90 
It is not known for what purpose the carbonized, button-like wooden 
object seen in Plate XVIII, figure 4, was used, but it may have been used 
in a game. The exact use of the perforated wooden disk m Plate XVII, 
figure 11, also is unknown. It is slightly concavo-convex and about re mc h 
thick; the hole is biconical. 
Certain articles, including a piece of textile, a fragment of rope, pieces 
of birch bark, and a wooden dish will also be described under this head. 
The piece of textile seen in Plate XVIII, figure 6, which is probably 
part of a bag, was found in a mixture of ashes and carbonized organic 
matter containing bits of bark (probably pine) , 21 inches, deep, in refuse 
deposit 15. It is entirely carbonized, which accounts for its preservation. 
This may be only a small portion of the original fabric, but there is sufficient 
to enable one to get an idea of the technique; even part of the border or 
selvage remains. 
The fibres are about as coarse as the jute used in the manufacture oi 
binder twine and are probably derived from the inner bark of the basswood 
rather than from the milk-weed or hemp nettle, both of which produce a 
finer fibre. 
The style of weaving seen in this fabric is what is known as twined, but 
instead of the warp elements consisting of only a single strand, as in othei 
pre-European fabrics illustrated by Holmes (1, Figures 5, 8, 20, 21) , they 
consist of two-ply cords, and the weft consists of four sepaiate strands. 
The warp consists of a series of parallel cords, probably arianged in the 
same way as the warp on a loom, with each cord bent at the end that was 
to form the top or border {See Plate XVIII, figure 8A) and the two loose 
ends tied together at what was to be the bottom of the bag. The cords are 
about i inch thick, loosely but evenly twisted from two strands, each of 
which is about yg inch thick. The cords are twisted to the light, but the 
component strands appear to have been twisted to the left The weft 
elements consist of four separate strands, about yo inch thick, which are 
unevenly twisted, apparently to the left. Each pair of the weft stiands is 
twisted half around on the other pair, enclosing each warp element separ- 
ately at the intersections {See Plate XVIII, figure 8B). The spacing of 
both w r arp and weft elements is regular, the former being about ^ men 
and the latter about inch apart (See reconstruction in Plate Xv III, 
figure 7) . The border or selvage is about i- inch deep and was formed by 
closely impacting the three upper rows of w'eft elements. 
The fact that the weave of this fabric is so open suggests that it is 
part of a pouch or bag rather than of some wearing apparel. It is known 
that the early Iroquois used bags and pouches; Megalopensis (page 158) , 
for instance, speaking of the Maquas (Mohawks) , mentions a Bag wdnch 
thev make of wild Hemp,” in which they kept their dried fish. 
' The discovery of the fabric suggests that other kinds of woven articles 
were used at this site. Cartier (page 163) tells us that the Hochelagans 
used “ four-cornered ” mats, “ woven like tapestry.” When we consider 
that the culture of Hochelaga is practically the same as that of the Roebuck 
site, it is altogether probable that mats were used here also. 
