Battle Rams to Ballistas: An Exploration on Wartime Technology in the Classical Era
Advisor: Odell
Internship Location: Remote
Onsite Mentor: Dr. Richard Leese, Conflict Archaeology Researcher and Consultant, CIRCA
Hello everyone, my name is Aadrit Talukdar, and my project is Battle Rams to Ballistas:
An Exploration on Wartime Technology in the Classical Era. Throughout my research, I will be combining literature studies of ancient texts from Mediterranean Classics with engineering experiments on technological designs. The literature studies will be using sources from my time period (5th century BCE to 5th century CE) around the Mediterranean, and using a combination of my own language knowledge with English translations to analyze the texts, specifically on details of siege weaponry used in notable wars. With that information, I hope to engineer replicas or smaller-scale models of these technologies to investigate their efficiency more closely.
My final product is meant to be both a comprehensive comparison of ancient texts relating to the technology and time periods I will be researching, and multiple engineered models that can be compared to visualize technological improvements and design changes.
Project Blog
Week 1: Blog Post - Introductions
Hello everyone, and welcome to the first issue of Ballistas to Battle Rams. My name is Aadrit Talukdar, and I will be your historian/engineer guiding you through my research journey.
My project will focus on how siege weaponry has changed across time and the impacts that evolution made on war. Siege weapons, in this case, apply to both the defensive technology used to fortify cities and towns, as well as offensive technology used to take over those locations. I’ll be researching technology from the Classical Era of Greece (5th century BCE) to the fall of the Roman Empire (5th century CE).
I chose this topic as a blend between my two interests: classics and engineering. Taking Latin for 7 years is a testament to my love of ancient history and culture, while engineering projects have filled my middle and high school years.
Through this project, I hope to learn how ancient sieges used the technology of their time, and recreate some of that technology myself (albeit on a smaller scale). Through both literature and engineering studies, I hope to find connections between the events and results of warfare to the evolving weaponry used.
The next few weeks will be literature studies, so get ready to read about ancient wars throughout the ages. After I’ve highlighted key points of comparison, I will start building and modeling the technology based on written descriptions and diagrams I found earlier. Throughout the process, I will be getting closer to answering my research question.
Stay tuned for the next post!
Signing off,
Aadrit T.
Week 2: An Introduction to Research via Caesar
Hey everyone, and welcome back to the second issue of Ballistas to Battle Rams. If you’ve forgotten after only a week, my name is Aadrit T, and I will be your host through this week’s progress update in my project.
When it comes down to historical research, a lot of the work comes down to reviewing different sources. As a student who took AP Latin, one of the biggest sieges that is well covered is Caesar’s Gallic Wars. During the 50s BCE, Caesar led his army throughout Gaul, trying to expand Rome, and wrote about his exploits in Commentarii De Bello Gallico, or Commentaries about the Gallic War.
One of things Caesar focuses on is how his wars are fought, mostly to show him in a better light. However, he’s never as focused on wartime technology as I’d like him to be. For the first start of my project, I wanted to focus on the sections he talks about his many sieges, namely in Books 7 and 8.
My research methodology is relatively simple. First reading through a translated version of Caesar’s Books, I would highlight the sections talking about specific sieges, like his attack on the Bellovaci, or the famous siege of Alesia. With these highlights chosen, I can do a deeper dive with the Latin itself, translating the words on my own to avoid the biases that come from a separate translator. Through this process, I can pick out words that may hold more meaning than it seems at first.
One find is through the siege of Uxellodunum, in Book 8, Chapter 40. Caesar uses the word “tormentes”, to describe his siege operation. Looking into it further, it shows how the term is used interchangeably between catapults, ballistas, and other machines for projectiles (as defined by the Oxford Latin Dictionary, the reputable source for Latin words). What this tells me is that throughout the Roman weapons history, catapults and ballistas can be considered in the same age, especially through seeing the authors that use such a word being in similar time ranges.
