168 ME. E. E. COWPEE EEBD ON THE GEOLOGY OE [May 1 89 5, 



[52]. In this rock the material included in the mosaic patches is 

 therefore of a felsitic nature. (The rock also exhibits the direction 

 of flow by the parallel position of its felspar-crystals, some of which 

 have corroded angles.) Another example is [74], in which the 

 orthoclase-crystals show a zone of second growth ; and in another 

 case [6] microlites abound in the cryptocrystalline base. 



Of microlitic felsites in which this stage of ' micropoikilitiza- 

 tion ' is shown there are very numerous instances from all parts of 

 the area examined, — Carn Gelli, Cwm Brandy, Goodwick, Strumble 

 Head, Carn Llys, near St. Nicholas, etc. [91], [92], [245], [61], 

 [215], [210], [212], [213], [26], [329], and [324]. In the case of 

 two from Carn Gelli [91], [92], perlitic structure is beautifully 

 shown ; and the two slides prepared from a rock from Carn Llys 

 [329], [324] are remarkable for having no porphyritic felspars — a 

 very unusual condition in microlitic felsites from this area. 



A further stage, in which the mosaic is very well and completely 

 developed throughout, is seen in some rocks from Strumble Head 

 and elsewhere. 



In many cases the inclusions consist of the microlites of the 

 original groundmass [238], [254], [223], [2] ; and in [87] perlitic 

 structure is visible. 



Amongst felsites with cryptocrystalline inclusions in the micro- 

 poikilitic mosaic may be mentioned [229] with large porphyritic 

 crystals, and [256], [246] from Strumble Head. 



The most complete ' micropoikilitization ' is shown in a rock from 

 Bigney [208], [221], and in another from Carn Llys [347]. 



In some cases a slightly different type occurs ; patches of clear 

 quartz containing no inclusions are scattered about in the midst of 

 a microlitic or cryptocrystalline groundmass [207]. Perhaps the 

 peculiar rock from the neighbourhood of Carn Gelli [210], in which 

 the groundmass appears to consist of true micropegmatite with 

 blurred and indistinct small felspar-crystals and microlites lying in 

 it, may be related to the true micropoikilitic type. 



But, putting aside these two last-mentioned exceptional and 

 aberrant types, we see that it is possible to trace without difficulty 

 a series of stages in the development of the micropoikilitic structure 

 from rocks with a simple microlitic or cryptocrystalline ground- 

 mass. The first stage in its development exhibits disconnected small 

 patches occupying only a small proportion of the whole rock ; the 

 next stage shows them more closely aggregated and usurping the 

 greater part of the rock ; in the third stage they pervade the 

 whole substance of the rock, but still have somewhat indistinct 

 boundaries and are of small size ; and in the last stage they form 

 coarse mosaic, with the individual patches clearly defined. Haworth, 

 in his previously quoted account of this structure, though considering 

 it original and a type of the holocrystalline structure, yet expresses 

 the opinion that this coarse, clearly defined mosaic is its most perfect 

 and complete development, thus arriving at the same conclusion, 

 though on different lines. 



The next fact to be taken into consideration in connexion with 



