A projectile weapon with a range of throw superior to the javelin.—PUBLIUS FLAVIUS VEGETIUS RENATUS
author of De re militari (Concerning Military Matters), written ~450
printed edition 1473
I have been a part of the throwing community in An Tir for many years. The majority of the activities that we’re involved in are modern creations based on period weaponry. I wanted to research thrown weapons that were designed to be used in combat situations, and several years ago, I discovered a weapon that the Romans used called a plumbata. My friends in the throwing community and I had never heard of this weapon and my goal with this exhibition is to share what I’ve discovered with the people of An Tir. Since I began my research, others in our Kingdom have introduced the plumbata in another modern creation by throwing it like a dart at a hard target. Research shows that this practice is not an accurate use of the plumbata. What follows is a summary of my research in this area.
What is a plumbata?
Many of you probably remember a yard game from the 70s and 80s called Yard Darts or Lawn Darts. In this game you would throw darts underhand at a ring set on the ground trying to land them into, or as close as possible to, the ring. These darts had a pointed steel tip, a large plastic fletching, and about a 4 inch grip behind the fletching.
Now, imagine a dart with a 4 to 6 inch long barbed iron tip and fletched wooden shaft with about a 4 inch grip behind the feathers. The iron tip and the wooden shaft are joined by a 3 to 4 ounce ball of lead. This is a Plumbata, a thrown weapon that Roman soldiers used with great effect during the 3rd – 5th centuries CE. The name, Plumbata, is derived from the Latin word for lead, Plumbum. The lead gives this rather small weapon added weight which increases both the distance they can be thrown and their penetrating power upon impact.
Plumbata Construction Variations
In examining images of the extant remains of Plumbata it is easy to see that there was significant diversity in the way they were constructed. These variations include length of head, point design, shape of the lead, and the cross section of the head. The metal head could be round, square, or circular in shape. The hollow, circular shafts often had the lead attached to the metal shaft rather than at the connection point between metal and wood. This was possible because the wood shaft was inserted into the head and held in place with 1 or 2 pins. The round and square heads were attached by having the lead cast around the connection point between the metal head and the wood shaft.
Plumbata Flight Tests
Since I first found out about the Plumbata, I have wondered two things: what is the best way to throw it and how far can I throw it? I have not yet finished construction on any of my own Plumbata, so I have not been able to conduct any flight tests. What I have done is research other’s testing of this weapon. I found two papers written by Robert Vermaat from The Netherlands. The first was a summary of results from his 2007 series of throwing tests. The second was from a series of tests he conducted in 2011. Vermaat tested several commercially made Plumbata for distance using both underhand and overhand throws.
I also found a paper written in 2010 by John Emery as part of his Bachelor of Arts of Archeology for the University of Wisconsin-La Crosse. He also considered the distance of flight for both over and under hand throws. For his tests he manufactured his own plumbata based on images and descriptions from period sources. Unlike Vermaat, he did not throw the weapons himself. He recruited, and trained, a group of volunteers who threw the weapons so he could focus on making observations and taking measurements.
These researchers used Plumbata that were different in size and weight so the results they produced vary greatly but there were some similarities. Both found that underhand throwing produced longer throws and required much less energy. They also found that, with practice, greater distances could be achieved.
In Vermaat’s first series of tests he found that in all cases, and with all weapons, underhand throwing produced the greatest distances. These differences varied by from 4 to 9 yards depending on the particular weapon used. His average distance for underhand throws was 35.9 yards, (108 feet). For overhand throws the average was 32.6 yards (97 feet). In his second set of tests in 2011, with different weapons, his averages were 35.4 yards underhand (106 feet) and 32.6 yards overhand (98 feet).
Emery conducted two sets of measured tests after several practice sessions. During these initial sessions he discovered that the overhand throw produced results significantly different than the underhand throws. These initial tests were not measured but he estimated the overhand throws averaged somewhere between 110 and 120 feet based on his pacing of the distances. This was 40 to 70 feet shorter than the underhand throws so he decided to end testing of the overhand method. His first series of tests produced an average distance of 170.9 feet. The second set resulted in an average of 173.7 feet and included multiple throws over 200 feet. Emery also noted that his longest throws resulted from the smoothest releases which produced less in-flight wobble. These were also the last throws that were made after the throwers had the most practice.
