Introduction Edit

The triggering mechanism is very vaguely described in the text of the cheiroballistra manuscript. Fortunately the manuscript diagrams (see Wescher 1867; Schneider 1906; Wilkins 1995) clarify the text a lot. Regardless, most of the dimensions are missing. This triggering mechanism borrows some ideas from Iriarte's (2000) earlier work. Wilkins reconstruction, though commendable, is based on the idea that cheiroballistra had a winch (1995: 14-17).

Look here for closely related articles:

Parts Edit

Fork Edit

Similarly to the claw, all authors agree upon the general appearance of the fork (ΕΖΗΘ). Wilkins (1995: 14) calls this component "double bracket and tenon". The part ΕΖ is the "two-pronged" (δίχηλος) part and ΗΘ is the rectangular tenon (τόρμος τετράγωνος). The double bracket is bored at ΤΥ to receive an axle. The axle is also pushed through the hole bored at Φ in the claw inserted inside the fork. Note that the letters in manuscript diagrams don't exactly match what we'd expect looking at the text.

FIXME: add updated CAD picture

FIXME: add updated photo

Practical tests have shown that the fork has a tendency to rise up from the slider when the ballista is cocked and shot. This can be prevented with a pin pushed through the side of the slider and the fork's tenon part. Alternatively the end of the tenon can be riveted so that it stays tight against the bottom of the slider. This latter option makes replacing the fork very difficult, and is not recommended.

The fork and tenon is fairly easy to make from two separate pieces.

Claw Edit

The claw (σχαστηρία, ΚΛΜ) is well-known from older artillery pieces, so there is little disagreement on it's general form (see Marsden 1971: 219-220; Wilkins 1995: 17; Iriarte 2000: 52-53). The claw has a 1 dactyl long incision and has a horizontal, round hole at Φ. This hole is used for the claw axle which also goes through holes in the fork.

FIXME: add an up-to-date CAD picture FIXME: add an up-to-date photo

The claw is easiest to make by forging.

Trigger Edit

Similarly to the claw and fork the general form of the trigger (δρακόντιον, ΝΞ) is well know and agreed upon. Wilkins (1995: 14) has translated this component as the snake (diminutive of δράκων, "small dragon"). In a sense, this is a good choice, because the σχαστnρία (claw) translates literally to "trigger" or "release mechanism" which can cause confusion.

FIXME: add an up-to-date CAD drawing

A round hole is punched or bored to the trigger (from top) at Ν. A corresponding hole is bored to the slider (ΓΔ) at Π. An axle is then inserted through both holes. This allows the trigger to rotate around it's axle on top of the slider.

Handle Edit

The form of the handle (χειρολάβη, ΑΒΓΔ) depends on how the cheiroballistra is cocked, even if we leave aside the highly questionable "winched cheiroballistra" theories. Translating literally, χειρολάβη means "something grasped with hand(s)", or "handle".

FIXME: add an up-to-date CAD drawing

In the text it's stated clearly that there's a round hole in the handle at Δ, which is used to attach the handle to the slider (ΓΔ). Wilkins (1995: 16) and Iriarte (2000: 53) did not believe the handle was actually used as a real handle, whereas Drachmann (1972: 493) and Baatz (1974: 62) did. There are good reasons to believe that handle was grabbed with both hands to assist pullback. Also, Iriarte (2000: 52-53) is very likely correct in that the handle was used to lock the slider to a pin in the case when the weapon had been cocked; if we don't start inventing new core components to the cheiroballistra, the handle is the only one that suits that purpose well:

  • It's very close to the end of the slider, beyond the fork and claw
  • There is a round (axle) hole in the slider for it
  • There is a rectangular hole at the end of the slider, apparently for accommodating the shaft of the handle

FIXME: add pictures of a real pullback handle

Pittarion Edit

The Π (pi) bracket (πιττάριον, ΟΠΡΣ) or pittarion is one of the most controversial parts of the triggering mechanism. Marsden (1971: 221-222) thought it to be a rivet-plate to which other triggering mechanism parts were attached to. Gudea & Baatz (1974: 62), followed by Wilkins (1995: 16) and Iriarte (2000: 53-54) interpreted it as thumbgrip for pushing the slider forward; based on my practical tests I find this interpretation somewhat suspect, as both the handle (χειρολἁβη) and the trigger (δρακόντιον) are more suited for pushing the slider forward.

FIXME: add up-to-date CAD drawing

Use of the pittarion as a rear sight is one possibility, but requires the pitarion to be made fairly high if a thread and bead tightened between the little arch act as the front sight.

The pitarion could have also been used to reinforce the grain of the wood behind the handle axle, but it's location is ill-suited for the job, unless we are prepared to change the numbers in the text and disregard the manuscript diagrams. More on this follows in the assembly section.

Real-life examples Edit

FIXME: add recent pictures of the triggering mechanism

Cheiroballistra triggering mechanism parts - 01 Cheiroballistra triggering mechanism parts - 02

Assembling the triggering mechanism Edit

The triggering mechanism of the cheiroballistra is described very poorly by P.H. Fortunately two different manuscript diagrams have survived. Also, two of the components, the claw and the trigger are known well from older types of artillery. In addition, there are various useful hints hidden in the text. All of the above combined with practical tests allow us to reach a fairly realistic reconstruction of the triggering mechanism.

