J. A. Thomson — Rocks of Western Australia. 213 



evidence of their internal structure, except, perhaps, the coarse or 

 fine texture. There are only a few garnetiferous amphibolites in 

 the Phillips River Goldfiehl for which a possible sedimentary origin 

 can be admitted. An examination of the jaspei's in the contact- 

 altered schists proves that they too are contact-altered, and are 

 therefore anterior in formation to the intrusion of the granite. 



The granites and porphyries have not as yet been examined in 

 great detail, but it appears possible to group the former into biotite- 

 microcline granites, hornblende-granites or tonalites, and albite- 

 granites. Each of these has its characteristic dyke accompaniment. 

 Many of the basic dykes consist of very fresh dolerites and 

 quartz-dolerites, and show so many points of similarity to the Karroo 

 dolerites of South Africa and the Mesozoic dolerites of Tasmania, 

 that it is not improbable that they are all of the same age. This 

 view has been elaborated in the paper read before the Royal Society 

 of New South Wales referred to above. Others are in a state of 

 alteration intermediate between these and the amphibolites of the 

 auriferous series. 



The jS'ullagine volcanic rocks are entirely difPerent in their state of 

 alteration from the rocks of the auriferous series. They are, as far as 

 examined, practically structureless aggregates of chlorite, epidote 

 calcite, and quartz. Some of the rocks intrusive into the Devonian of 

 Kimberley, however, appear to present a greater degree of constructive 

 metamorphism. One of them is an epidosite containing an amphibole 

 related to glaucophane. 



Y. Principles of Classification. 



The requisites for an entirely satisfactory classification of a rock 

 are a knowledge of — 



1. Its mode of origin, whether sedimentary, volcanic, or intrusive. 



2. Its age. 



3. Its original lithological character. 



4. Its state of alteration. 



If the age is known, the rest is largely a matter of petrographical 

 and chemical study. As far as possible the evidence for the age 

 of each of the main groups has been discussed above, hut the 

 principles involved will be brought together here for greater clearness. 

 It is not necessary to discuss the ordinary principles of stratigraphy 

 by which the aid of sedimentary rocks is determined. The age of 

 a given intrusive rock is fixed within certain limits — 



1. If it can be shown to be intrusive into any rock of known age. 

 Too great importance cannot be laid on the necessity of obtaining 

 evidence of contact-alteration in the surrounding rocks when the 

 intrusive nature of any rock is in doubt. 



2. If it is in turn traversed by intrusive rocks of known age. 



3. If it is covered un conformably by a sedimentary rock of known 

 age. To be entirely satisfactory this evidence should be supplemented 

 hy an absence of contact-metamorphism in the overlying rocks, and by 

 the presence of pebbles of the intrusive rock in the conglomerates of 

 the younger series. 



4. If it can be shown that its date of intrusion was before, within. 



