0. A. Derby — Magnetite Ore Districts in Brazil. 317 



An abandoned working badly obscured by forest shows the 

 ore masses in situ as rounded aggregates in soft decomposed 

 material which higher up the hill in the natural exposures is 

 charged with secondary siiica giving the siliceous masses above 

 described. The whole evidently forms a dyke some ten meters 

 or more in width cutting the Cambrian (?) schists. In the soft 

 material only decomposed mica (perhaps original biotite) could 

 be recognized. A washing gave an extraordinary abundance 

 of apatite and rare prisms of acmite. A loose block with 

 schistose structure consisting of a finely granular mixture of 

 apatite and acmite (the latter showing micaceous decomposi- 

 tion) serves to connect this cutting with another in the vicinity. 



A small dyke about 20 centimeters wide and about 100 

 meters distant from the one above described has afforded 

 specimens in which the original rock-type can be recognized. 

 It consists almost exclusively of orthoclase with some large 

 crystals porphyritically developed in a fine grained noncrystal- 

 line ground-mass with rare prisms of acmite and is thus a 

 typical augite-syenite.* Lower down the hill and apparently 

 coming from the same or a similar dyke are some large par- 

 tially decomposed, partially reconstructed (with secondary 

 silica) blocks of extreme interest. The soft material and the 

 totally reconstructed (with quartz, chalcedony and tridymite) 

 portions closely resemble those of the old mine above de- 

 scribed and the latter also contain enstatite. In other portions 

 in which the orthoclase is partially preserved in a kaolinized 

 state the introduction of secondary quartz gives the aspect of 

 an ordinary granite. The acmite is in great part transformed 

 to amphibole and finally to mica. It varies greatly in abun- 

 dance becoming in places the predominant element and with its 

 increase, apatite substitutes in great part, or wholly, the feld- 

 spathic element giving a rock composed entirely of pyroxene 

 and apatite. There is thus at this place an intimate associa- 

 tion of three apparently distinct types, viz : an orthoclase- 

 pyroxene rock with predominant feldspar, an orthoclase-apatite- 

 pyroxene and an apatite-pyroxene rock. These last two give 

 on decomposition an apatite-mica rock. 



A large road cutting on the opposite side of the ravine and 

 nearly in front of the locality last described is in totally decom- 

 posed schistose material in highly inclined layers. The 

 predominant rock is a pyroxene-apatite schist in which the 

 pyroxene (acmite) has, except in a few points, been wholly 

 altered to mica. Through this run narrow dyke-like streaks 

 (particularizations ?) relatively poor in apatite in which the 



* The daily increasing importance, wide range of variation and peculiar charac- 

 ter of the orthoclase-pyroxene combination seems to demand a simple non-com- 

 mittal term as a generic title for so-called augite-syenite group and Brogger's 

 name Laurvikite is here employed in that sense. 



