in the titaniferous iron-sands of Porth- 

 Dinlleyn, on the coast of Carnarvonshire. 



Barytes and celestine. — Both these minerals, more especi- 

 ally barytes, are met with as cementing materials, but (so far as 

 can be judged) never as detritus. Many of the sandstones and marls, 

 when broken up and separated by heavy solutions, give a large 

 residue. The barytes under the microscope consists, as might be 

 expected, of small rectangular grains bounded by the two good 

 cleavages {001} and {110}. It is associated with celestine, but 

 the two minerals are indistinguishable in microscopic grains. They 

 are easily identified, however, by their respective flame-tests ; and, 

 when occurring together, the presence of both is easily demonstrated 

 by spectroscopic methods. 



Celestine has been recorded from the Upper Marls of Peake and 

 Salcombe Hills, near Sidmouth, by Mr. S. G. Perceval. 



Cordierite ('?). — Only one grain which might possibly be re- 

 ferred to unaltered cordierite has been met with, and that from the 

 Lower Sandstone of East Town, east of Totland, in the Quantock 

 region. It occurred as a somewhat rounded grain, with traces of a 

 moderately good cleavage. It gave straight extinction, and showed 

 feeble pleochroism from pale blue-grey to an indistinct shade of 

 yellow. It contained a large number of minute opaque inclusions, 

 probably iron-ores. No pleochroic halos were noticed. If this 

 mineral is really cordierite, its occurrence is most unexpected ; for, 

 in rocks of this character, it would probably occur altered to pinite, 

 or some other decomposition-product. 



Staurolite. — On some horizons this mineral forms quite a 

 considerable portion of the heavy residues, and is one of the 

 most abundant heavy minerals. It occurs most often, in these 

 rocks, in angular and rough grains of a pale-yellow to amber colour, 

 measuring up to 0*5 mm. in greatest length. The angularity 

 (PI. XII, fig. 4 c) is due to the fraying-out of the mineral along the 

 prism- cleavages parallel to the form {110}. The cleavage parallel 

 to the form {010} is also a prominent boundary-plane. 



Very rarely it occurs in well-formed crystals. In the New Eed 

 rocks such good crystals as those figured in PI. XII (figs. 4 a & 4 6) 

 have not been met with. For descriptive purposes, therefore, I haye 

 shown two almost perfect crystals from more recent sediments, one 

 from the Pliocene of Lenham, and the other from the blown sand of 

 Newgale in Pembrokeshire. These crystals are tabular parallel to 



1 Proc. Liverp. Geol. Soc. vol. ix, pt. ii (1901-1902) p. 212. 



