222 THOMAS H. CLARK 



evidence of muddy waters, in which the coral fauna was extin- 

 guished. Other forms, more hardy, or less susceptible to sHght 

 changes in their habitat, could migrate with the retreating sea. 

 In Indiana, Ohio, and Michigan clear-water conditions prevailed 

 throughout the Middle Devonian, so that the presence there of a 

 somewhat restricted coral fauna above the Onondaga is not surpris- 

 ing. It is supposed that S. lucasensis took part in this westward 

 migration, reaching the northwestern part of the Ohio Basin in 

 Hamilton time. 



EXTENT OF THE ONONDAGA SEA 



The association of typical Onondaga corals with a brachiopod 

 found in the Hamilton of Ohio suggests one minor readjustment 

 of the paleogeographic maps of the continent in Onondaga time. 

 Spirifer lucasensis, originating in the East, migrated to Ohio by 

 some path no trace of which has to date been reported. In spite 

 of the intensive search for fossils in the Devonian rocks of New York 

 State, this Spirifer has not been found. It seems probable, then, 

 that this fossil did not follow along the line indicated by the present 

 outcrops of Onondaga in New York from the Helderbergs west- 

 ward. A glance at Schuchert's paleogeographic map for the late 

 Onondaga, or at the map for Onondaga time in Grabau's Text-Book 

 of Geology, shows a land area bordered approximately by Lake 

 Huron on the west, the Hudson-Champlain line on the east, the 

 present Onondaga outcrops on the south, and unlimited to the 

 north. Inasmuch as all of this area south of the forty-fifth parallel 

 was certainly covered by Ordovician sediments, which probably 

 extended as far north as Lake Temiskaming, it does not seem im- 

 proper to suppose that the Onondaga sea also covered much of it. 

 If the strand line be drawn from Montreal westward along the 

 forty-fifth parallel to Lake Huron, we shall have a much more 

 probable distribution of land and sea than the maps show. Coral 

 faunas require a wide, open seas, rather than narrow channels. It 

 is a question how close to an outcrop of Onondaga limestone one 

 can draw the strand fine. In this case, the cartographer may be 

 allowed more latitude than when dealing with detrital rocks. Such 

 a boundary as suggested would remove from the present maps the 



