PETROGRAPHY OF THE KEWEENAWAN 



639 



scopic results, by several authors, but only a few analyses are satis- 

 factory for the application of chemical classification. These ancient 

 basic rocks are so far altered in most places that the original can be 

 only roughly estimated from the present composition. However, it 

 seems clear that alteration results differently in different cases, and 

 the average of several tests of a rock-type will give valuable indica- 

 tions of the quality of the magma. It is therefore worth while t) 



Fig. 3. — Weathered volcanic breccia. About one-half natural size 



extend the chemical work, especially when material is very fresh, and 

 analyses can be made quite complete. 



Clastic Rocks. — The conglomerates of this area have all been 

 described in earlier papers and need no comment except that no 

 correlation of beds has been possible, and that they give evidence 

 that the bowlders were derived from both Keweenawan and older 

 formations. 



Tuff and breccia, of the type found by Dr. Berkey at Taylors 

 Falls, was found on the two tributaries of Tamarack Creek (Fig. 3). 



