400 J. CLIFTON WARD ON TTIE MICROSCOPIC STRUCTURE OF 



bling that of Cadcr Idris, "being fclspathic, somewhat harsh to the 

 touch, often porphyritic, sometimes columnar"*. 



The specimens which I have examined under the microscope arc 

 from the summits of Aran Bcnlyrm and Aran Mowddwy, and their 

 general appearance is represented in figs. 16, 17, and 18. 



1. Aran Mowddwy (figs. 16 and 17). — In the sliced specimens 

 of this rock, the characteristic folsitic structure is seen. Upon a 

 white, hazy, and milky-looking ground are scattered numerous small 

 irregularly shaped particles, some forming greenish and yellow- 

 ish-green patches (fig. 16) ; these seem to be all products of altera- 

 tion, and arc quite distinct from the felsitic base. Porphyritically 

 imbedded in the base arc scattered crystals of felspar, some of which 

 have margins apparently shading off into the milky matrix; others are 

 distinct orthoclasc twins j and there are a few instances of fragments 

 of triclinic felspar. Under polarized light the hazy felsitic base 

 breaks up into an irregular coloured breccia, very characteristic of 

 "this class of rocks and quite unlike any thing met with in other 

 groups of the volcanic series (fig. 17). As the polarizer or analyzer 

 is rotated the colours change to their complemcntaries ; but there is 

 nothing whatever in the felsitic base, when viewed by unpolarized 

 light, to show where the particular outlines of the coloured parts 

 will appear. The small particles scattered about the hazy base 

 maintain their distinctness while the change of colour goes on around 

 them ; and the thicker green patches appear dark, for the most part, 

 under crossed prisms. Many of the felspar crystals are much altered ; 

 but in several cases a twin structure is well seen in polarized light. 

 The following is an analysis f of this Aran Mowddwy feist one, 

 beside which I will place, for comparison, Durochcr's mean of analyses 

 by different experimenters : — 



Aran Mowddwy. Durocher's Mean. 



Silica 83-802 ' 754 



Alumina 7'(>80 15-0 



Potash 2-161 3-1 



Soda 4-229 1-3 



Limo -89G -8 



Magnesia -109 11 



Ferrous oxide '408"l ^ 



Ferric oxide -Ill J ^ 



Bisulphide of iron *191 



Phosphori c acid -089 



Sulphuric acid *017 



Carbonic acid trace. 



Loss on ignition -301 1-0 



100-00 100-0 



It will be seen that the percentage of silica in the analysis is un- 

 usually high, exceeding the maximum given by both Haughton and 

 .Ourochcr by 7 or 8 per cent. The percentage of alumina, on the 

 other hand, is low, while that of soda is fairly high. There is little 



* 'The Geology of North Wales,' p. 33. 



t This, together with three analyses of Cumberland rocks, given hereafter, 

 have been made for me by Mr. Jobn Hughes, F.C.S. 



