TRANSACTIONS OF SECTION C. 569 



of lime and the presence of potash and soda in varying proportions. According to 

 the relative proportions of the alkalies the rocks are roughly separated into three 

 groups : — 



(i.) Those in which there is a large excess of potash over soda, the latter being 

 present only in the smallest quantity (potash-felsites). 



(ii.) Those in which the soda, though present in considerable quantity, is yet 

 subordinate to the potash (potanh-soda-fehites). 



(lii.) Those in which thf> soda is in excess (soda-felsites). 



Group i. comprises felsites containing few or no porphyritic crystals : they are 

 composed mainly of a crypto-crystnlline or felsitic a^rgregato of quartz and felspar. 

 The few porphyritic crystals are not striated and consist of ortlioclase. 



Groups ii. and iii. embrace felsites in which a striated porphyritic constituent 

 is abundant. The felspar of the crypto- to micro-crystalline groundmass, how- 

 ever, is most probably ortlioclase, and it seems likely that the fluctuation in the 

 percentages of potash and soda in iliese rocks may be due to a variation in the 

 relative proportion between a porphyritic al bite- felspar and the orthoclase-felspar 

 of the groundmass. On the other liand it is not impossible that we may have to 

 deal with a felspar belonging to a triclinic potash-soda series (the anorthoclase series 

 of Rosenbusch). 



The modern equivalents of these ancient felsites are the rliyolites or liparites 

 and pantellerites, which have also been shown to be divisible into potash and soda 

 series. 



Some of the rocks in question show indications of having consolidated as 

 true glasses (pitchstone or obsidian). 



4. The Age of the Granites of Dartmoor and the Enfjlish Channel.^ 

 By A. R. Hunt, M.A., F.G.S. 



The examination of the crystalline and other rocks trawled in the English 

 Channel oft" the coast of South Devon has been this year brought to a close by the 

 sale of the fishing vessel which has for the past ten years supplied most of the 

 specimens. 



Thirty-two crystalline rocks trawled in the Channel have been microscopically 

 examined from time to time by the late Mr. E. B. Tawney or Professor Bonney, 

 and the analyses published in the ' Transactions of the Devonshire Association.' 



On comparing the granitoid rocks from the Channel, with nine specimens of 

 gi-anite, and of veins intersecting the granite or the adjacent sedimentary rocks 

 from the eastern flank of Dartmoor, the following points of ditierence may be 

 noted : — 



Tourmaline has not been detected in the ("hannel rocks, but abounds in Dart- 

 moor. Hornblende is absent from the Dartmoor granites but is often abundant 

 in those from the Channel. 



In the nine sUdes examined from the eastern border of the moor, as well as in 

 one derived from its south-western border, cubic crystals (presumably chloride 

 of sodium) are present in fluid inclusions in the quartz. 



In the fluid inclusions in the Channel rocks cubic crystals have not been 

 detected. IK^nco it would appear that the included fluid is water in the case of 

 the Channel rocks, and brine in the case of those Irom Dartmoor containing cubic 

 cry.stals. 



The bubbles in the Cliannel rocks are usually active, or at least easily moved : 

 those in the Dartmoor rocks are usually sluggish, if not immovable. 



Among the Channel rocks are gneisse.'^ both typical and hornblendic, and rocks 

 intermediate between granites and gneis.ses. 



The Dartmoor granites are not gneissic in character. 



Many of the ('hannel rocks indicate compression. The condition of the Dart- 

 moor granite is suggestive rather of fracture, with local crushings giving rise to heat 



' Trans. Deron. Assoc, vol. xxi. p. 238. 



