IORI One MOLOGY 
[MESSI OAR NILA Cla, SOOO 
WIENS, IMKOMIFINCIWAINUIRE, Ov ISIC IDSA ele 
GRANOLITES.* 
Most petrographers agree that the classification of granular 
rocks, if not of lavas, should be based on mineral composition. 
This resolves itself practically into the molecular composition. 
When we state that a rock is composed of quartz, mica, and 
orthoclase in certain definite proportions, we state the relative 
proportions of the molecules of which these minerals are com- 
posed, and this is true of all other minerals which are made up 
of a single molecule. But when we introduce terms such as 
plagioclase, which is composed of two molecules in ever varying 
proportions, we no longer treat of molecules as such, but of mix- 
tures of molecules. It seems quite clear that the molecular 
method should be applied throughout, when practicable, and in 
calculating the composition of the feldspathic rocks the plagio- 
clase should be resolved into the constituent albite? and anor- 
thite molecules, and the term plagioclase should not be used. 
This is particularly necessary with monzonites and diorites, for it 
is clear that if we define typical monzonite as a rock composed 
of equal quantities of orthoclase and soda-lime feldspar, we may 
«Published by permission of the Director of the U. S. Geological Survey. 
2We may treat the soda of the feldspars all as in albite, although some of it may 
be in the orthoclase. 
Vol. VIII, No. 2. 105 
