74 BULLETIN OF THE 



ters are greatly or entirely changed, and the case is different. We can 

 only proceed in safety when we know their field relations, or those of 

 rocks that are like them. In order to know the microscopic characters 

 of a sedimentary granite (if such a rock exists), it is necessary to study 

 one that is known beyond question to be sedimentary, and to compare 

 it with the most highly altered eruptive ones known. By this method 

 of proceeding, always taking as the basis rocks whose relations to their 

 fellows were known, diagnostic points. of great value would be found, 

 and a more just idea of the relations of the fragmental and non-frag- 

 mental forms obtained. The field evidence would have to be the 

 arbiter in all cases. If a rock is eruptive, or if it is sedimentary, it 

 is so, whatever may be its microscopic characters. In this class of 

 rocks lithology is at present weak, and assumptions take the place 

 of facts. 



Although in no case in this paper have we attempted to give any 

 elaborate microscopic analyses, but only a few of the more general facts, 

 we have shown plainly enough the fallacy of determining the geological 

 relations of highly altered rocks by microscopic analysis alone. Any 

 one who takes pains to read pages 533 to 599, inclusive, of the third 

 volume of the Geology of Wisconsin, will doubtless be convinced that 

 difi^erent observers, under the present method of microscopic analysis 

 and classification, reach widely different results in the study even of the 

 same specimens. He can further compare our description of the Picnic 

 point rocks with that given on page 567, and also with Mr. Brooks's state- 

 ment that granite dikes " have never been observed in the Marquette 

 series " (I. c, p. 452). It is of course worthy of remark, that of the litholo- 

 gists who have mici-oscopically studied the rocks of Lake Superior, Messrs. 

 Pumpelly, Wright, Irving, Hawes, Rutley, Julien, Turnebohm, and Wich- 

 mann, not one, so fiir as we can learn, had at that time any especial per- 

 sonal acquaintance with the characters of modern unaltered eruptive 

 rocks, except possibly the two last named. Yet it is essential that one 

 should be thoroughly acquainted with the unaltered forms of rocks be- 

 fore attempting to solve the most difficult lithological problems on the 

 globe, i. e. those relating to the altered forms. As descriptions of what 

 these observers saw in the individual specimens examined by them, their 

 ■work is undoubtedly of great value ; but beyond that there is a great 

 chance for differences of opinion concerning their conclusions, or the 

 basis upon which these were established. 



The sedimentary rocks of the Marquette district generally give evi- 

 dence of being shore deposits, and although they may have been deeply 



