654 AV. H. SHERZER RECOGNITION OF TYPES OE SAND GRAINS 



it may be pointed out that there were no crystalline formations capable 

 of supplying other minerals between the present deposits and their sup- 

 posed parent bed. Those materials that could have been thus supplied 

 would have been reduced to powder and removed. Furthermore, after 

 the land had once become mantled with the a3olian formation, it would 

 have been protected from further wind erosion. The reason for the de- 

 posit of iron oxide over the granules of desert sand has not been fully 

 explained. The most plausible explanation is that it results from the 

 deca}^ of iron-bearing minerals associated with the quartz. In the case 

 of the red sands of the Arabian desert described by Phillips (see page 6-1:3 

 of this paper), the sand is apparently stagnant, and there occurs an ex- 

 ceptional amount of this oxide, very probably due to the better oppor- 

 tunity for residual action. With no iron-bearing minerals present in the 

 Sylvania to be decomposed, it is not clear how any oxide could be formed 

 or expected. Just beneath the till covering of the bed the formation is 

 generally thus stained to a depth of a foot or more, but due to perco- 

 lating water from the recent deposit. 



So far as field data are available, the material making up the Sylvania 

 formation seems to have come from the northwestward, and very probably 

 was derived from the breaking down of the Saint Peter (Ordovician) 

 sandstone of eastern Wisconsin,'^ either by wave or wind action, or both, 

 and swept to the southeastward by prevailing northwest winds during 

 mid-Monroan (Silurian) time. This deposit is a very pure (about 

 99 per cent) aggregation of rounded granules of crystalline quartz, very 

 similar in appearance to the Sylvania, but of average coarser texture 

 and containing a greater variety of original material (see figures 1 and 2, 

 plate 47). Like the Sylvania, also, it is typically incoherent, readily 

 crumbling in the fingers, and believed by Berkey to have been itself very 

 largely an geolian formation.'* Owing to this lack of binding material, 

 the formation must have been readily disintegrated, and appears now in 

 Wisconsin in fragmentary condition. To the basal sandstones of the 

 Lake Superior region Berkey looks for the origin of the bulk of the Saint 

 Peter material (page 243), being "washed out by the retreat of the sea 

 and thereby assorted, then worked many times over by the wind in the 

 absence of the sea, and thereby still more perfectly assorted, and finally, 

 in the readvance of the sea, much of it was again worked over a last 



'^^ For a description of this formation the reports of the geological surveys of Wis- 

 consin, Minnesota (vols. 1 and 2), Iowa (vol. 1), and Illinois (vols. 3 and 5) may be 

 consulted. Also Owen's Geological Survey of Wisconsin, Iowa, and Minnesota, 1852, 

 and Eleventh Annual Report of the U. S. Geological Survey, 1892. 



''* Berkey : Paleogeography of Saint Peter time. Bulletin of the Geological Society of 

 America, vol. 17, 1906, p. 246. 



