196 SCIENCE PROGRESS. 



be at least equally significant. On the ground of chemical 

 composition alone, Rosenbusch (i i) has claimed as meta- 

 morphosed sediments various gneisses in Sweden, in the 

 Black Forest, in the Odenwald, in Bavaria, etc., as well as 

 certain hornblende-schists and allied rocks ; while he has 

 shown that other gneisses in Sweden, in Saxony, in the 

 Vosges, etc., give chemical analyses, which affords strong 

 presumption of their true igneous origin. Under the latter 

 head he also places the Saxon granulites, various halleflintas 

 and eurites from Sweden, and pyroxene-granulites, eclogites, 

 chlorite-schists, and talc-schists from various localities. The 

 day of sweeping theories of the origin of gneissic and 

 foliated crystalline rocks in general is past, and it is very 

 generally admitted that each area must be treated as offering 

 a separate problem in itself. We have yet to learn to what 

 extent the hypothesis of thermally metamorphosed sedi- 

 ments is applicable to large tracts of crystalline rocks, but 

 the present tendency of opinion, while leaving room for the 

 dynamic theories of the school of Heim and Lehmann, still 

 points to a partial vindication of the early Lyellian doctrine 

 of metamorphism. 



Passing on to the arenaceous rocks, we may remark 

 first that different types which would be roughly grouped 

 under this head give rise by metamorphism to widely dif- 

 ferent products. In the metamorphism of a pure quartz- 

 sandstone there can, in the nature of the case, be no very 

 sharply defined stages. The rock is not capable of any 

 transformation short of complete recrystallisation, and so it 

 will be practically unaltered except where a certain high 

 temperature has been reached during the metamorphism. 

 Some minor changes should, perhaps, be excepted, such as 

 the expulsion of water from the fluid-pores of the quartz, 

 an effect noted by Sorby at Salisbury Crags, Edinburgh. 



The quartzite which results from the complete recrystal- 

 lisation of a sandstone is, in general, easily distinguished, at 

 least in microscopic preparations, from a quartzite produced 

 by the infiltration of a siliceous cement. In the latter case 

 the difference between the original clastic grains and the 

 new-formed quartz-cement is clearly accentuated, while the 



