86 PEOCEEDIXGS or THE GEOLOGICAL SOCIETY. 



as spheriilites, let us see how far we can connect those whicli we 

 commonly meet with in igneous rocks. 



Commencing with the hyaline stage (omitting minerals of an- 

 terior consolidation, and restricting our remarks to the rocks con- 

 taining a fair proportion of felspar), we seem able to trace the 

 following order of phenomena in rocks which do not remain in a 

 condition so vitreous as the tachylytes, pitchstones, and obsidians : — 



A. (1) Formation of a large number of felspar microliths, so that 

 these become the most conspicuous objects in the ground-mass, in 

 which, however, more or less glass usually remains. This gives the 

 tracliytoidcd structure of many authors, and indicates a compara- 

 tively rapid fall of temperature, so that crystal-building is arrested 

 at a rather early stage. 



(2) Further development of the felspars, until their chemical 

 constituents are wholly, or almost wholly, removed from the magma, 

 followed by crystallization of the residue. This is the ophitic 

 structure of authors, most perfectly developed when the residue has 

 the composition of a pyroxenic mineral and itself crystallizes rather 

 coarsely — a structure, I think, indicative of more gradual cooling, 

 but still under no great amount of constraint. Is'ot usual in the 

 more acid rocks. 



B. (1) Structure minute ; but a glassy base* is not distinguishable. 

 Want of definiteness of external crystalline form, as though some- 

 times the separation of adjacent minerals had not been absolutely 

 perfect, or an irregularity of boundary, as though crystallization 

 had been simultaneous. Occasionally there is some approach to a 

 coordination of structure, a more or less imperfect micrographic or 

 spherulitic arraugement being visible. This is the petrosiliceous 

 structure of authors, indicative, I believe, of constrained consoli- 

 dation. 



(2) Generally a coarser structure than the last. The separation 

 of the minerals is more complete, and the felspars tend to have 

 rectilinear boundaries. The microgranulitic stimcture of authors, 

 which perhaps ultimately may be in great part classed with the 

 nest one, the remainder belonging to the first, and both belonging 

 chiefly to the less basic rocks. 



(3) Fairly coarse felspars, usually well-defined externally, espe- 

 cially in the more acid varieties, the granitoid structure of authors, 

 especially characteristic of granites, tonalites, most syenites and 

 diorites, and gabbros. 



This last method of crystallization seems to belong to very deep- 

 seated rocks, where consolidation has taken place under great 

 pressure and in the presence of confined water-vapour ; certain 

 cases of (2), and perhaps some of (1), indicate the same process 

 carried on more rapidly, and so occur frequently in vein-granites 

 and intrusive felstones. The first, or petrosiliceous type, however, 

 requires much discussion, and with some remarks on questions 

 relating to this, I must conclude this already too lengthy address. 



For many years the subject of the minuter crystal-building has 

 been present to my thoughts, although the pressure of what I may 



