686 CROSS, IDDINGS, PIRSSON, WASHINGTON 



tion of their optical properties and chemical composition with 

 that of the rocks containing them. The necessity is also appa- 

 rent for complete analyses not only as to the main constituents, 

 but also as regards some of the constituents hitherto considered 

 as rare or of little importance, but which modern investigation 

 has shown to be widely distributed and often of considerable 

 influence on the norm and mode. The constituents present usu- 

 ally in small amounts are distributed very unequally among the 

 various kinds of rock magmas, and it would save the careful 

 analyst much needless labor if the petrographer would indicate 

 to him those which it is advisable to look for and determine. 



EPILOGUE. 



Some who read this essay will, without doubt, object to the 

 method of classification herein proposed because it is new and 

 embodies new principles, since experience has shown that one's 

 mental attitude toward new methods is on the whole conserva- 

 tive, and tends to resist their introduction. It is, also, easier to 

 travel along a familiar and oft-trodden path, no matter how 

 crooked and obstructed it may be, than to hew out a new one 

 and provide a broader way for the future. 



Objection will doubtless be made that the system, on account 

 of the definite quantitative character of its divisions, throws 

 a greater amount of mental responsibility in classifying upon 

 the one who uses it. Thus, for instance, at the very outset, it will 

 often happen that megascopically a rock will be so close to one 

 of the lines separating two different classes that it will be diffi- 

 cult to know to which one it should be assigned. And it will 

 sometimes happen that later chemical or microscopical investi- 

 gation will show the preliminary classification to have been 

 wrongly made. Moreover, when such examinations have been 

 made, it will often occur that it is largely a matter of judgment 

 as to which of two adjacent Orders, Rangs, or Grads a rock 

 properly belongs. 



To some this mental responsibility is distasteful and unsatis- 

 factory; they desire a classification made up of a number of 

 simple divisions like neat pigeon-holes or compartments into 



