Taming the Onager

[P. Clodius Secundus]

Greetings Roman Artillery enthusiasts,
The following is my interpretation and design of the onager as described by Ammianus. I apologize in advance for being so long-winded. It is a lot of material to cover. For those who are not completely bored by it, I am preparing a full article on the matter including proper citations, bibliography, etc. In the meantime, I am posting this in its current form just to stimulate discussion over the holiday season. In the spirit of the season please be charitable and don't savage me too badly for my heresy. The translations included are a mixture of those commonly available (Loeb, Schneider, etc.). There are many who deserve credit, but I would especially like to acknowledge Duncan Campbell for keeping the the De Reffye Onager alive in the public consciousness.
Warm holiday wishes to all, and may Santa leave a fresh copy of Marsden's "Treatises" in you stockings,
Randi Richert, AKA Publius Clodius Secundus LEG III CYR.


The Onager was the simplest of all Roman torsion artillery weapons. It should, therefore, be the easiest to reconstruct. Unfortunately there are several factors that conspire to make the task more difficult. The first is a lack of artifacts or iconographic evidence to aid the researcher. Major pieces and some carved images of other weapons like the catapulta and ballista have been discovered. Technical manuscripts on them written by noted engineers such as Vitruvius and Heron have also survived. The dimensions and formulas they contain provide a solid basis from which to proceed. In the case of the onager there are only a few passing references to their tactical use and a description by Ammianus Marcellinius, a Military Officer, rather than an Engineer. Written from an historical perspective his four sparse paragraphs rely on descriptive imagery instead of technical jargon. Characterized as vague, mistaken, or ill-informed, his observations are often disregarded by leading scholars. I would suggest that this perception may be due more to the ignorance of the reader than the author. In his defense I will attempt to provide a rational explanation of his text and in so doing, point out where scholars may have erred.

The first paragraph, though seemingly straightforward, manages to introduce controversy. It can be translated thus;

The scorpion, which is now-a‑days called the wild ass, has the following form. Two posts of oak or holm-oak are hewn out and slightly bent, so that they seem to stand forth like humps. These are fastened together like a frame-saw and bored through on both sides with fairly large holes. Between them, through the holes, strong ropes are bound, holding the machine together, so that it may not fly apart.

Nearly all scholars agree upon this description this of the ground frame and spring bundle. Their versions of it vary little except whether the spring bundle is in the center or front portion of the frame. It is with the last statement, “holding the machine together, so that it may not fly apartâ€

[LCorneliusScaeva]

Great job! Laudes!

[Magnus]

Well done Randi...a laudes from me.

[Matt Lukes]

Very interesting Randi, I look forward to reading/ seeing more as you progress

[LCorneliusScaeva]

I was wondering why you chose to bend the 'bent arm' the way you did? Would the option depicted in the attached jpeg also work?

[P. Clodius Secundus]

Cornelius,
It's a very good question. I chose to bend the arm that way because I was trying to follow the text as closely as possible. He starts out describing the arm in the un-drawn condition. It is only later that he talks about it in the drawn position. Also, drawing down the arm so that the main part is horizontal re-introduces the same conflicts and inconsistencies that caused most scholars to dismiss him in the first place. The beauty of that minor little change is that it resolves so many issues. Questions like your's are exactly what makes this forum so helpful. Thanks for asking.

[P. Clodius Secundus]

Here are a few shots of my little De Reffye-Richert prototype "Ammianus' Revenge". The washers are 1.5" diameter. It's only made out of pine but it is a very sturdy little frame. The spring bundle is made from nylon military parachute cord commonly known as 550 cord due to its 550 pound rating. The washers are only about half full but it already has impressive strength. I've got a piece of hickory log with an appropriate shape drying to make the bent arm. Next is the spindle for the winch. In keeping with my goal of using the least ammount of hardware and keeping things simple I've changed the design a bit so that the winch spindle clamps in between the beams. The levers will fit into holes drilled around the ends of the spindle. The part where the rope wraps around the winch will taper in toward the center so that as the arm is drawn back further it actually gains mechanical advantage. I was going to put a ratchet on either side, but to keep things simple I'll just use the levers backed against the cross beam to hold it in place. Ammianus never mentions a winch, so why start adding things now.

[Gaius Julius Caesar]

I always assumed it was called mule because of what would happen to someone who got hit by the arm, or kicked by a mule. Pretty much the same effect.

[P. Clodius Secundus]

I always assumed it was called mule because of what would happen to someone who got hit by the arm, or kicked by a mule. Pretty much the same effect.