Another find is when Caesar describes Gallic wall constructions, in Book 7 Chapter 23. When describing how the mix of stone and wood is beneficial to defense, he mentions very specifically an “ariete”. Back to the OLD (my acronym for Oxford Latin Dictionary). The word actually comes from Aries, which you all know as the Ram constellation, but it’s also used to define battle rams. While it’s not technologically descriptive, it gives me a better sense of the timeline, as Gallic defenses, which are considered more “low-tech” are protecting against rams, suggesting that rams are in an earlier batch of siege weaponry than other devices like catapults or ballistas.
One of the next few steps for me is researching secondary works based on Caesar who may have done archeological or further literary research to see if the technological details I seek are there. I’ve also found more sources on wars themselves, like Florus’s summary of Roman warfare and the descriptions of the Punic Wars by Livy.
Stay tuned for the next update, which won’t be as long due to a Disney trip,
Signing off,
Aadrit T
Week 3: Secondary Sources Exploration
Welcome, one and all, to the third installment of your favorite blog series, Ballistas to Battle Rams. My name is Aadrit T, and I will be your classicist guiding you through my research journey.
Now instead of taking a trip down Ancient Avenue, we’ll be going through Modern Street. In other words, we’ll be looking at secondary sources about the primary sources of the sieges. For this blog post, we’ll be focusing on Caesar’s De Bello Gallico, covered in the previous blog post.
One of the key secondary sources I was aiming for was physical traces on the wars and sieges. For Caesar, his important presence has led to many archeological interests around his sieges, especially in what the Romans called Gaul.
One of the first papers I’ve started with was Fernandez-Gotz’s and Nico Roymans’s work in “Caesar in Gaul” [1], which is essentially a research of the genocidal properties of Caesar’s actions, especially surrounding the Eburones and Aduatuci. One of the more interesting highlights of the paper showcased evidence of lead bullets, fired via slings, used by Romans in Thuin. This gives me hope that other machinery may work out to exist, as there’s already proof of Roman Ballista bolts in Hod Hill (introduced to me by my External Advisor, Dr. Leese), which signals to me that remains of metal weaponry is not as uncommon that I may have thought it to be.
Branching off of that paper, looking through assorted bibliographies, I saw the Siege of Alesia was a major focus on recent archeological missions. Particularly, an analysis by the Madain Project [2] showcases what was left behind in such a big site. Something I learned was how Napoleon sponsored excavations of the area to build French culture, but the place is still under dubious authenticity.
Such archeology drives modern recreations of the sieges, like by Historia Civilis [3] or SLICE History [4]. While they may not be the most academic of all sources, they tie in many expert testimonials and archeological evidence cutting between the dramatic retellings. They also include sources that they have, which have useful information that I can further trace.
I do apologize if this update seems short due to the trip, but we’ll be backed to regularly scheduled action next week,
Signing off,
Aadrit T
Citations:
1. Roymans, N., & Fernández-Götz, M. (2015). Caesar in gaul: TRAC 2014, 70–80. https://doi.org/10.2307/j.ctvh1dw2c.9
2. Archaeological Site of the Gallo-Roman Town of Alesia. Madainproject.com. (2025). Editors, Retrieved on March 18, 2026, from https://madainproject.com/archaeological_site_of_gallo_roman_town_of_alesia
3. Historia Civilis (2015, April 24), The Battle of Alesia (52 B.C.E.) [Video] YouTube https://www.youtube.com/watch?v=SU1Ej9Yqt68
4. SLICE History, (2025, Jul 3) The Siege of Alesia: Caesar’s Greatest Victory I SLICE HISTORY [Video] YouTube https://www.youtube.com/watch?v=oufffrL0UME
Week 4: When Digital Resources Fall Short
Hello everyone and welcome to the fourth chapter of your favorite blog series, Battle Rams to Ballistas. My name is Aadrit T, and I will be your researcher guiding you through my project.
