314 MR. J. PARKINSON OiS THE ROCKS OF [May 1900, 



to total reconstitution. The testimony even of hand-specimens is 

 clear. Sometimes we find black fragments included in the granite ; 

 at others dark shapeless patches embodied in the surrounding 

 rock; finally, these completely disseminated among the other 

 constituents. Thin sections show that these resemblances are 

 carried into even the smaller details of rock-structure. Here are 

 the same small scattered flakes of mica and hornblende, the same 

 meshwork of long felspars, the same infilling quartz, the same 

 needles of sillimanite. Minor points of difference do indeed exist: 

 the mica is perhaps often more fibrous in character ; the granular 

 hornblende is wanting in great measure ; the long rectangles of 

 haematite are absent. The plagioclases are often zoned, and the 

 extinction of the central parts frequently indicates labradorite. In 

 one place we catch sight of fragments of diabase, less completely 

 disintegrated, where lingering traces of augite yet remain. 



A well-defined dyke, some 12 or 14 inches wide, cutting sharply 

 across a gneissose dyke belonging to the earlier brecciation, has 

 so close a relation to the granite (it contains both hornblende 

 and mica and some black fragments) that it is considered part of that 

 rock, although its point of origin could not be traced in the field. 



Accordingly we may conclude that the main body of the intrusive 

 granite was posterior to the intrusion which produced the various 

 hornblende-rocks connected with the diabase, though it is not sug- 

 gested that the difference of age was considerable. On the contrary, 

 at one point on the south side of Le Nez the rock producing the 

 earlier brecciation becomes much coarser, contains both orthoclase 

 and quartz, and is separable only with some difficulty from the 

 later granite. Comparing a series of sections, the latter is seen to 

 possess considerably more quartz, the orthoclase is characterized by 

 a microperthitic intergrowth, and, where mixture with the basic 

 rock has taken place, a texture and general structure are preserved 

 which indicate a true granite. Usually more mica is present in 

 such slices. Some microperthitic orthoclase is found in the Le Nez 

 series embedding the other constituents, and corroding the plagio- 

 clases, seeming to herald the approach of the next intrusion, that 

 of the true granite. 



The practical identity in the mixed rocks from the south and 

 north of the island shows, 1 think conclusively, that not only was 

 the intrusion of magma the same in both cases, but also the physical 

 conditions attendant on its intrusion were at least similar over this 

 comparatively large area. The products, however, of the earlier 

 brecciation, as I have called it, confined to Le Nez and its neigh- 

 bourhood, are so strikingly different that the operative causes can 

 scarcely have been the same. 



(5) Comparison with Rocks in Guernsey. 



The fine-grained black rock with granular hornblendes mentioned 

 on p. 311 differs from a diabase in important respects. The 

 hornblende is no doubt original for the most part, but the earlier 