There is an anecdote about one of the ballistriari who was unlucky enough to be standing behind an onager when it misfired due to an improperly loaded stone. They couldn't find all the pieces of him. That's a lot of power! Part of the difficulty with trying to reconstruct the onager is overcoming our assumptions and pre-conceptions. For example, some scholars translate iugalis temonis (yoke pole) as chariot pole. The problem is that chariots were race cars, not everyday transport. If I say automobile, the first image that pops into your mind is a family sedan not a Formula 1 racer. If Ammianus had meant the differently shaped chariot pole I'm sure he would have been more specific.

[Gaius Julius Caesar]

Actually, the first thing that would pop into my mind would be an F1 car!!! lol: :lol:

[P. Clodius Secundus]

Actually, the first thing that would pop into my mind would be an F1 car!!! lol: :lol:

I stand corrected. I should have been careful to specify normal people :lol:
This sort of supports my point about Ammainus writing for a genral audience. Here in the U.S. even if I were to say race car most would probably think of a NASCAR stock car instead of an F1. Part of the problem with the evolution of technology is that it's not just the individual machine that changes, but the basic components evolve as well. In 1900 the steering mechanism on a race car would have meant a tiller. Ten years later as speeds increased the steering wheel took over. That's why, when something an ancient author says doesn't make sense I assume that my understanding and perspective are faulty not theirs.

[LCorneliusScaeva]

The images in the original post are no longer appearing. Can this be fixed. I think this post has important information and I'm hoping the images can be restored.

[LCorneliusScaeva]

The images in the original post are no longer appearing. Can this be fixed. I think this post has important information and I'm hoping the images can be restored.

Actually, it looks like more than half the text of the original post is gone, too. Somebody please restore this post!

Thanks,

Jim

[P. Clodius Secundus]

Jim et al,
Since it doesn't seem that my original version can be resurrected I am re-posting it in revised and updated form...

The Onager was the simplest of all Roman torsion artillery weapons. It should, therefore, be the easiest to reconstruct. Unfortunately there are several factors that conspire to make the task more difficult. The first is a lack of artifacts or iconographic evidence to aid the researcher. Major pieces and some carved images of other weapons like the catapulta and ballista have been discovered. Technical manuscripts on them written by noted engineers such as Vitruvius and Heron have also survived. The dimensions and formulas they contain provide a solid basis from which to proceed. In the case of the onager there are only a few passing references to their tactical use and a description by Ammianus Marcellinius, a military officer rather than an engineer. Written from an historical perspective his four sparse paragraphs rely on descriptive imagery instead of technical jargon. Characterized as vague, mistaken, or ill-informed, his observations are often disregarded by leading scholars. I would suggest that this perception may be due more to the ignorance of the reader than the author. In his defense I will attempt to provide a rational explanation of his text and in so doing, point out where scholars may have erred.

The first paragraph, though seemingly straightforward, manages to introduce controversy. It can be translated thus;

The scorpion, which is now-a days called the wild ass, has the following form. Two posts of oak or holm-oak are hewn out and slightly bent, so that they seem to stand forth like humps. These are fastened together like a frame-saw and bored through on both sides with fairly large holes. Between them, through the holes, strong ropes are bound, holding the machine together, so that it may not fly apart.

Nearly all scholars agree upon this description of the ground frame and spring bundle. Their versions of it vary little except whether the spring bundle is in the center or front portion of the frame. It is with the last statement, “holding the machine together, so that it may not fly apart”, that the trouble begins. The late Eric Marsden called this observation “ridiculous” and suggests that Ammianus was “confused” since the frame must be solid. That accusation itself is baseless. Using only mortise and tenon joinery on the crossbeams a perfectly solid frame can easily be held together by the cross tension of the spring bundle alone. Aside from simplicity this method allows the weapon to be easily dismantled and transported when unstrung. This is not an issue when building a scale model for a museum, but for an army on campaign it is a vital concern. Any Officer, engineer or not, would understand the concept. The frame of my first scale model prototype is quite sturdy. Using only half-filled 1.5 inch washers an almost no pre-tension it would be impossible to separate the frame without using enough force to shatter the wood.

Back to the text;

From the middle of these ropes a wooden arm rises obliquely, and being set upright in the manner of a yoke-pole, and is twined around with cords in such a way that it can be raised higher or depressed. To the top of this arm, iron hooks are fastened, from which hangs a sling of hemp or iron. In front of the arm is placed a great cushion of hair-cloth stuffed with fine chaff, bound on with strong cords, and placed on a heap of turf or a pile of sun dried bricks; for a heavy machine of this kind, if placed upon a stone wall, shatters everything beneath it by its violent concussion, rather than by its weight.

