68 F. P. Mennell — Cordierite in Dartmoor Granite. 



detected microscopically.' Within the limits of the granite proper 

 the mineral does not appear to have been observed at all in an un- 

 altered condition, at least so I infer from the absence of any record of 

 it in the above-mentioned memoirs, as I must confess to a somewhat 

 scanty knowledge of West of England geological literature. Having 

 recently done a certain amount of geologising along the northern 

 margin of Dartmoor, which has not been investigated by the Survey 

 since the introduction of modern petrographical methods, and having 

 obtained some rather interesting results, of which I hope to publish 

 shortly a detailed account, I was glad to take advantage of an 

 opportunity presented by a visit to the South of Devon to make some 

 comparisons with the contact and other phenomena described by the 

 Survey. 



One of the excursions made "with this object was to Ivybridge, 

 where the chiastolite slate is well known to geologists and where 

 the other andalusite-bearing rocks have been well described by 

 Mr. Barrow. Going up from the railway station alongside the Erme 

 valley, over the track which leads past a quarry where the chiastolite 

 slate is well exposed, one soon comes upon the granite, of which 

 a second small quarry affords an excellent section, close to its margin. 

 Nearest the contact the rock is even-grained and rather discoloured 

 by alteration. Further away, however, a fresh, grey, biotite granite 

 with large white pseudo-porphyritic felspar crystals is seen. Through 

 this are sparingly distributed idiomorphic six-sided or somewhat 

 rounded stumpy prisms of a dark-green mineral, often J to ■!■ in. in 

 length. The habit and general appearance of the mineral are quite 

 sufficient for its identification as cordierite, though the exceptionally 

 well-marked basal parting gives some of the specimens a rather 

 unfamiliar aspect. Many of the crystals are semi-transparent and 

 glassy-looking, and even the more altered of those which are con- 

 spicuous in hand specimens are, for the most part, colourless and 

 transparent in thin section except roxind their margins. They show, 

 however, here and there, some regularly distributed brownish 

 enclosures like those seen in diallage. Faint yellow pleochroic halos 

 are also occasionally seen, as well as some yellowish stains which are 

 not pleochroic. The smaller crystals, that is to say, those which are 

 only a millimetre or so across, are altered throughout to greenish 

 matted ' pinite ' aggregates. A specimen which I secured, but have 

 not sliced, contains a little rounded patch about \ in. in diameter, 

 consisting of fresh cordierite and a reddish mineral which I take to 

 be andalusite. The latter is very common in the adjacent contact 

 rocks, and I have found it to be abundant along the northern margin 

 of Dartmoor in certain offshoots of the granite, but it is curious tliat 

 this fine example of a granite containing fresh cordierite in conspicuous 

 crystals should be found where that mineral does uot seem to have 

 been detected among the metamorphosed sediments.^ It may be 

 surmised that the absorbed material, which has recrystallised in part 



^ Dartmoor memoir, p. 38. Mr. L. J. Spencer kindly ascertained for me 

 that there is no British specimen of unaltered cordierite in the National 

 Mineral Collection. 



^ The word cordierite does not appear in the index to the Ivybridge memoir. 



