658 GEORGE W. BAIN 



original rock must be considered to remain behind throughout all 

 the changes. Alumina seems to fill this place to perfection, since 

 once it is precipitated it seems to migrate again only under extreme 

 conditions, such as resolution in a batholithic magma and discharge 

 in pegmatites. Some additions of alumina may occur but their 

 effect will be pointed out later. 



Fig. 2. — Highly silicated limestone. (Magnification about 30 diameters. Ordi- 

 nary light.) The photograph shows the initial development of metapoikilitic struc- 

 tures with increasing amounts of diopside and woUastonite (strong cleavage). The 

 cloudy area near the center of the field represents a pseudoeutectic intergrowth of 

 dolomite and calcite. Calcite forms the remainder of the field. 



Consider first the change from limestone to garnet rock (Figs. 

 2 and 3) . Then since the garnet rock is formed from the limestone, 

 replacement of the latter must continue until there are 239 times 

 as many molecules of alumina as before. If the garnet rock was 

 formed by simple recrystallization of the limestone, the difference 

 expressed in the difference column would remain unadjusted and 

 the excess material would then have to be removed in solution and 

 would produce a great decrease in volume. But all field evidence 

 points to practically no change in volume in this rock. 



