A preliminary test trench one yard wide and four yards long, was dug in
squares M6-9 to determine whether complete excavation was warranted.
Subsequent yard square test pits were sunk at P 7-8 and K 8. It was
hoped that these would indicate the nature of the geologic stratigraphy
and at least suggest the range of cultural materials which might be
expected. Since excavation seemed warranted, the entire area to be
worked was cleared of the surface layer of manure.
As set forth in the description of the site, the recent floor sloped
up toward the northwest (Fig. 8). Test excavations indicated that the
basal layer was approximately horizontal in a north-south direction,
but sloped gently up toward the west. There seemed to be no significant
correlation between geologic stratigraphy (see below) and typology
in any except the lowest layer, which had not yet been extensively
investigated. Because the evidence was as yet so scanty, it was
decided to excavate with reference to both the surface and the datum
point. Each square was treated as a separate unit and, except where
natural levels were apparent, excavation proceeded in four inch levels
measured from the surface at the northeast corner. Each level was also
correlated with the datum point.
Both typological provenience and the position of the basal layer
indicated that the major portion of the deposit had been laid down
horizontally. To test this hypothesis a careful watch was kept for
fragmentary specimens, in the hope of finding corresponding pieces from
separate locations. The relative positions of these would suggest the
nature of the surface on which they were deposited. Ten such fragments
found comprised five artifacts: three potsherds, one handstone and
one fragmentary grinding slab. Though some occurred as much as twelve
yards apart, each pair fell within the same vertical level. From
this analysis it appeared that the occupation levels were in fact
approximately horizontal.
Stepped excavation seemed advisable in view of the unconsolidated
nature of the deposit (Fig. 5). Vertical exposures more than two feet
high collapsed as soon as the lower portions became thoroughly dry.
These conditions made stratigraphic columns or balks impractical in
most instances. However, by the use of rip-rapping, it was possible to
leave a reference balk one foot wide on the north side of line N 5-11.
All material recovered was sifted through quarter inch mesh screens.
Since it was fairly loose and dry, most of the earth was removed with
shovels. Trowels, brushes and a flexible baby bottle for blowing were
employed when features or artifacts _in situ_ were encountered. The
location of the screening dump was determined by test augering of the
area between the site and the creek. This dump was separated from the
excavation by a catwalk and low earthworks. The latter served the dual
purpose of guarding against sliding from the dumps, and keeping out
water which often poured over the edge of the overhang in the extremely
rainy season of 1957.
[Illustration: LODAISKA SITE