298 REPORT OF NATIONAL MUSEUM, 1884. 



lineal descendant of the rude wicker fisli-trap. Imagine a number of 

 stakes driven into the ground pretty close together. A horizontal pole 

 is laid against them in the rear, aud by the wrappings of a withe 

 around the pole and each upright stake diagonally on the outside and 

 vertically on the inside, a spiral fastening is produced (Fig. 23). This 

 stitch crosses the two fundamentals in front at an angle and the hori- 

 zontal frame-piece in the rear at right angles, or vice versa, and the 

 lacing may always run in the same direction, or the alternate rows of 

 lacing may run in opposite directions, as in Fig. 23. As a matter of 

 fact, in soft and pliable material this operation constantly pushes the 

 uprights forward a little, giving to the fabric an appearance of the 

 back of a watch (Figs. 24-26). 



The Clallam Indians of the Selish stock make a carrying basket in 

 this manner (Figs. 22, 23), the frame (warp and woof) sticks being about 

 one-eighth inch in diameter, lashed in place with split ozier or root. 

 The Japanese also make a fish -trap similarly, with the exception that 

 the coiled splint passes alternately backward and forward, so that if the 

 horizontal were pulled out the fabric would tumble to pieces. The ob- 

 long oval shields of bamboo, made by the Bateke negroes of the Lower 

 Congo, imitate this structure exactly. The frame of the shield is an 

 oblong hoop on which are stretched splints of rattan, running longitudi- 

 nally on one side and transversely on the other, crossing at right angles 

 except at the plano-convex space at the ends. Splints of bamboo, about 

 one-eighth inch wide, are woven into these cross strands precisely after 

 the manner of the Makali basketry, the consequence being a series of 

 square stitches on tlie back and diagonal stitches on the front, closely 

 fitting, and coving the surface completely. Now, if the frame were cedar- 

 bark threads about the size of pack threads, and the lashing of white 

 sea-grass, we would have the Makah basket (Figs. 24-26). It takes 

 three sets of threads (Fig. 25), the radiated warp, the coiled woof, and 

 the spiral-biudiug thread, to finish the compound. No other area is 

 known to the writer where this peculiar pattern is wrought icto deli- 

 cate fabrics. The Makahs belong to the Nutka stock, most of which are 

 on the southwest shore of Vancouver Island, including the great group 

 of Aht tribes. No Aht basketry is in the Museum, but it would be ex- 

 tremely interesting to trace this unique method of basket-weaving 

 among all the tribes of the stock. Bands of serrate patterns are pro- 

 duced in color by using different wrapping threads, the principal one 

 being grass dyed black in mud. 



There is one specimen of the cedar-bark mat from Vancouver Island 

 in which the shreded bark which serves for warp is fastened at inter- 

 vals of about an inch by a chain-stitch instead of the twine. This must 

 have been a modern innovation; at least there is not another evidence 

 in this collection of savage acquaintance with the chain-stitch. 



The Clallams, adjoining the Makahs, but of the Selish or Flathead 

 stock, in addition to the fish-trap or bower style, are the first going 