This week’s progress included me doing the unthinkable: going to a library. One of the key issues I found from researching online is that many of my sources fell into a middle zone. Too young to become public domain like my primary sources, and too old to be readily published online like articles and videos. Going to a physical library helps me bridge that gap.
Over the course of the week, I made multiple trips to San Jose’s own Martin Luther King Jr. Library in search of the texts I needed. One of the key reasons I chose MLK is because it connected directly with SJSU, offering a wide array of academic sources for my research.
One helpful text was the Oxford Companion to Classical Civilization by Hornblower, Spawforth, and Eidinow. It essentially acts as a large dictionary-like source for all key terms in the Classical era. Using their entries on warfare and siege technology, I was able to find out how the siege technologies were used across the ages. In particular, sieges around the 4th century BCE by Greek tyrant Dionysius I included weapons similar to crossbows and stone throwers, supported with wooden frames and metal plating. This meant that higher power, torsion-based weapons like the Cheiroballistra could only be developed during the Roman era, where metalworking allowed for a complete metal frame.
Another helpful book was Civilization of the Ancient Mediterranean: Greece and Rome, edited by Michael Grant and Rachel Kitzinger. They provided much more support on the earlier end of the Classical era, which actually falls out of my original range. It covered how Assyrian-era sieges often relied on close combat siege warfare, which included large towers to scale walls and the famous battle rams to break down walls. In concurrent with the Oxford Companion, it showed how ranged weapons were introduced during Dionysius’s time and only got better as time went on.
While these texts didn’t have the exact technical specifications I was dreaming of (even if they had great diagrams of all the civil engineering), they pointed me in the right direction.
I hope to spend one more week finishing out my literature study, which I foresee ending now that I found good points of comparison. I found interesting papers referenced in these texts, such as Eric Marsden’s “Greek and Roman Artillery: Technical Treatises”, from 1971, which I can only hope have the details I need. After that side is done, I’ll be taking you down Part 2 of this project, the engineering and modeling (the more exciting part, in my opinion).
Signing off,
Aadrit T
Hornblower, S., Spawforth, A., & Eidinow, E. (2019). The oxford companion to classical civilization. Oxford University Press, Credo Reference.
2. Grant, M., & Kitzinger, R. (1988). Civilization of the ancient mediterranean: Greece and Rome. Scribner’s.
Week 5: My Final Step of Literature Research
Hello everyone and welcome back to another installment of your favorite blog series, Ballistas to Battle Rams. You already know me, so let’s jump right into the research.
When I say I’ve found my Holy Grail, that is no understatement. E.W Marsden’s “Greek and Roman Artillery - Technical Treatise” is where all of my background research and studying has led me.
Marsden’s research was mainly centered around artillery and siege warfare, similar to what my own interests are around. His paper centers around translating the workbooks of ancient engineers, like Heron of Alexandria and Vitruvius of Rome. In a fashion similar to loebs, one half is the text in its original language, say Ancient Greek or Latin, and the other is the translated version.
These workbooks have almost all the information I have been looking for, including the dimensions of the different pieces, how the mechanisms themselves work, as well as preliminary diagrams based on ancient descriptions.
Just starting out after getting the text, I’ve been reading through the section of Heron’s Artillery manual. His descriptions go into full geometric diagrams of the bow structures of a ballista-like object, detailing the exact geometry of many of the pieces. Additionally, he goes much into detail about how they discovered the methods for torsion-based launching, and where the supposed tension strings must lie on the entire machine.
With the entire text being so dense, I am still parsing through it. But while I’m doing that, I’ll mention a few questions and important answers I had while reading.
Number 1: It’s been thousands of years. How do I know the measurements and ratios?
Answer: One of the very helpful facets of Marsden’s paper is that it includes a section that converts the Graecian units of πούς (feet) and dactyls to millimeters and inches. The same also applies to Roman Pes and digitus, as well as units of weight for both. An interesting side tangent is how both Ancient Greeks and Romans used a unit of measurement that had a name closely related to feet, etymologically.
Number 2: How will I scale my project?
In terms of scaling, it will largely be based on the space I have to work with and test. I may have to go all the way down to toothpicks, or I may pull off 1:2 ratios. If materials may cost too much, I will also have to consider going smaller instead.
Engineering pictures coming soon,
Signing off,
Aadrit T
Week 6: The Road is Clear. What next?
Hello Everyone, and welcome back to the newest issue of Ballistas to Battle Rams.
After a long spring break, I have quite a few updates.
Looking through Marsden’s work, it becomes apparent that the only method of comparison can be with similar levels of a similar technology. Siege weaponry is boiled down into 2 forms. Non-torsion engines, which can be found described by Biton, a third century BCE historian, and Heron, a 1st century CE engineer, have significantly less power but are simpler to put together. Namely, the gastraphetes, or belly-bow, is a form of crossbow that, according to Heron’s instructions, can mostly be made through wood, with smaller metal pieces that can be worked around.
It might seem that this certainly has to be the way I progress, but if you read carefully, there’s another form of engine: torsion. Instead of tension, like a bow string being pulled back, torsion relies on the twisting power of springs and coils of sinew. Either in a V-shape to look like a bow still, or a single string to act as a catapult, the overall mechanism becomes stronger. Why? I won’t bore you with the physics, but it essentially allows for more potential energy to be stored up in a string within a confined area compared to just bending or stretching.
The whole point of this project is on the comparisons of technology across the ages, and tension-based mechanisms did not have that long of a life span in innovation. Within the 4 centuries between Biton’s descriptions and Heron’s descriptions lead to arguably very similar machinery based on schematics. That wouldn’t make much of an interesting journey if the differences only came up with testing.
So, I have decided to put myself through a stronger challenge of torsion-based mechanisms. For my points of comparison, I have landed upon these. One is the straight spring engine detailed by Heron. His measurement systems are based on ratios and proportions related to artillery, so it is entirely up to my discretion. Additionally is Vitruvius’s descriptions of the Scorpio, a small scale torsion ballista weapon that needs significant scaling (measurements around 19 Roman feet at the longest end, which roughly translated to 19 Imperial feet).
Engineering is often a lengthy process, so I am setting these as my goal, with the possibility of using Philon’s engine descriptions. This will be a slow and lengthy process, so I ask for patience with updates while I decode my sources.
Signing off,
Aadrit T
Post 7: Week Seven: Can He Build It?
Hello Everyone, and welcome back to Ballistas to Battle Rams.
The work has started. As you can see from the pictures, I bought the first rounds of wood and brackets, and will shortly begin cutting and joining. Most of the other supporting beams can be found with scrap wood around the garage, and the more complicated parts will have their own process. As a teaser, you are graced with a crude rendition of the schematic in Tinkercad, minus all the important moving parts as those are a little too complex for my mediocre skills.
If you'd imagine, the two holes up top would have wound sinew coils where the arms enter. That twists and stores the energy when the arms are pulled back by the slide and handspring (the circle and square on the large beam). Again, it's crude, you'll get a better understanding later on.
But more importantly, I have been actually creating those schematics based on Vitruvius’s own descriptions. This did not come easily.
Vitruvius is writing in Latin, which means a lot of the words he’s using I should recognize. But oftentimes, he’ll use a word I am unfamiliar with. Take, “plinthide” for example. The translator uses the word “joist” for it, but that makes no sense. Joists are a term used in roofing contexts and don’t make much sense to be used with supporting columns (as the rest of the passage is about). I take a look at the Oxford Latin Dictionary, and wouldn’t you know it, the use of plinthide comes from the word “plinthis”. A word, when it comes to any column or structural support context, that is only defined in this one text, in this one section. Searching up “plinthis” does pull up the word plinth. Which is a perfectly English good word that describes how it’s related to columns and supporting the ballista. This same thing happened many times over, where the translator uses words that don’t make it clear how the pieces connect, leading to me getting lost in piecing it together.
