[ Hello Slinging! ] Zwiebeltuete's Slinging

My experiences and skill

Important Information

Slinging is dangerous! A projectile thrown with a sling can kill a person. Slinging is especially at the beginning dangerous because there is no muzzle which points to the direction where the projectile will fly. The projectile can fly in any direction. With some styles the projectile can be thrown vertically up in the air so that you can effectively hit yourself. For additional information see for example the Sling Safety information.


Once I found www.slinging.org. I do not remember how, probably by accident. I got interested because of the sling's simplicity, so I made myself a sling.

I slung my first stone on 2005-09-24. I used for that a self-braided 60cm sling made from hemp yarn. The stones collected on an acre flew around 10-20 meters. The style was helicopter with one or two rotations.


Why do I sling? The fascination of the sling is its simplicity. A rope and three knots are enough and you have a sling. You can take it everywhere, although can't use it everywhere because of safety. And its fascination for me is the challenge to master this difficult instrument. Golf is a challenge for your whole life, but slinging is a challenge for your next three lives.


I use two types of slings:

For the braided sling I use sausage yarn (?, Wurstgarn) which is probably made from hemp. These sling look very nice, probably even like slings the ancients used, and work I put into them makes them special to me. Interesting is, that this type of sling is very silent. Here is a photo of my braided slings (The first sling is my first ever made sling with which I started.):

[ Braided hemp slings ]

For the knoted sling I use PP rope. It is very fast to make.


I tried many styles:

I never was any good with Figure-8 and do not practice it at the moment. All others I do practice. Underhand because it has a totally different flight path than the others. Greek Style because it seems to fit the best with documents about ancient slinging - also textual sources. The most practice I do at the moment with Apache Style. Nice about that style is that it is fast and (relatively) accurate.


From the distance of 10-20 meters I went up to 75 meters with 30-50 g marble pebbles in March 2006, with Apache Style and Greek Overhand. Unfortunately my distance went down to maximum about 45 meters thereafter. I did feel that I was not putting force to the projectile. So at the moment (April 2006) I ignore distance and started to sling for accuracy. I do now lazy slinging and also get up to 45 meters distance.


In April 2006 I started to sling for accuracy. I aim for a tree trunk (diameter ca. 25cm) at a distance of about 30-40 meters. Two things helped me to improve my accuracy:

Now in May 2006 I do not think the last point is helpful. So I just go with the first point. I tried to hit the tree trunk at a distance of about 6-7 meters (10 paces). Of my last 18 throws 3 hit it (17%) and 4-5 were near misses. (All with low power Apache style)

Today (2005-05-21) I again tried to hit the trunk at the same distance of 6-7m: 1 hit out of 33 throws. Not really good, but I had the whole last week no time to sling.

On 2006-06-09 I had 2 hits out of 50 throws with same conditions. Not really better. I think all of my throws were in an arc of about 30°. My most common error was 15° to the right high in the air. My second most common error was 15° to the left 5m in front of me into the ground. Many were also in an arc of 5° around the trunk, but that's not good enough to hit.

2006-06-11: 1 hit out of 55 throws with same conditions. I think I have reduced the common deviation to the right with the first point. The single hit was when I tried the second point. I am not sure if I improve or if these are just random performances. Help me!

2006-06-13: Same procedure as always. 2 out of 50. That doesn't look better, but I have the feeling to get better. It looks to me the deviation from the target has gotten smaler. I will try to make a video of me slinging. Maybe someone can give me hints based on that.

2006-06-14: Same procedure as always. 0 out of 50. Looks worse, but effectively all stones were in an arc of about 30°. So it is within the error range of the last days. I'll try to get tomorrow some movies of me.

2006-06-15: I have made a video of my Apache style throws. Maybe somebody can give me some hints on what to improve. The throws with large deviation are at the end. (apache-060615.avi, DivX, 4MB, apache-060615-cinepak.avi, Cinepak, 7MB, apache-060615.mpg, MPEG1, 4MB)

From this video(s) I got very helpful comments from Dale and siguy. In the next (and following) days I will try to follow these.

2006-06-16: I tried to follow the comments with success. 3 out of 50 hits on the trunk and maybe 5 or 6 in the branches above. The things I changed are: only little movement of my body ahead (I think this improves stability and therefore accuracy), turning of the hip, throwing arm hanging straight down, turning of the throwing arm more vertical, non-throwing arm on body. (I will try to make another video on the weekend.) With all that the power behind a throw increased, although without a quantitative measure. And I have the feeling my accuracy increased.

2006-06-17: 2 out of 10. All probably in a 15° arc. Have to change slinging site.

New site. A bush of size 60x80cm at a the same distance (10 paces, about 6m). Ammunition are now lumps of shale. 6 hits out of 50. Next weekend I will hit a pinhead at 140 meters. At least in my dreams.

2006-06-18: I have made another video of my Apache style throws, now IMHO with a much improved technique. (apache-060618-cinepak.avi, Cinepak, 3MB, apache-060618.mpg, MPEG1, 1.9MB)

Today 4/50 hit the bush. Nothing remarkable.

2006-06-23: Today I slung in the evening in the middle of biking. I didn't do that long because of gnat, only twelfe stones.

2006-07-10: The last weeks I slung in between biking with gnat repellent spray. