252 PROCEEDINGS OF THE AMERICAN ACADEMY. 



instability of phenocrysts as originally formed will be discussed more 

 fully in a later paragraph when the general question of the crystal- 

 lization of the rocks will be considered. 



In extreme cases, particularly where the rock shows otherwise 

 evidences of extreme modifications, the phenocrysts have been re- 

 duced to a fine mixture of two feldspars in size of grain very near to 

 that of the groundmass itself, and, if it were not for the preservation 

 of a part of the "streaks" and marginal portions of the phenocrysts. 

 one would be sometimes in doubt whether the rock contained feldspar 

 phenocrysts or not. The introduction of the granular material of 

 the groundmass along breaks in the phenocrysts is also a very 

 common phenomenon. 



iVnother conspicuous change is seen in the modification of the horn- 

 blende groups. The beginning of the alteration shows first in the loss 

 of the highly spongiform appearance of the groups brought about by 

 the development of minute rods or fibers of deep blue amphibole 

 about the edges or replacing the original mineral. Further changes 

 result in a general breaking up of the groups, the inner and more 

 massive portions, where- these occur, being also involved. Small 

 clusters of prismatic to almost fibrous, deep blue amphibole of the 

 riebeckite type, associated with abundant magnetite crystals, and 

 often with recrystallized or secondary feldspar and quartz, take the 

 place of the original group. At the same time more or less elongated 

 prisms and more irregular grains develop in the immediate neighbor- 

 hood, and to a greater or less extent, also make their appearance 

 throughout the rock generally. The aegirite associated with the 

 hornblende seems to be less easily affected but is nevertheless finally 

 invoh'ed in the alteration. Still further changes, in which both long 

 continued deep-seated alteration and more superficial decay doubtless 

 play a part, more or less completely destroy the original hornblende 

 and pyroxene and the resulting products, blue riebeckite shreds, mag- 

 netite with hematite and limonite or other ferruginous products, Ije- 

 come generally distributed through the rock and only clusters of 

 magnetite grains etc. remain to mark the position of the original dark 

 silicates. Of course the microliths of the groundmass also suffer a cor- 

 responding alteration. 



The quartz phenocrysts are often impregnated with the secondary 

 amphibole, and in the more advanced stages, are broken up and the 

 parts scattered. The quartz of the groundmass in these altered phases 

 i^. perhaps rather more distinct than in the normal rock. This is in 

 part because it is little affected by the changes, but also in part 



