498 THE MARQUETTE lEON-BEAEESTG DISTRICT. 



hornblende are seen lying in a greenisli-gray groundmass, and producing in 

 places a well-defined luster-mottling. The hornblende of these rocks is a 

 compact blue-green variety, in long columnar crystals that are idioraorphic 

 in the prismatic zone. Occasionally they are aggregated into areas resem- 

 bling those of ophitic augite, but usually the}' appear to be scattered indis- 

 criminateh' through the rock mass, Avhicli consists largely of chlorite, 

 epidote, and the remnants of yery much decomposed plagioclase. The 

 amphibole has undoubtedly been formed at the expense of previously 

 existing p^TOxene, for the hornblende rocks are In many instances but local 

 phases of well-defined altered diabases. The hornblende has grown until 

 it has extended beyond the areas formerly occupied by augite into the 

 matrix produced by the decomposition of the plagioclase. This hornblende 

 is always compact, and its crystals are often twinned. The rocks containing 

 them are "diorites" in structm-e as well as in composition, though, of course, 

 they are not diorites which have crystallized directly from a magma. 



In another class of the diorites, represented by the knob in the center 

 of sec. 12, T. 47 N., R 27 W. (Atlas Sheet XXVIII), the compact, apparently 

 idiomorphic amphibole is evidently a pseudomorph of augite. Remnants 

 of pink augite may still be detected In the indi^ddual hornblendes, and 

 occasionally nearly complete crystals of the minerals may be observed. 

 Rocks of this kind were originally augite-porphpites. 



One other exposure deserves to be mentioned, on accoimt of its pecul- 

 iar appearance. It is on the north side of the Duluth, South Shore and 

 Atlantic Railway track, in the garden of a house built on a hill about half 

 a mile east of the Ishpeming station. The rock exposed is a coai'se-grained, 

 slightly foliated one, with a smooth, glaciated surface, mai-ked by concentric 

 or spiral lines, resembling on a large scale the perlitic cracks in glassy 

 rocks. When broken the fresh fracture of the rock presents no unusual 

 featm-es. The phenomenon noticed on its exposed surface is apparently 

 that of spheroidal weathering, for the partings which produce the lines do 

 not extend any considerable distance below the sm-face. 



Contact phenomena around the eastern greenstones are rarely seen. 

 The only evidence of endomoqjhous contact action noticed in any of the 

 eastern knobs was observed in that forming the foot-wall of the open pit 



