34 JOURNAL OF THE 



the 'mesh-structure' is distinctly brouo-ht out, and the 

 orig-inal <2franular nature of the rock clearly seen, as 

 shown in iig-ure 6. 



The variation between the different mineralog-ic va- 

 rieties within the same rock mass, and the essential 

 unity of the whole peridotite <iroup, have been pointed 

 out above in the descriptions of the various t3^pes. 

 Dunite, harzburg-ite and enstatite rock are often found 

 to blend into each other as inseparable parts of the 

 same rock mass, with no banding- or irreg-ularit}^ of 

 structure, whatever, between the different tj'pes. These 

 phenomena are such as are referred to mag-matic seg"- 

 r^g"ation or differentiation in well recog"nized ig-neous 

 rocks. In the enstatite bearing- type, the crystals of 

 this mineral are much larg-er than the gfrains of olivine 

 and are in the form of broad, flat plates; yet, when the 

 rock has not been sheared, no trace of parallel arrang-e- 

 ment has been detected. Even when considerably 

 sheared and laminated, only a partial parallelism has 

 been effected between the en.^tatites, and they have 

 usually been altered into talc. 



Similar transitions from dunite to amphibole-picrite 

 are found within the peridotite area at Buck Creek (PL 

 II.). Here, too, and also throug-hout a considerable 

 territory to the southwest, transitions are found from 

 dunite, the pure olivine rock, to troctolite, the coarse 

 olivine-feldspar type; and the latter, ui turn, passes in- 

 to the pure anorthite rock, anorthosite. '^ 



The extremely fresh condition of the olivine, even in 

 the surface exposures, is a very striking- feature of 

 these peridotites. Most of them, it is true, show some 

 tendency to serpentinization when examined with the 

 microscope, and in a few cases they have underg-one 

 complete transformation; but a number of specimens 



