|CMi#1: Sesterce of Nerva 97 AD. Click the picture to see two grazing mules in front of a cart. The only clear depiction of a Roman collar harness off use. Between the two collars is the cart pole. The function of the two further lines to each collar is unclear. Perhaps they are separated reins. The collars are padded and the wooden pads at the end of the bows are visible too. Only the straps to fix the pads in position on the mules were to small to be depicted here. Interesting is the attachment of the collars to the pole. It seems of rigid design. The attachment point is clearly not at the top of the collars but at the sides.|
Only find of a Roman collar with wood pads still on the iron
core. It happened in the Great Moore of Le Rondet, Swiss, in 1963 and 64.
The site is located between lake Neuenburg and lake Murtensee.
Among hipposandales, horseshoes and a lot of other things were
about 100 iron collar bows.
The wood pads are about the size of a human hand. Fine fibrous wood
of apple-, pear- or mountain ash trees was used. The inner side
is flat and polished. The other side has one or two iron hooks
and is fixed or moveable attached to the iron bow. "These iron bows
were often found in Roman castels or army camps. They are usually
called saddle bows. The still attached wood pads of the Le Rondet
find show the bows as parts of Roman horse collars."
[Schwab, Hanni: Le Rondet - eine roemische Militaerbruecke im Grossen Moos, Arch. Korrbl. 3, 1973, 335-343]
The iron core of Roman collars from the find of Neupotz, southwest Germany,
ca. 270 AD
The width varies from 195 to 305 mm, the height from 162 to 215 mm and
the weight from 0.68 to 2.204 kg.  Note the asymmetrical bow ends of the
right collar. That is probably a collar wich is rigidly connected to
a cart or wagon pole like in CMi#1. All others are symetrical with
no trace of the traction attachment point. Therefore the majority
of the collars was not rigidly connected to a wagon but by cords
[1 Sigrid Alfoeldy-Thomas: Anschirrungszubehoer und Hufbeschlaege von Zugtieren. In: E. Kuenzel, Die Alamannenbeute aus dem Rhein bei Neupotz, (1993), p. 331ff]
A Roman ox collar, used for ploughing in North Africa
around 225 AD. Usually (at least today), only yokes
are applied to oxen. But this unique Roman mosaic shows
one ploughing team using a yoke and this one using a
collar. In this unique design, the traction point is
as low as possible. That offers an advantage in
plough design. A straight plough shaft is now
usable without increasing the traction force to
the ox for reasons of geometry. The plough is now
of more simple and robust construction, the
bending load in the shaft is reduced significantly. That is
probably the reason why this type of plough is used
for the high force application to break up the ground,
whereas a conventional one is used for the lighter
job of ploughing in the seed.
Actually, not oxen (castrated bulls) were used but real bulls. The Romans seemed to be able to handle them. Today's bulls are considered to be too difficult to control for rural work and oxen are preferred. But it is suggested that some working duty is of benefit for breeding bulls. [Schlipf, Praktisches Handbuch der Landwirtschaft, 1958, p. 342]