A REVIEW OF IGNEOUS ROCK CLASSIFICATION 73 



obtained from the salic ratios. It is not clear whether this is 

 due to the rocks belonging to a different series, to a defect in 

 the method of obtaining the symbol, or to defects in the method 

 of calculating the norm. 



It is thought that enough has been adduced to show that 

 the utilisation of the salic ratios throughout the classification, 

 instead of only in Classes I., II., and III., as at present, would 

 at least give better results than the present method, especially 

 in indicating serial relationships between rocks. Moreover it 

 would do away with an awkward asymmetry in the Quantitative 

 Classification as at present constituted. 



It may be objected that as some rocks in perfemane are 

 totally devoid of salic constituents it would be impossible to 

 treat them as suggested in the foregoing discussion. This 

 is certainly true ; it is a defect inherent in a dichotomous 

 method. The difficulty also occurs in the Quantitative Classi- 

 fication in the pure quartz-rocks at the other end of the series. 

 Here it is impossible, or at least inexpedient, to carry the 

 symbol representing the magmatic position of the rock beyond 

 the Order. Thus a quartz-rock from Secucuniland, South 

 Africa, and a quartz-granite from Eskdale, Cumberland, are 

 represented by the symbol I. 1.-. -(see Iddings, Igneous Rocks, 

 vol. ii. p. 31). Similarly at the other end of the series, and 

 under the method of calculation adopted in the foregoing 

 discussion, rocks devoid of salic constituents would simply 

 have the symbol V.-.-.-., indicating the Class in which they 

 fall. The matter is of small practical importance. For example, 

 it would affect only five rocks out of 34 belonging to Class V. 

 (perfemane) in Iddings' tables of rock-analyses — that is, five out 

 of over 2,000, considering the whole series of analyses. 



Suggestions towards a Quantitative Modal Classification 



It will be evident from the foregoing that the writer con- 

 siders the only classification which will meet the immediate 

 needs of petrographers is one that is at once quantitative 

 and modal. A quantitative classification is admittedly purely 

 utilitarian. However constructed, it will probably "correspond 

 to nothing that has occurred in the evolution and differentiation 

 of igneous rocks," but it is none the less necessary for their 

 comparative study. Petrographers will always require a pigeon- 

 holing arrangement whereby they may accurately docket rocks 