P.H. gives only a few relatively certain measurements of the triggering mechanism:

  • The length of the incision in the claw is 1 d.
  • The distance between the fork's tenon and the end of the slider is 5 d.
  • The distance from the trigger axle to the handle axle is given as 4 d.
  • The distance between the pittarion's two posts is given as 4.5 d

The are a couple of problems with these measurements. The more obvious one is the distance between pittarion's two posts, 4.5 d, which would mean that they would not even fit inside the case, let alone the slider. Iriarte (2000: 53-54) suggests that the numeral Δ (4) migh be corrupt and would have to be read as Α (1) instead. This is fairly plausible, and would place the posts 1.5 d from each other and thus allow them to (barely) fit inside the slider's width. Iriarte (2000: 53-54) placed the pittarion between the handle and the fork, whereas Wilkins (1995: 16) placed it in front of the fork. I've placed it between the handle axle and the end of the slider to reinforce the grain of the wood.

The second problem is less obvious. If we follow the instructions closely and place the handle's axle 4 d (Δ) away from fork's tenon, it will be very close, less than 1 d away, from the end of the slider. In practice this means that the grain at the end of the (wooden) slider will split if the handle is pulled or used to lock slider in shooting position. We then only have four options:

  • Start inventing "missing parts" in form of metal reinforcements
  • Change the numeral Δ (4) to Α (1) and thus place the handle axle under the trigger and possibly foul the trigger's movement
  • Assume the handle was not stressed to any significant extent and invent some missing part to lock the slider to the case
  • Violate both the text and the manuscript diagrams by moving the handle forward in the slider

The location of the handle is all the more interesting as there is plenty of empty, unused space between the trigger axle and the handle axle. Why insist on making design decisions that necessitate metal reinforcing, when one can do without? Simply by placing the handle axle farther away from the end of the slider the whole structure becomes much more robust, and no metal plating should be necessary. Furthermore by placing the pittarion between the handle axle and the end of the slider we further reduce the risk of wood splitting along the grain. Thus in my reconstruction the handle is attached as close to the claw and trigger as possible, and the pittarion between the handle axle and the end of the slider.

In a nutshell there is a good reason why the handle axle should be where I've put it, and no good reason why it should be at the end of the slider. So, in this case, I rather trust common sense than the manuscript, as they're clearly contradicting each other.

One more thing requires further discussion. The text is very unclear of the purpose of the four-sided (παραλληλόγραμμος) hole in the slider at Ξ. The exact same word, given in LSJ as "bounded by parallel lines", is also used in connection with the four-sided holes in the little ladder beams. When P.H. spoke of the tenon in the fork and the dovetails in the slider and the case, he used the term τετράγωνος instead; this term usually means a square, but can mean anything that has four sides. So, the exact terms used the text do not help us at all in deciphering the purpose or location of this particular four-sided hole.

In any case, the hole in the slider at Ξ is mentioned when P.H. speaks of fitting the handle and the slider together through the round hole (ΜΝ) at the end of the slider. In the single diagram where it's shown, the rectangular hole is located between Μ and Ν. Iriarte (2000: 52-53) interpreted the rectangular hole as a lengthwise, rectangular slot extending forward from the end of the slider to the base of the handle. In his reconstruction the handle rotated freely up and down, which was necessary to lock it to it's anchor (a strong nail) at the end of the draw and to unlock it after the shot. I think this interpretation is basically correct and I have chosen to use it myself. Wilkins' (1995: 16-17) alternative interpretation of the handle would not work in a stomach-cocked weapon.

Preventing bowstring slipping under the claw's finger Edit

Practical tests have shown that the bowstring will slip under the fingers of the claw fairly easily. There are two ways to combat this nasty phenomenom. First and foremost the fork/claw axle has to be of just the right size. If the axle hole has been drilled with, say, a standard drill bit 6mm in diameter, you can't use a standard bolt 6mm thick as the axle - the fit is simply not tight enough. What you can do is take a slightly thicker (6.5mm) round steel rod for an axle and file it down until it fits snugly.

The second way to prevent bowstring slippage is to sink the claw's finger to the slider. If the slider is wide enough you can carefully drill small holes into it for the fingers. Alternatively you can just give the fingers a good smack with a hammer to make shallow dents to the slider. Both of these methods allow the fingers to rise up slightly without allowing the bowstring to slip under them.

In either case you should fit the end of the claw and the trigger together after the above things have been done.

Operation of an Iriarte-style handle Edit

A round Iriarte-style handle can be made to automatically lock to the pin in the case if the pin points backwards instead of forward (original blog post). This means that when the handle contacts the pin, it will rise until it finally drops behind the pin. These pictures illustrate the mechanism perfectly:

Handle auto-locking - 01

Handle auto-locking - 02

Handle auto-locking - 03

This mechanism works in fully automated fashion, but only if the slider is fairly close to horizontal. Unfortunately this is not the case during cocking, so a small tap with hand is necessary to lock it, but this is a significant improvement over having to operate the handle entirely manually. Thanks to my friend Boris for pointing out this very simple improvement!

Additional sources Edit

Manuscript diagrams of the triggering mechanism have been published in

  • Wescher 1867: 127-128
  • Schneider 1906: 150-151
  • Wilkins 1995: 15

Look at the Bibliography page for availability details.