262 COILED BASKETRY IN BRITISH COLUMBIA [ETH. ANN. 41 
we will also call attention to the irregularities in their work which are 
instructive because they show how attempts have been made to solve 
the arising difficulties. 
The imperfections will be taken up first in a general way, from the 
standpoint of the adjustment of the designs to the field, and coincident 
with this, as occasion demands, the attention of the reader will 
be called to any minor points that may present themselves. The 
relation of designs to each other, as regards spacing or incongruity, 
will also be touched upon, as well as the adjustment of figures made 
necessary by wrong calculations. Next occurs the question of sub- 
dividing the surfaces of imbricated areas, such as the vertical stripe, 
and the lesser mistakes that result from the varying width of stitches 
or from distracted attention, wrong calculation, or inherent imability 
on the part of the worker to keep in mind changes of rhythm. It 
will also be interesting to discuss some specimens technically and 
artistically almost perfect. Some problems arising from the decora- 
tion of lids, when the designs are carried over from the walls of the 
basket and converge there, together with the related question of 
designs on oval trays, are also important. 
We shall first discuss the various types of designs in use by the tribe 
as burden basket decorations. We must recall the several distinct 
styles of distribution which have already been treated, namely, the 
horizontal, diagonal, vertical (including the vertical stripe), all- 
over, and large single patterns. It has been observed that the 
Thompson have tried and still use all of these in the application of 
their designs to baskets, so that they offer a more varied and inter- 
esting study than the similar specimens manufactured by the sur- 
rounding tribes. In the application of large single designs, one to a 
field, which is perhaps the easiest type of decoration and strangely 
enough one of the rare ones, the first problem is to center it, 
which is generally accomplished by eye, but sometimes aided by 
means of rough measurements. These designs must be, in most 
cases, symmetrical, and shght inaccuracies such as arise from the 
varying width of stitches are usually not very obvious. Plate 46, c, 
shows a basket, which, while having more than one figure to a face, 
at least has only one on each level, and each of these designs center- 
ing on the same principle as a single large design. It is interesting 
to note how the upper figures have been increased in size the better 
to fill the larger field. (See also pls- 28, f, g; 29; 31, e; 46, c; 51, h.) 
One of the simpler distributions is the horizontal banded arrange- 
ment which runs completely around the basket (pls. 8, d; 9, a; 21, d; 
27, c, d, f; 36, c; 41, i; 54, b). Here, at least, with plain continuous 
lines, there is not the difficulty of spacing isolated designs or vertical 
stripes and providing for fillers, etc., or of doing what seems to us so 
