The case, slider and crescent-shaped piece form the core of the cheiroballistra into which other (metal) parts are attached.
The case is the lower part of the cheiroballistra stock with a female dovetail groove running down it's length. The upper part of the stock, slider, has the male dovetail which allows it to slide on top of the case. Although the description of the case (e.g. Marsden 1971: 213) is relatively clear compared to most other sections in Heron's cheiroballistra, it can still be interpreted in a number of ways.
The part describing the location of the projecting block (ΚΘ) is corrupt in all manuscripts and does not make sense as is. A simple solution to this corruption was suggested first by Prou (1877: 120-121) and repeated later by Iriarte (2000: 48). Both simply substituted ΑΘ with ΛΘ and the text makes perfect sense. While this theory sounds most plausible to me, other explanations have been suggested by Marsden (1971: 218), Wilkins (1995: 11-12), Schneider (1906: 149) and Baatz (1974: 70).
The actual purpose of the projecting block has confused pretty much every researcher, as Iriarte (2000: 48) points out. I have interpreted it simply as a support for the little ladder holding the field frames. This is the simplest solution to keep the little ladder, the field frames and the little arch from moving backwards when the weapon is cocked. Of course the T-clamps are still needed to keep the little ladder in place, but without support from the projecting block the T-clamps would have to be much, much heavier and stronger.
As Wilkins (1995: 12) notes, removing wood along ΛΘ and ΑΚ as suggested by Heron (e.g. Marsden 1971: 213) seems silly. It seems clear that Heron is not thinking like a carpenter, who would have simply glued or nailed a piece of wood to bottom of the board ΑΒ and be done with it - as did I.
Full CAD drawing of the case below. Side, bottom and top views:
And the case from front:
The slider has a male dovetail corresponding to the female dovetail in the case. Although a relatively simple component, it's exact form is still not clear. There are two competing interpretations for the slider's cross-section:
- Most scholars (e.g. Marsden 1971: 218, Wilkins 1995: 11) have reconstructed the slider from two pieces forming a "T" shape. The lower part of this composite construction was the male dovetail to which the upper part was attached. The upper part simply rests on top of the case.
- Iriarte (2000: 52) suggests that the slider was made from one piece.
These differing interpretations stem from the fact that Heron did not state how wide the female dovetail should be; he only gives it's depth (1d) and length (46d). He also says that the slider should be "about" 2,5d wide and 1,25d high. The "T" proponents take 2,5d to mean the width of the upper (non-dovetail) part of the slider, whereas Iriarte (2000: 52) suggests that the male dovetail itself - being the only part of the slider - was about 2,5d wide.
I've personally followed Iriarte's interpretation as it is simpler and requires one to make fewer questionable assumptions. In addition, using a T-shaped slider places the slider too high, making the bowstring rub against the stock.
The "about" (see Iriarte 2000: 52) in slider width requires some discussion. If the slider was 2,5d wide, then only 0,5d (or ~1cm) of wood would be left on both sides of the slider. This is not much, but might be enough for durable operation. Nevertheless, I've made the slider slightly narrower (2d).
Below the slider viewed from side, bottom and top:
And from the front:
Crescent-shaped piece Edit
The crescent-shaped piece (ΗΒ) has a rectangular hole in it's middle and it's attached to the end of the case. Apparently a rectangular tenon was pushed through this hole and into the stock to keep it firm. None of it's dimensions are given in the manuscript, so all the measurements are based on practical tests:
The crescent-shaped piece is used to push the slider backwards with stomach pressure. This interpretation is almost universally accepted by all scholars, with the notable exception of Wilkins (1995; 2000, 2003) and Marsden (1971). There is no need to find any other explanations unless one is predetermined to interpret the cheiroballistra as a winched weapon.