In addition, it seems I was wrong in my initial conclusion that Vitruvius is working in the scale of 19 feet. In truth, Vitruvius puts all measurements in terms of “foramen”, or the hole where the springs are put through, so Marsden writes all units as fractions of F, for foramen. Therefore, I can completely scale this to what I need it to be, much like Heron. However, I am quickly realizing this will be much more work that I can reasonably do. I may have to pivot to only designing the Scorpio, and doing trials on different versions of it based on small inconsistencies in the translations, as suggested by my faculty advisor.
Signing off,
Aadrit T
Post 8: Blog Post Eight: Just Trust Me on This.
Hey y’all, and welcome back to your favorite blog series. Let’s get into the updates.
I’ve started construction and cutting, as referenced by the photos in this drive. I’m using a drive because the blog post system is very unfriendly to Google Pixel photos.
As you can see, the photos are of the first half of the construction, which includes the hole-carriers written by Vitruvius and the supporting beams. The engineering dimensions are one of the pieces I had the most trouble with.
In his descriptions, he makes the frame much thicker than what I am making. This is most likely because his devices are much bigger, and require more support for the tension and torsion of the coils. For the sake of my own sanity in buying materials and cutting them down to size, I chose to downscale some dimensions, making the overall frame thinner than specified, but still around the same size with respect to the arrows.
Not pictured here [I forgot to take pictures] is the recent work I did on attaching the different components. Because I chose to use thinner materials, it means that the choice of jointing (mortise and tenons) had to be swapped for a more modern solution of angle brackets on most of the corners. Mortise and tenon joints, for the non-woodworkers reading, are an extrusion and intrusion (peg and hole) that joins two pieces. Because the wood is thinner at the joint areas, cutting a true joint would be too difficult and unstable for this project. If I were in the ancient times, it would be easier because of the use of thicker wood stumps and carpentry experience higher than using a circular saw.
A little Latin Nerd out moment if you will: mortise and tenon joints are often described as male and female connectors, which is also common in electronics terminology. What I found very interesting is that the word describing this joints, at least according to the OLD, Cardo, cardines, actually has two genders. Based on which one it falls under, it specifies the type of joint. I just found that similarity very cool.
Now, if you’re getting your pitchforks on how brackets seem too weird and modern for this project (but not power tools and drills for some reason), I will tell you that Vitruvius also uses brackets and similar mechanisms on the outer walls of his Scorpio.
I already have leftover wood to start developing the slide and casing, and will work on procuring the main engine for the weapon soon. The arms are the long wooden pole from last week, cut to the length necessary. We are nearing the end of Phase 1 of the Project.
Progress is slower than I optimistically anticipated, but we’ll make it work.
Signing Off,
Aadrit T.
Post 9: Week Nine: Sin-ew
Hey everyone, and welcome to what is one of the final installments of Ballistas and Battle Rams.
Vitruvius does something very interesting with his schematics in the way he takes a lot of things for granted. One of the key components of any ballista is the slide and trigger. You’d expect something this important to have a lot of description, but if you simply read the texts at face value, it would all be rectangles. There’s not really a good description on how the trigger interlocks, where it goes in relative position, and how the slide is designed. This means I have to do a lot of groundwork in creating the design.
Take the slide for example. There’s no real talk about it other than the length, which I try to keep as faithful as possible. In terms of the trigger and claw, there’s not much other than the Greek words that are present in the Latin. I’m sure, given infinite time, I could recreate the machine to a higher degree of perfection, researching the Graecan engineers of Heron and Philon, but I must make do.
So I landed on a makeshift gear system. The arms are tied with excess artificial sinew (don’t get me started on sinew) and pulled back by a single block, with small rails to keep it straight. A small divot at a specific length is where it slides in, and a trigger holds it in place. To make it more stable, a DIY windlass (essentially a rope on a cylinder) is used for better pullback, as instructed by Vitruvius. Pulling back on the trigger releases the block, launching the block and arms into resting state, throwing whatever projectile.
