MR. O. A. DERBY ON NEPHELINE-ROCKS IN BRAZIL. ZoO 



as phon.olite. Aside from this type, there is a group of basic rocks, 

 of basaltic aspect, which is thus far only known in small dykes, but 

 which may be suspected to occur in more important masses in some 

 of the unexplored ])ortions of the mountains. 



As at the Po^os de Caldas locality, there is abundant evidence 

 that the foyaite and ]jhonolite are but different phases of the same 

 original magma. In fact, part of the distinctly granular rocks 

 present as well-marked a division of the crystalline elements into 

 two generations, as pheiiocrysts {Einsprenglinye) and groundmass, 

 as do the typical phonolites, and might either be considered as 

 phonolites, with a coarsely granular groundmass, or as lb3'aites, 

 with a phonolitic structure. Por the study of the origin of the 

 eruptive mass, the distinctively phonolitic type is in its modes of 

 occurrence the most instructive ; it appears both in the form of 

 dykes and as effusive sheets. 



Only the dyke-phonolites, or loose masses that are presumed to come 

 from dykes, have been found in a state of preservation such as to 

 permit of satisfactory petrographical study. Among these, nearly all 

 the prominent characteristics of mineral composition and structure 

 may be observed. Eoth nephelinitoid and trachytoid phonolites 

 abound ; while, as will be seen farther on, there is a tendency 

 towards the leucite-phonolite type. The rocks of the first type 

 frequently exhibit the ocellar structure ; while those of the second 

 have a well-marked fluidal structure, and appear to pass into true 

 trachytes. Fragments of this last type are comparatively rare, 

 except in the bed of the Santo Antonio, where they have been traced 

 to a dyke, 2 metres wide, which cuts the gneiss a hundred metres 

 or so above the dam. A peculiar feature of some of these rocks is 

 the presence of polygonal granular inclusions or aggregations, 

 which, being common also to the foyaitcs, will be discussed after the 

 description of that group. 



A considerable number of phonolite dykes have been seen in the 

 cuttings through gneiss outside the area of the peak proper ; and, 

 in one or two instances, they have been met with at a distance of 

 several kilometres away from the eruptive centre. These peiipheral 

 dykes of phonolite are seldom less than one, or more than two, 

 metres in width, the dykes that are notably less than a metre in 

 width being universally, so far as present observations go, of a more 

 basic type of rock. Within the eruptive area — that is to say, the 

 area where foyaite occurs — the dimensions may be greater, one 

 dyke, forming a fall on the small river 8abino, being over ten metres 

 wide. In only two cases have phonolite dykes been seen cutting 

 foyaite : one is at the very summit of the peak ; the other in a 

 loose block at the foot (see fig. 4, p. 261). 



High up on the Eio do Ouro, and well within the crater-liko 

 depression, the river flows for a considerable distance (more than 

 100 metres at the least) over a mass of phonolite which is un (Ques- 

 tionably effusive. The limits of the mass were not seen, but it was 

 found extending laterally some scores of metres to the right, where 

 it forms a fall on a branch stream. Although sufficient to form 

 considerable cataracts in the river-bed, the rock is so decomposed 



