chap, xiii. Rocks of. 423 



Tliere is some fine-grained syenitic granite, orbicularly 

 marked by ferruginous lines, and weathering into 

 vertical, cylindrical holes, almost touching each other. 

 In the gneiss, concretions of granular feldspar and 

 others of garnets with mica occur. The gneiss is 

 traversed by numerous dikes composed of black, finely 

 crystallised, hornblendic rock, containing a little glassy 

 feldspar and sometimes mica, and varying in thickness 

 from mere threads to ten feet : these threads, which 

 are often curvilinear, could sometimes be traced running 

 into the larger dikes. One of these dikes was remark- 

 able from having been in two or three places laterally 

 disjointed, with unbroken gneiss interposed between 

 the broken ends, and in one part with a portion of the 

 gneiss driven, apparently whilst in a softened state, 

 into its side or wall. In several neighbouring places, 

 the gneiss included angular, well-defined, sometimes 

 bent, masses of hornblende rock, quite like, except in 

 being more perfectly crystallised, that forming the 

 dikes, and, at least in one instance, containing (as 

 determined by Professor Miller) augite as well as 

 hornblende. In one or two cases these angular masses, 

 though now quite separate from each other by the solid 

 gneiss, had, from their exact correspondence in size 

 and shape, evidently once been united ; hence I cannot 

 doubt that most or all of the fragments have been 

 derived from the breakiug up of the dikes, of which we 

 see the first stage in the above-mentioned laterally 

 disjointed one. The gneiss close to the fragments 

 generally contained many large crystals of hornblende, 

 which are entirely absent or rare in other parts : its 

 folia or laminae were gently bent round the fragments, 

 in the same manner as they sometimes are round concre- 

 tions. Hence the gneiss has certainly been softened, 

 its composition modified, and its folia arranged, sub- 



