Bogged in winter blues, I struggled to get the rungs done.


I chose round bar for the rungs because it is very simple, and makes a much nicer hand hold when hefting this thing around.  Eventually the hex head bolts will all be replaced with square nuts.   For the historical purist, screw threads are contraindicated.  However, for the experimentalist, flexibility in how this contraption is assembled makes sense from a development standpoint.  Rivets are more permanent than ideal if one anticipates changes…

A set of lower struts will connect to the ladder at the same point as the these rungs.  The struts will connect back to the stock, much like on Firefly, and stabilize the stock and framework with great rigidity.

The Occamists may object.  But then, they always do.



Slogging along with the ladder fit-up, Phoenix’s core framework gets together for the very first time.



Because the artifact of the Elenovo kambestrion is unusually robust, with it’s wide and heavy duty stanchions and generous loop dimensions, this design theme has been carried over into Phoenix’s kamarion and ladder construction.  Even though she is only seven tenths the size of Firefly, Phoenix will have the ability to handle much higher loads safely.  By the time struts are added to stabilize the kamarion, the limbs will be far and away her weakest link. Which is exactly what we want for developing maximum power.  In my experience, single component failure is always easier to deal with than multiple things going wrong.

It takes a fumbly old man about five minutes to get all the wedges and keys knocked into place properly.  No doubt a team of well trained ballista techs could get it done in a minute or so.



This basic design with the cross-wedge and keys came from Mr. Iriarte’s paper on the inswinging theory.  I added the idea of the steel cross-wedges rather than wooden ones shown in his paper.  Given the pounding I intend to give these parts, we don’t want any soft spots caused by inappropriate joinery.




Or in this case, half a gallon of cutting fluid.


No, that is not a horizontal mill.

Mess, courtesy of metal chips that built up under the magnet block I use to anchor the return hose.  But the cuts look good.


I have observed two types of historians during the course of this little torsion project.  There are those I call the Occamists.  These stern fellows have taken it upon themselves to remind us all of what is strictly a part of the historical record and what is not.  Like faithful pruners of their gardens, they rarely meet ideas that don’t inspire them to acts of excision.

Then there are those of us who are perfectly well aware of the razor-like insights proffered by the Occamists, yet still find it valuable to extend our observations by using a combination of imagination and logic in an attempt to flesh out those areas of ancient technology that belong to the realm of the not strictly known.

The Occamist faction seems to delight in reminding their more, shall we say, creative colleagues, that they are susceptible  to a dreaded disease known as “blue sky thinking”.  Often, these dour guardians of historical precision,  take it upon themselves to administer dire correction to anyone who strays outside the fold of absolute certainty.  Or at least, their version of it.

Of course the trouble is, when it comes to the study of ancient catapults, what is actually known for sure makes paltry pickings for further study and examination.   This is where the arts of extrapolation and thoughtful experimentation must take over to light the way.  Without a considered use of imagination to suggest how these machines were built and affected the ancient world, the whole area of historical catapult study becomes moribund and repetitive in the extreme.

What we do know is far exceeded by what we do not know.  We can gage the depth of this ignorance by those few scholars that have tried to replicate ancient catapults and generally failed miserably to get any decent performance out of them.

However, there is a modern breed of experimental archaeologist that prefers to blend the “known” with an experimental approach that is oriented towards developing high-performance.  And by this methodology, further our understanding of how these remarkable machines probably worked and looked.

What the Occamists might call “blue sky thinking” , I prefer to call, “ripe fruit on low hanging branches”.  Perhaps there is no irrefutable direct evidence that a particular type of technology existed, but because it would have been so easy to achieve, and is a natural outgrowth of well known technologies that are a part of the historical record, and, very importantly, because this “ripe fruit” technology would confer substantial benefits to it’s users, there is every likelihood it did exist.  Enough anyway, to make looking for the signs of it a worthwhile endeavor.  Absence of evidence is not evidence of absence.  And because the Occamists are constitutionally averse to using their imaginations in these matters, they make poor readers of what signs do exist.

Take for example, the subject of shooting lead bullets (or glandes) from ballistas that normally only fire bolts and light javelins.  In my experiments with Firefly I was able to make this double bowstring, with a pouch,  in an afternoon.



It was my first attempt at making a sling string, and yet in short order I was able to boost that one pound lead egg up to 318 feet per second.  This next video shows the effect at 50 yards of one of these projectiles travelling at 295 fps on four sheets of 5/8″ plywood.

