2 8o SCIENCE PROGRESS 



we must not be content to label the rock with the name of 

 one of our present groups, or enumerate the minerals which are 

 present and the characters they present. We must make a 

 quantitative determination of the relative amounts of the 

 minerals, and an optical determination of their composition, so 

 far as the present state of our knowledge permits, and must 

 finally, from these data, calculate the chemical composition of the 

 rock, verifying it whenever circumstances permit by a chemical 

 analysis. Even where a glassy magma exists we may be able 

 to form some idea of its composition by the specific gravity 

 and refractive index, assisted to a limited extent by the colour, 

 magnetic susceptibility in strong fields, fusibility, and the simpler 

 chemical characters. 



Wherever the chemical composition of a rock has been 

 determined, a careful comparison should be made with other 

 rocks having a similar composition, and the differences that 

 present themselves in mineral composition and structure care- 

 fully studied and if possible explained. 



But it is not only quantitative determination of the minerals 

 that is important. An estimate, as close as circumstances permit, 

 should only be made of the amount in which the different rock 

 types are developed in each district, and their relations with 

 one another in space should be carefully studied. 



By these means workers in the field and with the microscope 

 may co-operate with those engaged in experimental research 

 on the fusion, differentiation, and solidification of silicates, in 

 tracing out the true history of igneous rocks and the physical 

 principles in accordance with which they have come into 

 existence. It will only be when these aims have been at 

 length accomplished that we shall be in a position to frame a 

 truly scientific and universally acceptable system of classifica- 

 tion of igneous rocks. 



