SYSTEM A TIC PETROGRAPHY 49 I 



as examples, many of the forms derived from them by diagen- 

 esis or metamorphosis, but does not outline the system for dis- 

 criminating and naming these alteration products. Some of the 

 metamorphic rocks, such as gneiss and mica schist, may be 

 formed from several primary rocks. 



The proposition made by Walther is manifestly rather the 

 work of a geologist than of a petrographer (as was pointed out by 

 Brogger in discussion, when it was presented to the Congress). 

 Like many discussions of principles concerned in the systematic 

 problem, it is not sufficiently worked out to show a practical 

 result, and does not fully test the adaptability of the chosen fac- 

 tors for petrographic system. But it seems to the writer that in 

 this renewing of effort to treat the non-igneous rocks in logical 

 systematic manner lies ground for hope that something more 

 than an arrangement for convenience may develop during the 

 early years of the twentieth century. 



Returning to a consideration of the principles adopted by 

 Walther, it may be remarked that the first one would be excellent 

 if the processes of rock formation were all open to examination. 

 Unfortunately, they are not so, in all cases. Many igneous rocks 

 and nearly all of metamorphic origin have resulted from pro- 

 cesses we cannot see in operation and can only imperfectly 

 imitate in experiment. The fourth rule is not a necessary con- 

 sequence of the facts stated under II and III. It is open to 

 argument whether the processes which originally produced a 

 rock are more deserving of recognition in petrographic system 

 than the processes which have greatly or entirely changed the 

 characters and perhaps even the composition of the original mass, 

 making the rock now accessible to our studies. 



As to rule V it can hardlv be said to warrant the application 

 made of it, in establishing the three divisions of unstratified, 

 stratified, and dike forms for all kinds of rocks. Where the 

 relations expressed by these terms have some genetic connection 

 with the properties of the rocks they may perhaps be adaptable 

 to classificatory purposes, but there is no logical reason for 

 applying this principle in unqualified form. 