For his trials, Vermaat used Plumbata that were longer and heavier than Emery. The fletching on his commercially made weapons was disproportionately larger than those used by Emery and this increased drag may help to explain some of the differences in distance achieved.
Plumbata Archaeological Finds
The Plumbata is a late period Roman weapon. Most of the extant finds have been in or around Roman fortifications active during the late antiquity period. The earliest dating of weapons found that can, with certainty, be known as Plumbata was in the 270s, with most finds being dated from the 4th and into the early 5th centuries. As of yet, no Plumbata have been found in Roman fortifications that were abandoned before the last third of the 3rd century. The latest finds so far have been dated as late 5th or early 6th centuries.
These weapons have been found over a vast area ranging from Britain, Germany, Austria, Slovenia, Hungry, Greece, and as far as Georgia. They’re use seems to have been centered in and around the Roman province of Illyricum, which corresponds to parts of modern day Albania, Montenegro, Bosnia, Herzegovina, and Croatia. In this region were two legions of 6000 men each that were described as preferring the Plumbata over all other weapons. These legions were given the titles of Jovian and Herculean by the emperors Diocletian and Maximian who reigned together from 284-305 ce. These legions were the preferred troops of these emperors due to their extreme prowess with the Plumbata. The use of this weapon seems to have spread out from this central region.
A friend asked if the location of the extant finds of Plumbata coincided with army on army battle locations that Rome was involved in from the era that the finds were dated from. In comparing a timeline of Roman wars with Plumbata find locations, here is what I found. A note, there were many small skirmishes and conflicts which took place in and around Roman held territories which I have not taken into account for this post since the question focused on army on army wars.
When Diocletian was named emperor in ad 284 there already existed 2 legions noted for their use of the Plumbata who were stationed in Illyricum where some of the earliest finds were located. These were dated around 270 ad so I began my search here. In 279 Probus launched a campaign against the Vandals in Illyricum. It seems probable that the legions used Plumbata in these battles and their fame for their use began here. In 285, in the region of the Morave River east of Illyricum (present day Serbia), Diocletian defeated Carinus. Here again, it is most likely that Diocletian would have used his best legions (which he had named Jovian and Herculian a year earlier for their skill with Plumbata) as his primary troops. In both of these regions Plumbata have been found.
There have been several Plumbata finds in Britain where, between 286 and 296 the Carausian revolt took place. Many have also been found in Germany where, in 357, the Battle of Strasbourg took place between a coalition of Germanic tribes known as the Alemanni and Emperor Julian. Others have been found throughout much of Southern Europe where Roman troops were stationed during this time.
It is worth noting that most of the Plumbata found have been at the sites of Roman fortifications. These were built in areas where conflict was either expected or taking place and then abandoned when the conflicts were resolved. So, to answer my friend’s question I would have to say that there is definitely a direct correlation between areas where major battles took place and where Plumbata have been found. My research also seems to indicate that Plumbata use spread from the region of Illyricum outward to the rest of the Roman Empire based on the dating of the finds.
The website Roman Army Talk has a list of extant Plumbata heads (from page 44 in the Plumbata thread, moderated by Robert Vermaat): Currently there are 175 published finds: 31 from Serbia, 30 from Britain, 16 from Slovenia, 15 from Italy, 15 from Austria, 14 from France, 10 from Hungary, 9 from Croatia, 7 from Germany, 7 from Switzerland, 5 from Georgia/Abchasia, 4 from Romania, 3 from Bulgaria, 3 from Greece, 2 from Liechtenstein, 2 from The Netherlands, 1 from Belgium, and 1 from Slovakia. Added are 88 from doubtful or unprovenanced origins (up from 80) for a total 263.
Creating a Plumbata
Since I first discovered what a plumbata is, I’ve wanted to construct my own for several reasons. I wanted to be able to throw them and at the time couldn’t find any place to purchase them. As I read more about this weapon it became clear that there were many disagreements on such things as how they were made, how long they were, what they weighed, and how far they could be thrown. I believe that the only way I could recreate this Roman weapon was to literally create them myself and adjust these variables to make as effective a weapon as possible.