Click for vid: glandes 1

In the next video we see the first rough shot prior to sighting in.  It struck a good foot above the plywood targets, slipped through the camo netting and tarp, ploughed through a dense hay bale and smacked a two foot splinter out of the 2″ X 8″ douglas fir backstop, and then went on to God knows where .  You can hear the sound of it hitting that fir backstop if you listen carefully.

Click for vid:  glandes 2

Here is a view of one of these whirring beasties speeding downrange.  (Unfortunately Firefly is not visible as she is hidden inside the open door of my shop.  There was no way to move her back then.)

Click for vid:  glandes 3

If you turn up the volume and look at this last video on a big screen, in a darkened room, you can gain a sense of how these projectiles must have appeared to their recipients. (Okay, hypothetical recipients.)  Just visible is a glint of lead in the left of the open doorway, followed by a streak of grey to the left of center screen, and then a small splash of dirt on the lower left as the glans grazes the top of the dirt backstop. The knocks you can hear after that are from the glans crashing into the tree line 40 yards further on.

In my last excursion into glans testing (back on  April 9, 2012, see archives), a sling string was used to launch 14 glandes at once, each weighing 31 grams.

sling string 2


The effect of this “shotgun” pattern can be seen as seven of the fourteen glandes blew through  a 1/2″ thick plywood patterning board at 35 yards.  The seven that missed during this shot could been seen going through the air in a similar pattern directly to the right of the plywood.  So the actual pattern is twice what the video actually shows striking the plywood.

Click for vid: glandes 4

Please keep in mind this patterning exercise was really just a first attempt, conducted in a single day.  How effective would a shotgun technique like this have been in ancient combat?  Especially if the technique was well developed and several times more devastating than the one seen in the video?  That is largely a matter of context.  For opposing an assault through a breach in a wall, or one that was funneled into an appropriate kill zone, the shock of it would likely be quite demoralizing for the attackers. Especially if multiple machines fired at once or in volleys. Perhaps at lower power settings the Romans would have found it useful for crowd control.   Who knows?

The point of all this is to illustrate the concept of “ripe fruit on low hanging branches”.  The Occamists, with their penchant to oversimplify, are not in a position to consider the existence of these easily attainable technologies.  I know they consider themselves to be defenders at the gate, keeping at bay the hordes of speculators that threaten their ivory towers, but let’s face it, the Occamists are also somewhat bereft of imagination and therefore unlikely to connect any of the fairly obvious dots.  Especially the pragmatically opportune ones.

Ejecting high speed glandes was a part of this project that was accomplished without a lot of undue fuss.  In other words, performance like that shown in the above videos was achieved with very little effort on my part.  Why the Romans wouldn’t have included a simple, inexpensive sling string for shooting glandes and the like in their normally bolt shooting ballistas, seems remote to those of us that have actually done the experiments.

“Blue sky thinking”  or “easy pickings”?  As my wife Rebecca says,  “it is better to know that you believe, than to believe that you know”.  The Occamists, and their predelicton to streamline all avenues of thought, are not well suited to exploration.  Probing the unknown is what this game is about.  That requires imaginative and physical engagement,  tempered with careful caveats, not just smug criticism devoid of any sense of the potentialities involved.


Okay, this is weird.  The blog post that follows just popped itself out of the archives and redated itself to the present.  Probably my fat fingers teasing me again.  I’ll not correct the error and just figure providence gets a say here too.  Here it is, ancient in it’s own right:


— I have been guilty of certain apriori approaches in the development of Firefly that have caused a lot of unnecessary effort.  Perhaps the greatest of these has been my fixation on protecting the field frames and the kamarion from the shock of the limbs crashing home.  Originally I had assumed that leather buffers would be a good idea to prevent damage to the machine from so much energy having  to be absorbed by the framework.  Great idea, until the leather buffers broke down and caused endless problems with bowstrings breaking as they absorbed all the shock.

After the buffers I moved on to heavy bronze hardstops intended to spread the load over a greater area of the stanchion,  I really thought I was on to something because now the limbs had a positive stopping point and the bowstrings were well protected from the limb shock by simply adjusting their length so that they were just long enough not to be over stressed.  The frames showed no damage with these bronze hardstops, so obviously this approach was working.

This phase with the bronze hardstops lasted quite a while.  Eventually my  desire to increase the limb rotation past 90 degrees caused the removal of these  pretty bronze parts, and with it (for reasons I don’t remember) my concerns about frame damage.

When we have experienced the pain of all the things that didn’t work we are ready to learn. —


Increasing limb rotation beyond 90 degrees?  Because, you know, back then I’d dreamed there were fabulous kinetic riches to be had with an extra long power stroke.  Oh! the innocence of the young!