The Groma
The primary instrument used to lay-out a legionary camp was the groma. The groma
consists of four plumbed lines
, fila cum ponderibus iunxerunt, at right angles centered
about a fifth plumbed line which established the datum point. Use of the groma was
quite simple. The surveyor, the
agrimensor or sometimes known by the instrument of
his trade, the
gromaticus, would first establish the location of the datum point. This was
accomplished by rotating the swing arm of the groma so that the central plumb line fell
on top of the datum point. After establishing that the groma was level, by confirming that
all of the pondera were at the same height, a line was the made by sighting along the
three collinear plumb lines of one axis of the cross at the ends of which the four plumb
lines were attached, the
corniculi. It is hypothesized that the corniculi rotated about the
end of an offset arm, the
ancon, which itself rotated about the main supporting staff.  
The central point about which the corniculi rotated and above which the datum point
was established was known as the
umbilicus soli. A second line at a right angle to the
first line could then be made by sighting a pole along the line formed by the
perpendicular plumb lines of the corniculi.  An excellent site having a very good diagram
of the groma in use may be found
here.
Until the turn of the last century, the only description that we had of the groma was from
the ancient literature. The primary source is a document on land surveying known as
the
Corpus Agrimensorum (I'm still looking for an on-line reference for the entire book).  
For example, the use of the groma, from which its construction may be inferred, is given
by both Frontinus, and by M. Junius Nipsus.  Nipsus in particular describes the
dropping of a plumb line from the umbilicus soli to establish the datum point.
A complete groma, or at least parts of a groma from which a groma was able to be
fashioned, was discovered in Pompeii.  Matteo della Corte made a
reconstruction.  
Recently there has been some discension as to the accuracy of della Corte's
reconstruction.  In particular, the existance of the Ancon, was doubted.  However, if the
accounts from the roman literature are to be believed then some sort of swing arm for
the corniculi had to exist.
One of my many desires is the construction of a usable groma.  As a first step, in the
following are some two dimension CAD drawings of the groma and its components.  
Unfortunately, the only reliable dimensions that I have are for the total width of the
cross arms of the corniculi, of 100cm, and of the total height, as reconstructed by della
Corte, of 170cm.  All other dimensions that I have are guesses.  The creation of these
drawings remains but a first step.  As construction evolves so too may the size of
things.  First the profile of the groma itself:
The main supporting staff is made from wood with the iron or steel ferramentum slipped
onto its base.  The following drawing follows the form of the ferramentum found by della
Corte:
I hypothesize the ancon could be fashioned from a block of wood.  Something firm, yet
light weight, like ash may be used.  My first visualization of how the corniculi and the
ancon would go together is illustrated below:
The details of the pivot points remain somewhat ambiguous.  I mustn't forget the
pondera. The sizes tended to vary, but they were basically conically shaped bronze
castings.  A central knob on the top had a pair of holes at right angles.  The looped
filum inserted at the top was secured by means of a pin.
Groma side view
The Ancon
The Pondus
Ferramentum