47 o SCIENCE PROGRESS. 



from different sides, and we will therefore notice in turn 

 the different factors which, from time to time, have been 

 made to enter into schemes of rock-classification. The 

 development of these schemes shows a progress from a 

 purely artificial towards a natural system, the latter being, 

 however, a goal sought rather than an end attained. The 

 earlier workers seem to have held no conception of a 

 natural system of classification, even as an object to be 

 aimed at. Their view is perhaps not unfairly represented 

 by the opening sentence of Pinkerton's preface in 1803 : 

 " Le seul avantage qui resulte de tout systeme me'thodique 

 en histoire naturelle, est de soulager la memoire ". 1 



The first factor that we shall notice in the various 



schemes of classification of igneous rocks is geological age. 



The rival systems in the early part of this century were 



often referred to as the "geological" and the " mineralogi- 



cal ". They were to some extent identified with Plutonism 



and Neptunism, but the two principles in rock-classification 



may be traced back to an earlier date. Linnaeus, Wallerius, 



and others of the Swedish school in the middle of the last 



century, embraced the rock-types known to them in their 



mineralogical systems of the inorganic kingdom ; but they 



had also concluded, from observations in their own country, 



that the several broad divisions of rocks occupied each its 



own definite position in the crust of the earth. In other 



words, they recognised a stratigraphical succession of rocks, 



which they believed to be of general application. This 



idea is more familiarly associated with the name of Werner, 



who elaborated it into a system based on his researches in 



the mountains of Saxony. In these early speculations, 



which were indeed the beginning of stratigraphical geology, 



the important conception of an ordered succession of rocks 



was pushed so far as to obscure the distinction between 



truly bedded sediments and igneous rock-masses ; while, 



further, the succession observed in a single area was 



assumed to hold good universally. Werner's scheme of 



1 I have met with this earlier work of Pinkerton only in the French 

 translation, Esquisse dune nouve/k classification de mincralogie. 



