I gg BUREAU OF AMERICAN ETHNOLOGY [bull. 118 



firmly attached to the main body of the seat. It was harder than the 

 interior and had a surface which took on a high polish as a result of 

 use. So hard was this surface that it may have been subjected to fire 

 hardening in some cases. However, there was no tempering material 

 in the matrix, and in cases where fire discolorations were observed, the 

 seat had been under a burned structure which had collapsed upon it. 

 It is possible that no agency for hardening the surface was used except 

 kneading of the clay while soft, and continued drying. If in this dry- 

 ing any cracks developed they were filled ; for none were found in the 

 surface when the seats were first exposed. When they were uncovered, 

 however, they absorbed moisture readily. Most of this work of explo- 

 ration was done in midwinter, so that soon after exposure these seats 

 would be exposed to rain, snow, and often to freezing temperature. 

 After two or three wettings the surface would check and deep cracks 

 would appear in the body of the material. After freezing, crumbling 

 soon began. Always the surface would flake off, which seems definitely 

 to suggest that it had been plastered on to the main body and was not 

 thoroughly attached. The interior portions of the clay body did not 

 scale off in the same way. 



In plate 128, a, is shown the seat in the primary floor of Bowman 

 Mound No. 1, Site No. 2. The top step of this seat was slightly dam- 

 aged by workmen before its presence was discovered. A portion of 

 the top step has been restored. This seat had really three steps. The 

 lower one shown in the photograph was a clay platform some 3 inches 

 thick which extended to both front and rear of the seat, as shown in 

 figure 4. Plate 128, &, is the seat on Mound No. 2, Site No. 6, Harvey 

 Hill farm. The seat on the primary floor of Structure No. 2 on the 

 Caryville Site No. 5, shown in plate 129, a, is exceptional in that the 

 lower step has a semicircular outline instead of a rectangular form, 

 which is the usual method of construction. It was built so closely 

 against the wall of the town house that the logs came up through the 

 back of the seat, as shown in plate 129, &, which presents the longi- 

 tudinal section of these post molds. The seat on the secondary floor of 

 Mound No. 2 on the Harris farm, Site No. 9, which is shown in plate 

 130, a, is notable in that the surfaces of both steps are slightly concave. 



Plate 130, Z>, shows a seat being uncovered. The hard surface of the 

 seat separates readily from the superincumbent earth. There is, there- 

 fore, little danger that the surface of the seat will be marred during 

 excavation, unless it be because of carelessness of unskilled and un- 

 trained workmen. Plate 131, a, shows the seat on the primary floor of 

 Mound No. 2 of the Harris farm, Site No. 9. This picture was taken 

 after the seat had been subjected to several rains and freezing at night. 

 The surface was flaking off and large cracks were developing in the 

 main body. The large post mold at the left-hand corner of this seat 



