MINERALOGY AND PETROLOGT. 987 



The oldest known rocks.] 



and such resultant rock would be uniformly of about the same normal character as 

 the Archean or Taconic diabases. On the other hand all later granites would show 

 great initial variations. If they originate from the Archean they must show alliance 

 with the Archean granites, and with their variations. If paleozoic sediments be 

 fused, or even later rocks, great departures would ensue from the types of Archean 

 granites, and these variations would be emphasized by the greater susceptibility of 

 acid rocks to endomorphism, combined with their greater viscosity. The alkalino- 

 acid element is easily transfused either into (No. 552) or from (Nos. IB, 5, 648) a 

 molten rock. The ferro-magnesian magma is chemically more stable in contact on 

 acid rocks, but as a molten mass it is more fluid. 



The so-called " peripheral phases " of the gabbro are peripheral only by accident. 

 They might occur, and do, at other places in the great mass. The granulitic structure 

 occurs about Muscovado lake, well within the gabbro. The concentration of iron is 

 well known in the coarsely crystalline gabbro, making large deposits. The ferro- 

 magnesian phase in the form of highly magnetited gabbro is the common feature 

 at Mayhew lake and at Frazer lake (No. 1041), while olivine (fayalite) and magnetite 

 sometimes compose important masses (Nos. 1336, 1843). These " phases" have been 

 much noted and studied, but they seem to be phases both of the Keewatin green^ 

 stone and of the Taconic gabbro, representing an intermediate state of recrys- 

 tallization. 



THE OLDEST KNOWN ROCKS. 



In Part I of this volume it is stated that the oldest rocks of the state are the 

 Keewatin greenstones of the Kawishiwin, and in the preface is a diagrammatic scheme 

 of the structural relations of the Archean as presented in this report. 



It is only within a few years that American geologists have entered seriously 

 upon the attempt to subdivide the rocks of the Archean upon a chronological scale. 

 Mr. A. C. Lawson divided the sedimentary Archean of the Rainy Lake region into 

 Coutchiching and Keewatin, the same year in which the writer divided it into Ver- 

 milion and Keewatin. In each case it was recognized that there was a later granite 

 which is intruded into these rocks, and Mr. Lawson dwelt on the importance and the 

 significance of this later intrusion. He showed that, either as massive granite or 

 granite-gneiss, this later rock spreads over very extensive areas, and constitutes a large 

 part of the mass which, under the term Lower Laurentian, embraces both metamor- 

 phosed sediments and intrusive granite. He restricted the term Laurentian, how- 

 ever, to this igneous rock which, while younger by reason of intrusion, yet structu- 

 rally in many places lies below the sedimentary rocks of the Archean. The sedi- 

 mentary division of the Archean was by Lawson designated Ontarian, this term 

 covering both Coutchiching and Keewatin. The earlier division of the " Laurentian," 



