308 THWAITES AND TWENHOFEL THE WINDROW FORMATION 



assumed with considerable confidence to be of Pliocene age, since it bears 

 gravels which belong to the Pliocene Lafayette formation-' (13). Trow- 

 bridge appears to hold a similar view (13, 23, 24, 26, 27, 28). Chamberlin 

 and Salisbury (10) discussed these gravels under the Pliocene, but sug- 

 gested that they may be of older age. 



CRITERIA RELATING TO CORRELATION 



There are two problems involved, the correlation of the occurrences 

 with each other and their correlation with deposits of known age. In the 

 discussion which follows the two problems are not separated. Criteria 

 which bear on the age of the gravels of the Windrow formation may be 

 divided into six groups: (1) lithological similarity, (2) topographical 

 position, (3) fossils, (4) age of underlying formations, (5) relation to 

 overljdng formations, and (6) history since the advent of glaciation. 



LITHOLOGICAL SIMILARITY 



The correlation of the different occurrences of the Windrow formation 

 with one another is in large part based on their lithological similarity- 

 Such differences as exist are quantitative, not qualitative. This fact, 

 taken in connection with other evidence, leads the writers to the conclu- 

 sion that the Windrow deposits as far west as Mitchell County, Iowa, are 

 of the same age.^ The gap of 120 miles from Floyd County to Guthrie 

 County, Iowa, is more difficult to bridge. There seems little doubt that 

 the Guthrie County beds are Cretaceous, and could a correlation of these 

 basal Cretaceous gravels with those of the Windrow formation be defi- 

 nitely established, the age of the latter would be determined. The corre- 

 lation of the Windrow formation with the high-level gravels of Illinois is 

 even more difficult and hesitation is felt in attempting such. Whether or 

 not the upland gravels of Buffalo County, Wisconsin, belong to the Wind- 

 row formation is an open question. They are much less like it than is 

 the Cretaceous of southwestern Iowa, but, on the other hand, they are not 

 like any known drift in that they lack igneous rocks. The larger stones 

 are possibly all of local origin and none found is striated. It seems im- 

 probable that all the igneous rocks could have been destroyed by weather- 

 ing, for they are present in what has generally been considered the oldest 

 known drift, the Nebraskan. It is certain that these gravels antedate the 

 valleys, and so are presumably of pre-Pleistocene age. 



^ In this connection it should he stated tliat the writers do not assume the gravels 

 were necessarily deposited within a period of time represented by any marine formation. 

 An entire period may have been involved and some portions of the gravels may well be 

 somewhat older than others, hut they are believed to have been deposited within a space 

 of time throughout which there were maintained the same general conditions of deposi- 

 tion over the area of distribution. 



