Pap. No!"2lT' JOHN H. KERR RESERVOIR BASIN — MILLER 179 



the Hamilton Focus, wliile Harrington (1922) reported finding the 

 same type in sites on the upper Temiessee River. Grifiin (1938) de- 

 scribes this ware as "Type III" from cave sites in the Norris Basin, 

 Alabama, and (1939) as "fabric plain plaited" in the Wheeler Basin, 

 Alabama. Haag (1939) described this type under "Long Branch 

 Fabric Marked," also a limestone-tempered ware, as associated with 

 the Copena complex. 



Guernsey and Kidder (1921) describe "Twined weaving" in bags 

 most competently. They state: 



Tlie bags are flexible seamless sacks with full, round bodies and long, grad- 

 ually constricted necks. They range from 1^2 inches to 2 feet or more in 

 length. All are made in the same way, of close twined weaving; the majority 

 of specimens have both warp and weft of two ply apocynum string, though some 

 have yucca warp and apocynum weft. The combination of apocynum warp and 

 yucca weft is rare. (By no means is the present writer insinuating that these 

 fibers were used by the aborigines of southern Virginia and northern North 

 Carolina.) 



Our study of the weave was begun by examining the bottoms of the bags in 

 order to make out how the preliminary "set-up" of the warp cords was accom- 

 plished. By dissecting several fragmentary specimens we found that there 

 were two methods, one common, the other rare. The former was as follows: 

 six long strands were laid across each other, three above and three below ; the 

 middle strand of each set runs out straight, the others are bent so that their 

 ends radiate from the common center. There are thus produced twelve orig- 

 inal warps. The second method consists of twisting three strands about each 

 other and then bending their ends so that they radiate and form six warp cords. 



The above systems are very simple and practical and avoid the ugly lump and 

 the potential weakness in the fabric which would have been the result of knotting 

 the warps together at the base. The method of inserting the weft also obviates 

 knotting ; a single long string is worked over and under the radiating warp cords 

 close about their common center ; in reality, it is pulled up very tight and holds 

 the warp firmly together. When a circuit of the spoke-like warps has been made, 

 the two ends of the weft string of course come together ; they are then combined 

 into a single strand of twined weaving, which continues spirally around and 

 around to form the body of the bag fabric. . . . (p. 66). 



We have not mentioned the actual method of inserting new warps. Two ways 

 were employed. In one the string to be added was looped and laid between two 

 of the old warps thus forming two new ones ; the first two or three turns of the 

 weft attach the new strands to the old warps on either side of them holding 

 all firmly in place; the next turn of the weft takes in each new element sepa- 

 rately and the weaving continues normally. 



In the second method, the strand to be added was doubled into a loop, mak- 

 ing, as before, two new warps ; the string at the bend of the loop was twisted 

 apart into its two component plies and one of the old warps was threaded 

 through the resultant opening; the loop was then slid up the old warp and 

 brought close against the last woven turn of the weft, thus producing a pair 

 of new warps one on each side of the original one ; on its next revolution about 

 the bag the weft takes in the two new warps and holds them solidly. . . . 

 (p. 68). 



We now take up the twining of the weft, which is perfectly simple and regu- 

 lar. It begins at the very bottom and continues in a close spiral to the mouth. 



