384 THE VERMILION IRON-BEARING DISTRICT. 



are characterized by large poikilitic plates of liypersthene several inches in 

 length, which show bright reflecting faces. Most of these rocks when 

 examined under the microscope appear as granular aggregates of the 

 various minerals enumerated and give no clue to the original rock from 

 which they were derived. Some of thenl, however, still contain the 

 rounded areas to which reference has already been repeatedly made, and 

 show conclusively that they have undergone a more or less complete 

 recrystallization. In these the areas are always outlined by a zone of 

 magnetite, rarely with some hematite. In some cases this magnetite occurs 

 as a very fine dust; in others the magnetite is in relatively large masses. 

 Ordinarily the boundaries between these areas and the adjacent quartz of 

 the groundmass are sharply marked by the magnetite zone. When the 

 areas are close together the magnetite border of the one unites with that of 

 the ones adjacent, and such union tends to destroy the regularity of the 

 areas. Indeed, when the areas are closely crowded they run together more 

 or less. . When, in addition to being close together, the interior of the areas 

 is occupied by magnetite, as is commonly the case, the resulting rock is 

 composed of a mass of magnetite with little quartz and none of the rounded 

 granule areas are visible. In many of the areas quartz is the chief con- 

 stituent, ixL relatively coarse grains. Within these grains occur dust-like 

 crystals of magnetite which are accumulated either at the centers of the 

 grains or just within their peripheries. Outside of the areas occurs the 

 matrix, which consists now of coarsely crystalline quartz. When viewed 

 between crossed nicols, however, it is seen that the large quartz individuals 

 of which the matrix is composed pass across the boundary and extend into 

 these areas. 



We can readily see how such a rock as this might be produced from 

 the one already described (p. 380), in which essentially the same conditions 

 existed, with the difference that the rounded areas of silica and limonlte 

 were bounded by limonite, and that the quartz was in fine fibrous spheru- 

 litic aggregates with limonite at the centers and bounding their peripheries 

 (PL XII, A). By dehydration of the limonite there would be produced 

 hematite or, if insufficient oxygen were present, as appears to have been the 

 case throughout this region, magnetite. With the limonitic rocks tliere is 

 found associated ferruginous carbonate, which also contains undoubtedly 

 some lime and magnesia. Recrystallization of this material in combination 



