106 BUREAU OF AMERICAN ETHNOLOGY [RULL. 65 
patterned bottom of a twined sandal. In our collection there is a 
small fragment of a similar object (A-1385, Ruin 5). It will be 
remembered that there are two types with the offset or jog, one 
plaited of whole leaves (I, a, 2), the other the twined type just 
discussed (II, 6). For the former a pattern would have been con- 
venient in order to measure the proportions of the growing sandal, 
place the jog correctly, etc., but it would not have been necessary. 
For the latter, with its complicated series of concentrically looped 
warps (fig. 38), a measure of some sort must have been aimost indis- 
pensable for the correct preliminary laying out of the strands so 
that they should be of the right lengths and have the proper curves. 
During the weaving in of the weft-strands, also, a standard pattern, 
such as would have been provided by the stones, must have been a 
great help in keeping a check on the proportions of the growing 
fabric. Dr. Snyder’s? theory as to the use of the stones is as follows: 
A long strand of yucca fibre was first wrapped around the stone from one 
end to the other at intervals of half or three-quarters of an inch. This pre- 
liminary step served as a basis for securing the intertexture of prepared ma- 
terial, as the stone served for a guide for shaping the form. Upon these trans- 
verse filaments the work was commenced and prosecuted, first on one side 
of the so-called last and then on the other, by the “over and under ” method 
of broad plaiting. When this process of interweaving the damp and flexible 
strands had been conducted sufficiently far on both sides of the model to fix 
by it the shape and dimensions of the sandal, they were probably cut apart 
and finished separately. 
An examination of sandals of any type, however, shows this idea 
to be erroneous; none have been made in pairs and cut apart, and 
in no case does the warp run otherwise than longitudinally. 
A word as to the distribution of “ sandal stones.” They are com- 
mon in the Mesa Verde ruins, and are figured by Fewkes* and Nor- 
denskiéld,* Morley reports them from the McElmo,* and Professors 
Cummings and Kidder excavated several on Alkali Ridge, San 
Juan County, Utah, in 1908. Mr. Richard Wetherill states’ that 
they occur in Marsh Pass, Chaco Canyon, Grand Gulch, Allen Can- 
yon, and Montezuma County, Colorado. We know of no case of their 
having been found outside the San Juan drainage. 
Type II, ¢ (pl. 40, a). Of this:class we found but one specimen, a 
very fragmentary sandal, apparently of child’s size. The sole is 
made of stout yucca string in coil without foundation. The twists 
are pulled very tight, giving a flat, even texture. Netted to the 
edges of the sole are shreds of what was evidently an attached “ up- 
per” or, more probably, a sock-like legging. The stitch is the same, 
coil without foundation, but is very much looser. The thread of 
11897, pp. 129, 130. 41908, pl. xxxix, s. 
21909, pl. 21, j. 51897, p. 248. 
$1893, pl. xxxix, 6. 
