TiioMAs.] THE "OHIO DISTKICT. 563 



closure shown iu N». 3, I'l. xi, "Aucient MomimcTits," where thei-e are 

 two walls s(une distance apart, is probably an instance where the in- 

 closed area was contracted subsequent to its first occupancy. This i,s 

 inferred from the fact that tlie inner wall passes over a mound (j>robabiy 

 natural) which it seems would have been avoided iu the original plan 

 when the line might have beea fixed wherever desired. It is also pi'ob- 

 able that some of the short walls about the chief gateway of "Forti- 

 fied Hill," Butler county, shown in I'l. VI, "Ancient Monuments," are 

 additions to the original plan. The reasons for this supposition become, 

 more apparent upon a personalexainination of the tapographical featui'es 

 of the locality. The easy appr oach at this gateway renders it the weak 

 point of the fort. 



The work with the double wall above mentioned and some others in 

 central and southwestern Oliio, similar in character, except that they 

 usually consist of a single wall and ditch, belong to a slightly difterent 

 type than the "hill forts" referred to, and constitute the third class 

 above mentioned. They occupy peninsulas iblnied by the bends of the 

 iivers or larger streams, and the headlands at the junctions of deep 

 ravines or uniting streams, and consist of a curved or irregular line of 

 wall and ditch across the isthmus, or running from one bank to the 

 other, the precipitous descents at the sides and around the point forming, 

 the defenses at these points of the area. Some of these works bear a 

 strong resemblance to the works iu the northern part of Ohio, which 

 we have ascribed to the tribes of the llurou-Iroquois district, and, pos- 

 sibly, may be due iu part to-some tribes of this fiunily. It is to be 

 observed that a few of the same type occur in Kentucky. They may 

 have been erected by the people who built the more elaborate hill forts, 

 yet I am rather inclined to doubt this. However, as they are simi>le in 

 character and are found at widely separate points in the moiuid area* 

 it is unsafe to consider them of ethnic sigifificance without further 

 liroof. We will probably be justified in assuming that they can not be 

 ascribed to the builders of the geometrical works which constitute our 

 second class of inclosurcs. • •' 



• The works of this second clitss are usually more or less regular iu 

 form and are located on level areas, generally in the river valleys. 

 The leading types of this cl.ass are the well known circles, squares, and 

 octagons, or, as they are collectively termed, " geometri(!al works," tbuud 

 chiefiy in the Scioto valley and about i!^ewark, Ohio. Others which 

 may be included in the class, though varying from the typical forms, 

 are found in the Kanawha valley. West Virginia. Most of those of 

 southern and central Ohio are figured with comparative accuracy by 

 Messrs. Squier aud Davis iu "xVucient Monuments." But when deduc- 

 tions are to be drawn depending upou accixrate measurements, it will 

 be necessary to take into- consideration certain errors in th«ir work 

 pointed out by the present author iu a bulletin entitled "The ttirculai-, 

 Square, and Octagonal Earthworks of Ohio,". published by the Bureau 



