roTH] HAMMOCKS 395 
tious than to insert it behind the warps thread by thread, and hence 
St. Clair’s statement of half a century later becomes quite intelligible. 
“With the huge looms of the old-time Arawak on the Corentyn some 
7 feet long and 8 or 9 wide, with similar warp, they would begin the 
weaving by letting off, or rather taking off, the first row of threads 
and passing through it a roller composed of a piece of stick with 
thread upon it, by way of a shuttle, which is then knocked down by a 
heavy piece of hardwood. This is done backward and forward until 
the work is finished ” (StC, 1,295). “The heavy piece of hardwood ” 
referred to here is again undoubtedly the presser. To weave these 
huge hammocks on such primitive lines must have required a large 
amount of patience and time. Indeed, the last-mentioned author 
goes so far as to say while some of the cleverest worked them beau- 
tifully in openwork patterns, in either way, it took months to finish 
one (StC, 1, 295). The most recent mention of these articles is proba- 
bly the following: “ Hammocks with a very close texture are more 
lasting, but are no longer manufactured by the Kalinya (Carib) of 
the (Surinam) lowlands ” (PEN, 1, 130). 
476. When the body of the cotton hammock is completed on one or 
other of the lines laid down in the preceding section (pl. 129 A), the 
projecting ends of the weft and the terminal loops of the warp may 
undergo certain modifications, even additions, for purposes of orna- 
ment or utility. Thus, running along the sides of the hammock 
may be more or less of a tassel or fringe work, etc. (pl. 129 B), on 
occasion even decorated with feathers? (pl. 129 C), while at the ends 
the terminal loops may be interwoven in special patterns prior to 
the insertion of the scale lines. It is through secondary modifica- 
tions such as these that it is often possible to pronounce a correct 
opinion as to the local source of origin of the article. 
477. Space must be found now for a description of the Warrau 
hammock woven from the sensoro variety (sec. 60) of ite twine, 
known as the “ purse net” from the peculiar nature of its mesh. It 
appears to be identical with the chinchorro of the Orinoco (sec. 459). 
I can recognize no classificatory relationship between it and any of 
the other hammocks described. On the one hand, it appears to 
be all warp; on the other, the weft, if such it can be called, forms 
‘both warp and shuttle. A description, however (WER, mr), will 
probably make these points clearer. It is woven on the flat at such 
height from the ground as may be convenient. The frame consists 
1 Conspicuous among his treasures (the President’s at Manaos) were some hammocks 
magnificently bordered with featherwork by Indians on the Rio Negro. The patterns 
represented flowers, leaves, and birds, and the colors were most gorgeous ; but such was the 
labor connected with procuring the great variety of brilliant feathers in the first place, 
and that of arranging them afterwards, that the hammocks were of almost fabulous 
value (BL, 394). 
