378 H. C. COOKE 



west and southwest of Hudson Bay. What its further extent may 

 have been is not known. The nature of the rocks that composed 

 the plateau and formed the floor on which later rocks were laid 

 down is yet unknown. The first event of which record remains is 

 the extrusion of vast amounts of lavas, which spread in sheets over 

 the greater part of it. Many of these lavas possess pillow struc- 

 tures and are interbanded with beds of tuffaceous sediments. 

 Both of these lines of evidence indicate subaqueous extrusion; 

 so that either the sea covered portions of the plateau at this time 

 or the earher extrusions of lava so disorganized the drainage as to 

 create large lakes, in which ellipsoidal lavas and bedded tuffs were 

 laid down. In the northern Quebec district the oldest lavas found 

 are of basaltic composition. They were succeeded by more acid 

 types as extrusion went on, so that the basalts are overlain suc- 

 cessively by porphyritic basalts and andesites. Between the ande- 

 site flows, beds of chert are occasionally found, which probably 

 represent portions of the load of magmatic waters accompanying the 

 flows. Near or at the top of the andesite flows, beds of coarse 

 tuff are found locally, indicating that the period of andesite extru- 

 sion ended with volcanic explosions. The period of volcanism 

 closed with a number of local and small extrusions of quartz 

 porphyry. In places this porphyry forms a breccia instead of a 

 massive flow. The brecciated texture has been shown to be due 

 porbably to subaqueous extrusion, and thus corresponds to the 

 pillow structure in the more basic lavas. 



At the close of the extrusive period deposition of sediments 

 began. The lowest beds are of a rather basic, tuffaceous composi- 

 tion. These beds soon give place to others which have more the 

 composition of impure sands and are now altered to micaceous 

 schists containing a few garnets. Higher in the series a further 

 change in composition takes place, probably by an increase in the 

 lime content of the beds, which is marked by the appearance of 

 garnets in great numbers. Interbedded with the garnetiferous 

 mica gneisses are beds of garnetiferous hornblende gneiss and an 

 occasional thin bed of crystalline limestone. These sediments, 

 locally termed the Nemenjish series, are correlated with the Gren- 

 ville series. It is as yet uncertain whether they are to be con- 



