Sometimes you excavate a feature that just defies all logic and every once
in a while we find new features types which either have not been defined
before, have been misidentified or simply written off as a natural anomaly.
Recently, near Middletown, Delaware at the Elkins A site (circa 1740-1780),
Hunter Research, Inc. excavated a feature which we had initially considered
to be a well or the cone for well, (located 45 feet from an mid-18th century
stone-lined root-cellar).  It has now become one of the most puzzling and
exciting features at the Elkins A Site, and possibly on the Route 301
project to date.  This circular feature is situated on the side of a shallow
gully that defines the northern boundary of a mid 18th century homestead.
The feature measures approximately 15 feet in diameter and extends three
feet below the plowzone (See photos in attached blog).  There are posts
positioned approximately every three feet around the base perimeter.  Each
post is angled inward 10 to 30 degrees with rocks jammed in around the base
to prevent the post from falling into the pit.  Several green sandstone
slabs situated in the upper center of the fill exhibit no signs of trimming
or shaping (they do however contain thousands of fossils not thought to be
related to the function of the feature).  No other artifacts have been
recovered from this enigmatic feature.  The lack of artifacts suggest that
it predates the house, as otherwise household refuse (in the form of a sheet
midden) would have undoubtedly been part of the upper fill.  There is no
central shaft (ignore the central rectangular test and two small auger tests
shown in blog image). This area has experienced between one and two feet of
soil deflation.  The fill inside the pit appears to be the same soil
excavated from the pit suggesting it was banked against the pit or packed
against branches woven between the posts and either fell back into the pit
when the posts were removed or rotted away.  So what is it?  We have had
several professionals visit the site with interesting ideas, but none are as
yet completely satisfactory. We have ruled out a well, privy, cistern,
spring head, windmill, brick/clay pit or silo.  One exciting suggestion is
that it is a wolf trap.  Construction of wolf traps was mandated in this
area during the 1670s and 1680s as attempts were made to eradicate wolves as
pests.  The small crossroads settlement of McDonough, a couple of miles
south on what is now Route 13, was originally called "Trap".  A competing
possibility is a "field root pit" type of root cellar used to store staples
such as potatoes, carrots and turnips.  Field root pits are often
rectangular but can be circular. Some also have peaked roofs.  We have taken
soil samples for flotation and chemical analysis which is underway.  We also
have carbon samples as we cannot rule out a prehistoric origin.  Your
thoughts are welcome.  See the attached Route 301blog sponsored by DelDOT
below for digital photographs, weekly updates regarding our site as well as
two other data recoveries along the corridor. 
http://blogs.deldot.gov/category/us301/archaeology-updates/

Bill Liebeknecht, MA, RPA
Principal Investigator
Hunter Research, Inc.
Trenton, New Jersey