A^^'DESITES OF DEVONSHIRE. 505 



chlorite in No. 909, and chlorite, probably replacing augite, occurs 

 in Nos. 903, 904, 906, 919. 



Brown mica occurs in the groundmass of No. 889, 



Magnetite is very abundant ; indeed, it often occupies the greater 

 part of the angular spaces between the felspars, especially in the 

 less crystalline portions of the rocks, usually near the upper or 

 lower surface, e. g. in Nos. 897, 906 a, 908 a, 911, 918 n. It is very 

 generally altered to reddish ferruginous matter, which gives most 

 of the rocks a reddish colour even in the thinnest sections. 



Bronzite is mentioned by Prof. Biicking ^ as a constituent of the 

 rock of Dunchideock (locality not more precisely given). I have not 

 observed it in any of the specimens that I have examined. 



Amygdules are very common. They are most frequently occu- 

 pied by calcite or by zeolites (showing a grey interference-colour of 

 the first order) or by both.^ 



(b) (Quartz Inclusions. 



As already mentioned, these occur in ITos. 912 a, 913, 914, 915, 

 928 A & B, 929 A & c, also in 912 b. They vary in size, from 2 to 3 

 millim. diameter and less, are usually of a rounded form, and are 

 encircled by a rim of small augite-crystals, often obscured by 

 secondary calcite. Each inclusion consists as a rule of a single 

 individual, but aggregates of numerous individuals with interlocking 

 margins, as in a quartzite, occur, e. g. in Fo. 912 c. The evidence in 

 favour of regarding these quartz-crystals as inclusions is : — 



1. The basic nature of the rock, shown by the presence of olivine 

 and plagioclase. 



2. The rounded form of the crystals. 



3. The presence of reactionary rims of augite-crystals. 



4. The complete want of any uniformity in the distribution of 

 the quartz-crystals throughout the rock, which is in some places 

 rich in them, in others free from them. 



5. The fact that the rocks are interbedded with sandstones, 

 whence the quartz could be obtained when the rock was erupted 

 through the underlying beds. 



6. The most convincing proof is the presence at Knowle Quarry 

 (No. 912) of an oblong eye of quartz about 3x1 inch embedded in 

 the rock. I have the specimen in my possession. 



I am aware that Mr. J. S. Diller does not regard points 1 to 3 

 as proof that quartz is not (in some cases) an original constituent of 

 basalt ; indeed, they are all present in the case of the quartz-basalt 

 of Cinder Cone, 10 miles N.E. of Lassen's Peak, California, which 

 he describes,^ and with a section of which I have compared the 

 Devon basalts, but when combined with points 3 to 6 the evidence 

 appears convincing. 



1 In Mr. Ussher's paper, p. 248. 



2 In the ' Challenger' Eeport, Deep-sea Deposits, 1891, pi. xviii. fig. 4 gives 

 a good idea of the zeoHtes in No. 928 b, and figs. 2 and 3 of those in No. 929 a. 



^ 'A late Yoleanic Eruption in Northern California and its peculiar Lava,' 

 Bulletin U.S. G-eol. Survey, No. 79 (1891), pp. 24-28. 



