REVISED PRINCIPLES OF CORRELATION 477 



Beferring to another general idea, it seems no less certain that circum- 

 stances must have often arisen that tended to exclude certain species, or 

 even whole faunas, from continental basins to which, under preceding or 

 succeeding more favorable conditions, they enjoyed free access. These 

 changes in physical conditions were responsible for two facts that are 

 exceedingly important in correlation by faunas : ( 1 ) That most provinces 

 were at times invaded by waters of different oceanic basins, each at such 

 times bringing with it the fauna peculiar to its own realm, the others 

 being excluded; (2) they occasioned the recurrences of species and faunal 

 aggregates — that source of much trouble in the past and now a constant 

 menace in correlation by marine faunas. Instances are multiplying, and 

 some of the best have not been published. I may mention two that have 

 been proved in the past year. Both are earlier occurrences of faunas than 

 had been known before. 



FAUNA OF THE MENDOTA DOLOMITE, AN EARLY OZABKIAN FORMATION 



The first of these recurrences affects the peculiar fauna of the Mendota 

 dolomite of southern Wisconsin. As now known, the true Mendota fauna 

 comprises a total of 28 species; 16 of these are gasteropods, 3 pteropods, 

 6 trilobites, and 3 brachiopods. In general aspect the fauna is decidedly 

 Ozarkian, but with some reminders of Upper Cambrian species. The 

 latter, however, constitute a smaller proportion of the whole than is 

 apparent in the fauna of the succeeding Madison sandstone. 



We are indebted to the efforts of Dr. Samuel Weidman and Mr. T. F. 

 Thwaites, of the University of Wisconsin, for the discovery of a dolomitic 

 ledge in the Saint Lawrence formation that in both its lithologic char- 

 acter and faunal contents closely resembles the true Mendota dolomite. 

 This ledge is developed to the west of Madison in hills bordering the 

 valley of Black Earth Creek, between Black Earth and Mazomanie, Wis- 

 consin. It should be mentioned, further, that the bed lies in the middle 

 part of the Saint Lawrence, beneath the Dikelocephalus minnesotensis 

 zone, and that the Saint Lawrence is the second formation beneath the 

 top of the Cambrian, as now defined in the Upper Mississippi Valley. 

 The Jordan sandstone lies between the Saint Lawrence and the over- 

 lapping and consequently varying base of the Ozarkian. 



So far as collected, the fauna of the Black Earth dolomite, as this 

 Cambrian bed may be provisionally designated, consists of 13 species. 

 Ten of these are strikingly like species found more or less abundantly in 

 nearly every exposure of the fossiliferous part of the true Mendota; in 

 fact, the fossils from these two zones are so much alike that if the prob- 

 ability of their distinctness had not been suspected the Black Earth species 