In the extant versions that have been recovered there is a great deal of variability in their construction methods. I believe this to be due to a number of things including: the time period when they were made, who was making them, the available material on hand, the skill of the maker, and their intended use. My current goal is to create a weapon that will be thrown by hand and achieve maximum flight distance. To do so I will be adjusting the weapons length and weight as well as fletching size, number, material and placement on the shaft. Skills I will need to develop include learning how to forge the iron heads, make molds and cast the lead onto the shaft, and how to fletch effectively. I’m in the infancy of this learning process, the following information describes my initial attempt.
Forging Plumbata Heads
I had an opportunity to work with Master Grendel to learn to forge iron heads for plumbata. I want to start by saying this man is an amazing teacher. In the few hours I spent with him I learned more about metals and forging than I had in the other 59 years of my life. What follows is a description of how we made a barbed heads for plumbata.
We started with a piece of iron flatstock that was approximately ¾” by ¼”, heated it, and began to shape the shaft. We wanted a ¼” by ¼” shaft. In the photograph, you see us using a power hammer to do this initial shaping. This was done simply to speed up the process, the rest of the work was done by hand. We next cut the barbs by driving the heated bar onto a sharpened piece of hardened steel that was clamped into a vice. Once the barbs were formed we reheated the bar and cut it off just beyond them. Shaping the point was more complex. It required hitting the piece at an angle without the edge of the hammer hitting the anvil or flattening the barbs back against the shaft. The last steps were to make sure the shaft was straight and quench it. What we produced was a fairly decent rough cut barbed plumbata head. With some work on the grinder it could easily be sharpened and made into an effective weapon.
We made 2 barbed heads that day. I watched, learned, and photographed as Grendel made the first one. He then asked me if I wanted to try it (of course), handed me a hammer (not his good one, he didn’t trust my skill), and talked me through the process. The photograph of the completed head is my first attempt not just at making a plumbata head, but at forging as well. Again, I have to thank Master Grendel for sharing his knowledge and skill.
To construct my own plumbata I needed to learn how to cast lead which is used both as a weight for the weapon and to strengthen the connection point where the wood shaft and metal head came together. Pure lead melts at 621.5f, a temperature low enough to melt over a wood or charcoal fire. I decided that, for my first attempts, I would purchase a modern pot to melt the lead in but as I learn more about the process I’ll try it over a fire. I began with some old fishing weights that were my father’s before he passed away. These weights were pretty oxidized and dirty so I expected them to put off a lot of fumes which are very dangerous. I was not wrong on my assumption. I set up the melting pot outside on a covered deck so I was able to place the lead in the melting pot and go inside the house to watch the process safely away from the fumes. Once the lead had melted and the smoking stopped it was ready to pour.
I was a little concerned about being able to get the rings that were molded into the fishing weights to tie the line to out of the lead before pouring it into the molds. I found that this was not a problem. The lead was denser than the rings so they floated on top with the rest of the slag. As I poured the molten lead into the molds the slag and the rings stayed in the pot. Once all the lead had been poured I just had to invert the pot and tap it gently and this waste simply fell out.
In my research I had discovered that it was possible to pour molten lead directly into a mold to connect the wood and metal parts of the plumbata. I wanted to test this theory so I used a splinter of wood to stir the molten lead before pouring it trying to push the slag away from the pouring spout. After using this same stick about 20 times to do this there was just a little charring on one end. This seemed to indicate that it is possible to pour the molten lead into a mold connecting the two parts of the plumbata without destroying the wood shaft.
As you can see in the images, I used an old aluminum mini muffin tin as a mold to cast my purified lead. I wasn’t sure how well this would work but after using it there was absolutely no apparent damage or discoloration. This is important because one theory on making plumbata is that 2 part copper molds were used to form the lead. Copper has a higher melting point than aluminum so it is feasible that it could have been done this way without destroying the molds.
Assembling the Plumbata
Experiments Throwing Plumbata
I’ve wanted to throw a plumbata ever since I first learned of them. I purchased two smaller versions of the weapon and anxiously awaited their arrival. When they were delivered I was disappointed that one was broken in the box so I only had one to throw. My first attempt to throw it met with limited success. This weapon, as described by the anonymous author of De Rebus Bellicis in the fourth century A.D., has, “at the lower end of the javelin are fixed flights to give speed, with enough space left above these flights, of course, for the fingers of the holder to be able to grasp” (Emery, 2010, p.5). The weapons I purchased only had 1.25 inches of shaft beyond the fletchings to grasp, not even enough room for a secure grasp with 2 fingers. When I tried to throw the weapon by just holding this end the results were erratic at best making a smooth release with force virtually impossible. My best throws were achieved by lightly grasping the weapon around the leather fletchings, which is not at all how the weapon is described as having been used. I believe this to be a design flaw. I eventually broke the second of my weapons and will rebuild both with slightly longer shafts to allow for a better grip.
I threw the plumbata both underhand and overhand. My research suggested that an underhand throw would produce greater distances and this proved to be the case. My test consisted of 10 underhand and 10 overhand throws. The data tables show the distances I was able to achieve. In both cases this distance increased as the test went on. I attribute this to my muscles warming up during the trials and to the increased experience I was gaining with each throw. The data for the underhand throws shows a great deal of variation. This occurred because I changed how I gripped the weapon several times. I found that I had little control when I gripped the weapon with two fingers behind the fletchings and these throws produced the shortest distances. In future tests I will use a single grip and release method to try to eliminate much of the variability.
As I stated earlier, my current goal in creating and throwing plumbata is to achieve maximum range with the weapon. I believe the distance they can be thrown also indicates their overall effectiveness. My thought is that, the further they can be thrown, the more you can throw before an enemy comes within hand to hand fighting range. The quote form the Roman historian Vegetius, at the top of this exhibit, indicates that they were utilized in this way, as a weapon used at distance. As I began my throwing trials I wanted to attempt both over and under hand techniques to compare average distances achieved. Others have published throwing trials (Robert Vermaat in 2007 and 2011, and John Emory in 2010) with their results indicating that an underhand release always produces greater average distances. My initial trial showed the same results.
With an underhand throw you not only can achieve greater distances, but it is easier to achieve greater height as well. This is important when you consider how this weapon was used. When throwing something like a spear, or pilum if you’re a Roman soldier, the throw is basically straight on and can be blocked or intercepted by a shield held in front of a defender. The plumbata is thrown with a higher arc meaning it will drop at a steeper angle. In effect, it can drop behind a shield and still hit the attacker. Against a cavalry charge, they could drop behind the rider hitting the horse in, what was usually, an unarmored area. Plumbata were also used by defenders in seige warefare. From inside a fortified area they could be thrown over walls without the person throwing them ever being in harms way. One last benefit of throwing underhand I will mention is that an underhand throw is more natural and causes less stress, and damage, to the body of the thrower. A baseball pitcher usually has to rest his arm for several days before performing again. A softball pitcher, who throws underhand, is usually able to pitch day after day. A soldier with a bad shoulder cannot be as effective in hand to hand combat as a healthy one.
It is possible that plumbata were sometimes thrown overhand. To the best of my knowledge, and in the research I have done, I have found no indication that this was their intended use since their range and effectiveness would certainly be reduced. One can imagine, however, that a soldier, with plumbata in hand, might have thrown it overhand at an approaching enemy before drawing a hand held weapon. It is even probable that they were tossed from the top of fortified walls onto the enemy below with an overhand throw. In my opinion, this could not have been a common practice nor one that soldiers would have trained to do.
1200 inches = 100 feet
|TRIAL 1 DATA|
- As I am not as young as I once was, before any future tests I will need to take the time to stretch and warm up to protect myself and eliminate some of the testing variability.
- Although I was not making any attempt to be accurate with my throws, all of them, except for 2 that were badly mis-thrown, fell within just a few feet of the tape measure I had laid out. This indicates that with further practice a great deal of accuracy should be possible.
- Greater distances can be achieved by throwing plumbata underhand.
- A smooth release, which produces far less in flight wobble, also results in the greatest distances being achieved. I believe this wobble will be decreased as I practice with the weapon.
For my second throwing trial I rebuilt both of the plumbata I had purchased and made a few changes to their design. The most important change was adding 2 inches to their length by extending the distance between the fletching and the end to provide a better grip. In the first trial I couldn’t effectively grip the weapon behind the fletching because there simply wasn’t enough space. The throws I attempted by holding the plumbata this way, as it is described in my research, were very erratic and hard to control. To make this change required me to construct a new wooden shaft. I choose to use an oak dowel thinking that it would be a lot stronger than the rattan that was used in the original ones that both broke fairly easily. I attached these to the plumbata heads and reattached the fletching on these new shafts.
Another change I made was to change the number of fletching pieces. When I bought them, they used 4 leather pieces as fletching. I changed this to 3 on one of them thinking that the decrease in drag might increase flight distance. The plumbata with 3 failed after the second test throw so the data set is too small for any real conclusions. However, when I was warming up by throwing both weapons, in a very relaxed way, they would land within a few inches of each other. This seemed to indicate that the fletching number wasn’t a significant factor in their flight distance.
When throwing for distance, I noticed immediately that the longer grip made a difference. My average distance increased as did my ability to control the weapon’s release. As in the first trial, I had some difficulty getting consistent releases. On throws 2 and 9 my release point was too high, and on throw 10, it was too low to achieve maximum distance. I believe that with more practice this problem would be solved. A concern I had when redesigning the weapons was that the fletching might not be far enough back along the shaft to effectively stabilize them. When throwing, I noticed that on every release the back end of the weapon wobbled up and down through much of the flight. I believe this decreased the distance that could have been achieved and may have been caused by the fletching placement. Before my next flight test I will again increase shaft length but this time I will be increasing the distance between the weapon’s head and the fletching hoping that this will increase its in flight stability.
To me, the angle of impact is an important aspect of the plumbata. They are designed to strike at a steeper angle than a spear or an arrow. This allows them to strike behind a shield or other frontal defense. I believe this to be one of the primary advantages of the plumbata over other forms of missile weapons. The above picture shows a typical angle of impact for a plumbata landing.
1440 inches = 120 feet
|TRIAL 1 & Trial 2|
|THROW||TRIAL 1||TRIAL 2|
- The added grip length increased distance and the ability to control the throw.
- There was no noticeable difference in distance between the 3-fin plumbata and the 4-fin plumbata. This was a surprise to me.
- The next plumbata I construct will have a greater distance between the head and the fletching. Hopefully, this will decrease the wobble.
- More practice should increase both distance and consistency.
My third throwing trial was with the plumbata I created. This weapon was similar to the previous ones I had thrown, since I had used the one of them to make the mold I cast the lead in. I also used the same type and size of wood for the shaft. There were several differences, however. I made the shaft 2 inches longer because I wanted the fletching to be further from the head thinking this might lessen the wobble I had experienced in trials 1 and 2. The tip on this weapon was not barbed. I was hoping the weapon would not fail as all my previous ones had. With a barbed head it was difficult to pull it out of the ground putting a great deal of stress on the connection point between the head and shaft. I also used feathers for the fletching on this one.
With my first throw, I noticed that I could hear the air passing over the fletchings. This caused me to wonder if the fletchings were too large, thus causing too much drag. I noticed that the weapon landed at a steeper average angle. This was probably because, due to drag, the weapon slowed more in the air causing a steeper descent. The average distance thrown in this trial was less than in trial 2, likely also due to the increased drag of the larger fletchings.
The fletchings used on this plumbata were a right-wing feathers. This caused the weapon to spin in flight, which greatly reduced and/or eliminated the wobble issues that I had in my first two throwing trials.
I had less variation in throwing distance in this trial, than in the previous two trials. I attribute this to better weapon design and more experience with the weapon. Also, the distances gradually increased as I became more familiar with this particular weapon.
1356 inches = 113 feet
- The fletching was probably larger than it needed to be in order to achieve maximum flight distance.
- The spin caused by the fletching eliminated the wobble that occurred during earlier tests.
- The non-barbed point was much easier to extract from the ground. This caused the weapon to NOT fail, unlike previous attempts with barbed plumbata.
- Throwing underhand uses a different muscle group than throwing overhand (like other thrown weapons). All 3 throwing trials ended with a pulled muscle. More practice is needed by this thrower.
- I want to change the design of plumbata to test flight distance. This includes different weights, different shaft diameters and lengths, different fletching sizes and shapes, and different tip styles.
- Casting – I need to learn more about different mold styles for casting lead. And, I need to learn more about casting, in general.
- I am interested in further research into Roman Army tactics and how missile weapons would have been utilized.
- I would like to continue to research the period use of thrown weapons in combat.
- Practice – the more I throw, the more consistent I am with placement and distance, and the further I am able to throw.
Plumbata – Teaching
- Emery, J. (2010). Experimenting with plumbatae and observations on their behavior. [Unpublished bachelor’s thesis]. University of Wisconsin-La Crosse.
- FECTIO. (2011). Retrieved March 02, 2020, from http://www.fectio.org.uk/fectio.htm
- Intercisa Múzeum. “RÓMAI PLUMBATA BARACSRÓL Intercisa Múzeum Dunaújváros.” Intercisa Múzeum, Retrieved from http://intercisamuzeum.hu/galeria/romai-plumbata-baracsrol.
- Notitia Dignitatum Imperii Romanii. (1436). (Lamy, P., Illustrations). Retrieved from National Library of France. Manuscripts Department, view 119, folio 58r. (Original work published ca. early 400s AD)
- Roman Army Talk. (nd). Plumbata. Retrieved from https://www.romanarmytalk.com/.
- Vermaat, R. M. (2007) Testing late Roman plumbatae 1 – Veerse Dam 2007. Retrieved from https://www.academia.edu/30544051/Vermaat_Robert_M._2007_Testing_Late_Roman_Plumbatae_1_-_Veerse_Dam_2007
- Vegetius. (2019). De Re Militari: Complete Official Edition (J. Harper & L. Adet, Trans.). Dubai: Harper – McLaughlin – Adet Publications.
- Vermaat, R. M. (2011). Testing late Roman plumbatae 2 – Breezand 2011. Retrieved from https://www.academia.edu/30545939/Vermaat_Robert_M._2011_Testing_Late_Roman_Plumbatae_2_-_Breezand_2011
- Vujović, M. (2009). The plumbatae from Serbia. Journal of Serbian Archaeological Society, 25, 203-208.
- Class: The Plumbata, Online ~ (April 2020)
- Thrown Weapons Champion, Barony of Glymm Mere ~ (June 2017 – June 2018)
- Class: Spear Throwing, Barony of Glymm Mere ~ (July 2015)
- Thrown Weapons Champion, Barony of Glymm Mere ~ (June 2015 – June 2016)
- Thrown Weapons Champion, Barony of Glymm Mere ~ (June 2014 – June 2015)
- Kingdom Protector, An Tir ~ (July 2013 – July 2014)
- Guardian (Glymm Mere) ~ June 2014
- Thrown Weapons Champion, Barony of Glymm Mere ~ (June 2013 – June 2014)
- Thrown Weapons Champion, Barony of Glymm Mere ~ (June 2012 – June 2013)
- Thrown Weapons Champion, Barony of Glymm Mere ~ (June 2011 – June 2012)
- Class: Spear Throwing Basics, William Tell Accademia ~ (August 2010)
- Class: Introduction to Thrown Weapons, River’s Bend Defender’s Tournament Accademia ~ (June 2010)
- Thrown Weapons Officer, Shire of River’s Bend ~ (August 2009 – present)
- Thrown Weapons Champion, Barony of Glymm Mere ~ (October 2009 – June 2011)
- Class: Introduction to Thrown Weapons, William Tell Accademia ~ (August 2009)
- Thrown Weapons Champion, Barony of Glymm Mere ~ (June 2008 – October 2009)
- Thrown Weapons Officer, Barony of Glymm Mere ~ (January 2008 – August 2009)