86 LOWER PENINSULA. 



grained sand rock, with little calcareous cement, and rarely fossilifer- 

 ous. Prof. Winchell named these strata Napoleon group, to distin- 

 guish them from the lower, often very fossiliferous beds, com- 

 posed of more fine-grained micaceous sand rock, with seams of 

 harder ledges rich in calcareous cement, which lower division he 

 names Marshall group. He asserts the two divisions to be always 

 separated by a seam of shale several feet in thickness, but such a 

 regularity in the disposition of the rock beds of this horizon does 

 not exist ; shale beds are found everywhere in alternation with 

 the sand-rock ledges, and in the different exposures of limited 

 vertical extent, it is almost impossible to tell which of the special 

 sand-rock ledges or shale beds we have under observation. 

 Neither is the presence or absence of fossils in the rock beds a 

 feature to be relied upon. A large portion of the lower beds, con- 

 sidered to represent the Marshall group, contains no fossils, yet 

 the absence of fossils at Napoleon is no proof that those particular 

 strata do not contain any. The rock of the quarries at Stony 

 Point is so absolutely similar to the Napoleon sandstone, and is 

 generally so barren of fossils, that nobody would doubt its identity 

 with the other. It is only lately, by the more extensive opening 

 of the quarry, that fossiliferous seams were discovered, which were 

 before not known to exist. 



It has been stated that Prof. Winchell considers the sandstones 

 of Marshall as the lower terminus of the carboniferous rock series, 

 typically distinct by its fossils from the next subjacent shaly beds 

 which he connects with the Devonian rocks by the character of 

 their fauna. Such a difference in the fauna is not perceptible ; the 

 fossils of the Marshall sandstones and the subjacent shales are not 

 only generically in full harmony, but a great number of species 

 are common to both. This lower shale formation is the surface 

 rock in the south part of Hillsdale County and Branch County ; 

 the transition from the upper sandy division to the lower is not 

 defined, as should be expected in rock beds with two distinct 

 faunas ; the beds are in fullest conformity of deposition, and the 

 material composing them does not change. In the upper we have 

 a sand rock with subordinate beds of shale ; in the lower we have 

 the same sort of shales alternating with subordinate beds of the 

 sand rock. The outcrops of the shale formation in the counties 

 mentioned are never of great vertical extent, and their hori- 



