BEACHES OF LAKE WARREN. 759 



new facts, and this can be clone best, as it seems to the writer, bj' subdivision and 

 restriction in nomenclature, as is sometimes done in the biological sciences. The 

 whole series of lalces here described might be called the Warren lakes. This would 

 be one way of preserving Spencer's nomenclature, but in the writer's opinion this 

 use of the name would be unfortunate. A collective name ought to have some geo- 

 graphic significance. The name Erie-Huron, which is used here, serves this purpose 

 admirably, and the name Lake Warren may then be applied in a more restricted 

 waj' to that one of the several separate lakes of the series which most closely corre- 

 sponds to Spencer's original idea. The Forest beach marks the widest extent of the 

 Erie-Huron glacial waters, and was the last and most extensive lake of the series. 

 It seems more appropriate, therefore, to call this stage Lake Warren than to apply 

 this name to ahj' of the higher, less extensive stages. 



The I^ake Warren beaches as here discussed and as mapped iu 

 PI. XXVI, inchide the whole of a complex series which occupy levels 

 about 50 to 75 feet below the Belmore beach. The members of this series 

 are more distinct from Cleveland eastward than around the western end 

 of the Erie Basin. Wind-drifted sand greatly confuses the shore features 

 in the latter reg'ion. 



Perhaps the upper ridge, to which Spencer applied the name Arkoha 

 beach, is sufficiently distinct from those below it to justify separating it from 

 the Lake Warren series. Taylor has considered it the product of a distinct 

 btit transient lake, and does not include it in this series. In southern Mich- 

 igan it lies along the outer border of the same belt of sand whicli in Ohio 

 constitutes Gilbert's fourth beach, but being composed usually of gravelly 

 material, it may be distinguished from the belt of clear sand which it borders. 

 Its gravelly ingredient seems to disappear in northwestern Ohio, so that it 

 can not well be separated from the remainder of the sand belt. The Wood- 

 land avenue beach at Cleveland seems likely to be a continuation of the 

 Arkona beach, and the ujjper member of the series from Cleveland eastward 

 should probably also be thus considered. The upper limits of sand ridges 

 in the district south of the western end of Lake Erie is also about at the level 

 of the Arkona beach. 



The writer has examined this series of beaches at but few points in 

 Michigan, and as they are soon to receive further study in that State, the 

 description will extend no farther north than the Ohio-Michigan line. Tlie 

 only detailed work which the writer has attempted on these beaches is in 

 the district between the Maumee and Vermilion rivers in northwestern 

 Ohio and in Erie and Genesee counties, N. Y. The portion north of the 

 Maumee had been mapped by Gilbert, and the })()rti( m between the Vermilion 



