116 



other hand, spready by seeds as well as vegetatively. It has 

 widened out from a strip 0.5-1.0 meters wide back of the 

 Jiincus, to large mats 5 to 7 or more meters in diameter, which 

 occupy the sand area between the relic dunes and the grassy 

 sand plain. The maximum development occurs just below the 

 grass. Where it overlaps into the grass the vegetation is 

 noticeably heavier than in either of the two associations. This 

 is aided by a few secondary species, especially Sporoholus cryp- 

 tandrus and Monarda punctata. The grass {Poa compressa) 

 forms a permanent vegetative covering, but the Potcntila dries 

 up during the winter. During the growing season much blow- 

 ing is prevented by the covering of Potentilla, but wSth the 

 removal of this in the fall the blowing of sand commences, 

 and during the winter the amount that is blown away is notice- 

 able. Much of it tends to accumulate southward of this area. 

 From the state line south to Winthrop Harbor the shore 

 dips away from so direct an attack from the waves. Here 

 the J uncus is able to retreat from the shore line, and although 

 relic dunes still are being formed, there is still a relatively wide 

 zone of Juncns behind them which protects the grassy plain 

 from being Washed away. 



T. E. Savage ]:)rcsented the following paper: 



THE GRAND TOWER (ONONDAGA) FORMATION OF 



ILLINOIS, AND ITS RELATION TO THE JEF- 



FERSONVILLE BEDS OF INDIANA. 



The name (irand Tower formation was proposed by 

 Kcyes^ for tlie Devonian strata in southeast Missouri, which 

 were considered the ecjuivalent of the Onondaga and tlie 

 Oriskany of the New York section. Ihe name is taken from 

 the town of Grand Tower, in Jackson county, Illinois. The 

 term is here amended to include only that part of the De- 

 vonian strata in southwest Illinois, and adjacent portions of 

 Missouri, which is the western representative of the Onondaga 



1 Keyes. Mo. Geo. Survey. Vol. Ill, p. I^D. IHiH. 



