168 BARBARA TSAKIRgIS
although some weights were unintentionally preserved when they were baked
hard in the conflagration that destroyed the Phrygian weaving complex. At the
Hellenistic site of Euesperides in Libya, the loom weights were also fashioned
from unfired clay, but unlike the earlier weights at Gordion, the latter were
formed in imitation of the conical terracotta loom weights used at contem-
poraneous sites in mainland Greece (Megias and Wilson 2008 : 52). At several
sites in Greece, a few metal weights appear in publications, and we must ask
whether clay weights were also used in the houses there, but not recorded by
the archaeologists, or whether metal weights were routinely used and some
were later melted down so that the valuable metal might be used for other
purposes.^4 The practice of recycling, of both metal and clay, is amply attested
in Greek houses, the first material reused because of its value and the second
because of its abundance.
Spinning was an essential first step in textile production and the stone and
terracotta spindle whorls used to increase the force of the rotation of the
spindle must be included in any consideration of domestic textile manufac-
ture. Because the whorls are fewer in number than the weights, they are often
overlooked in discussions of household textile production, and some putative
whorls might have been large beads rather than implements for textile produc-
tion. As most spindles were wooden, few survive. Whorls, too, were sometimes
fashioned of wood.^5 Terracotta covers for the thigh (epinetra) were used for
preparing roves of wool before the spinning began (Barber 1991 : 45), but they
are not represented in large number in domestic contexts; rather, they are best
known as dedications in sanctuaries or as grave goods. The top face of the epi-
netron, often decoratively incised with a scale pattern, provided a rough surface
for separating debris from the fibers and beginning the process of spinning the
rove of fiber into thread (Figure 7.1).
Heinrich ( 2006 ) has explained the seeming paucity of epinetra by suggesting
that leather or rough cloth might often have served instead of the fragile or
expensive objects to provide the necessary roughness for the task.
The mechanics of spinning and weaving on the warp-weighted loom are
well known. While spinning could be done almost anywhere, even when the
spinner was walking, weaving could be accomplished only by standing in front
of the loom for a very long time; the common denominator of both tasks was
considerable time spent doing each. Carr ( 2000 :) proposed that the experi-
enced Greek spinner made 100 meters of thread an hour, but experiments
conducted by Mårtensson, Nosch, and Strand ( 2009 ) resulted in somewhat
slower rates of spinning thread, the speed determined by the weight of the
whorl. Thus a whorl weighing four grams allowed 35 meters of thread to be
produced in an hour, 8 grams 40 meters, and 18 grams 50 meters. Warmer gar-
ments could be woven of two-ply rather than single-ply thread, and thus the
thread for them would take longer to make; plied thread was also a practical