This passage, describing the throwing arm and placement of the buffer which arrests its motion, is the primary source of contention among modern historians. The two most popular designs, those of Sir Ralph Payne-Gallwey and E. Schramm, are quite familiar to enthusiasts. Their relative merits have been argued for years. Unfortunately, I believe that this has all been wasted effort. They differ mostly as to the structure and placement of a heavy wooden framework to which the buffer pad is mounted and positioned so that it stops the arm in a vertical or near-vertical position. This framework is not mentioned in the original text and completely unnecessary. The first, and I believe most nearly correct reconstruction of the weapon, was made back in the mid 1800s by the French artillery officer Verchere DeReffye. His version allowed the arm to travel forward and down onto a large buffer placed in front of the engine exactly as Ammianus describes. There is one seemingly minor point of the text which De Reffye missed which is the source of much of the confusion which surrounds this document. I am referring to Ammianus’ comparison of the arms configuration to that of a yoke-pole. Prior to the introduction of the padded horse collar in the Middle Ages, carts and wagons were pulled primarily by a neck yoke. Due to the higher hitch-point of this system the pole not only rose at an oblique angle from the axle, but the forward section was also usually bent dramatically upwards to meet the yoke. A quick survey of contemporary reliefs will confirm this fact. It is a seemingly minor detail that changes the whole equation.

Ammianus continues;

Then, when there is a battle, a round stone is placed in the sling and four young men on each side turn back the bar with which the ropes are connected and bend the pole almost flat. Then finally the gunner, rising to his full height, strikes out the pole-bolt, which holds the fastenings of the whole work, with a strong hammer, thereupon the pole is set free, and flying forward with a swift stroke, and meeting the soft hair-cloth, hurls the stone, which will crush whatever it hits.

Applying the “bent” arm concept to De Reffye’s design eliminates one of the most glaring contradictions used to discredit Ammianus. Namely, when a straight arm is nearly flat (horizontal) it should be close to the ground. Why then, should the gunner have to rise to his full height? Some have suggested that it meant he stood exaltedly as if puffed up with pride. I think they are trying too hard to make up for a faulty design. If a bent arm is factored in it easily makes sense without having to torture the translation. When the sling end of the arm is horizontal, it is still far off the ground. On a large machine the trigger, which I believe would be attached near the bend in the arm, could easily be far enough off the ground that one would have to reach up to strike it. An additional benefit of the design is that a longer sling can be used, gaining greater mechanical advantage. Competitors in modern Punkin’ Chunkin’ events have resorted to digging a “Carbo Hole” behind their engines to gain range using a longer sling. Another significant trend among them (including the world record holder) is allowing the arm to continue forward of the vertical where it is arrested by a series of wires and springs functioning much like the ground mounted buffer Ammianus described. That pretty much destroys the argument that a vertical buffer frame is necessary, since there is neither a practical, nor historical justification for its existence. I am currently working with a local team to convert an existing Payne-Gallwey type to a De Reffye/Richert forward buffer and bent-arm configuration. They have years of experience working with this weapon, so any change performance will be readily apparent.

Ammianus concludes with;

And the machine is called tormentum as all the released tension is caused by twisting (torquetur); and scorpion, because it has an upraised sting; modern times have given it the new name onager, because when wild asses are pursued by hunters, by kicking they hurl back stones to a distance, either crushing the breasts of their pursuers, or breaking the bones of their skulls and shattering them.

Whether or not one doubts Ammianus’s terminology, my design resembles a scorpion as much, or more so, than the traditional Schramm or Payne-Gallwey versions. Its “sting” is also more upraised than De Reffye’s. The last sentence, explaining the connection to the wild ass is significant. Notice that the association contains no reference to the weapon’s recoil, only to their propensity to hurl stones. Those who have added the un-needed and un-evidenced buffer frame to their engines exclaim with misplaced pride that their machines “kick like a mule”. To an artillerist, the need to arrest this violent recoil is just wasted energy.

[D B Campbell]

The late Eric Marsden called this observation “ridiculous” and suggests that Ammianus was “confused” since the frame must be solid. That accusation itself is baseless. Using only mortise and tenon joinery on the crossbeams a perfectly solid frame can easily be held together by the cross tension of the spring bundle alone.

Slightly unfair to Marsden, Randi. What he meant was that the machine cannot simply be two side timbers with strong ropes to hold them together/apart. Almost everybody is agreed that Ammianus has simply missed out the cross-timbers, needed to make a solid frame. (In case you think this point is obvious, the French scholar Jacques Fontaine managed to sketch the machine without cross timbers, so that the ropes really are all that holds the sides apart!)