Theoretically, if not for the existence of sinew. The main component of the entire build is the elastic that keeps the arms taut. Sinew is traditionally animal tendons, which store lots of power. However, the artificial sinew I have chosen to use for efficiency, came thinner than I thought. This means it takes a lot more of it than the thicker ropes you see in other builds, so I’m a little delayed in wrapping it up for use. This sucks, and it’s something I should have accounted for, but it’s more manual labor than problem solving.
Be tuned for next week’s blog post for the final results [If you’re really sneaky, you might even find it in the Drive before the blog goes live].
Signing Off,
Aadrit T.
Post 10: Blog Post 10: The final steps.
Hey everyone and welcome back to your favorite blog series. Let's jump right in.
My time has been weaning off of engineering and more towards presentation now that the final deadline is coming up. But worry not, engineering stories are still present. If you look at the drive, you can faintly see the improvement of the design up to the present.
Many things have been happening that the blog posts may not have covered. One of which is that the machine works. The arms are finally stretched and supported with metal beams, and tied with a sinew rope to the block trigger. A lot of this was improvisation as many parts aren't described by Vitruvius, but just called the Greek name. My hypothesis is that for ancient engineers, these are staple terms, like how we would understand what a pulley is without mechanical descriptions.
But in either case, the block trigger was causing the most problems. Too much stress one direction and friction would stop the machine. Too much in the other and it would fly away. My solution? Building a system of rails on the support beams, and having an H shaped block to hook onto the rails. That way, vertical motion is controlled, while horizontal is still kept in check by the larger beams.
The rope was also changed for the windlass to a less sticky nylon. One of the problems was that the windup was getting jammed because the sinew was too sticky of a material. I think, while there's no real description of this, that this is the intended solution, because it would make sense for the ropes to be of a fiber and the elastic of sinew.
Another interesting thing I found out was that ancient engineers actually had a method of testing tightness uniformly. Vitruvius writes that the machines are wound up and tested to be on the same note. As in, a tight string being strummed to compare how wound it is. I find this very interesting, especially because it's a pretty accurate system for people who didn't have much else to work with. But alas, he doesn't say what note it is, so I can't tune it properly. Nor do I think I would get to the same place.
My machine is way less powerful than what Vitruvius and engineers would like, partly because I made it that way. The sinew bands are much thinner than normal because I didn't want it to go out of control, and I was right. Even with my limited strength, it was already putting a lot of pressure on my build. With the right tools, it can be improved. but the weakness right now is optimal. But fear not. The drive has a few videos of the block snapping back to throw the pole I use as a bolt a non-negligible distance. And that, to me, is a win.
I hope to see you all on Saturday's presentation, where I will be bringing in my build for display (but not testing, because I'm not that crazy).
Signing off for the penultimate time (right?),
Aadrit T
Post 11: Blog Post 11 - Fin.
Hello everyone, and welcome to the final installment of the blog series you've all come to enjoy, Battle Rams to Ballistas.
Presentations went wonderfully, and if you missed it in person, you can view it through this recording: https://youtu.be/CbjcgyB3Ygc, and if you want the slides, they're here: https://docs.google.com/presentation/d/1rKjKf8YfUz_nh20FvhcL3scNDxk2sy7cKwsM3V8AmJc/edit?usp=sharing
Thank you so much to everyone who's helped me throguhout my project. To Ms. Odell for helping me a ton with literature review and translation work, as well as keeping me accountable. To Dr. Leese for guiding me through what combat archeology research is despite me having no experience. To my parents for helping me document and film my work (I forgot to take pictures and videos a lot), and helping me with the larger parts of building.
This project really opened my eyes to what I could use my Latin and Classics interest towards, while also keeping that engineering side of me. I saw the difficulties that came from trying to translate things that had no references and also building things on my own with little reference.
Once again, thank you to everyone who showed up and helped me throughout my project, even in the smallest ways,
Signing off for the last time,
Aadrit T.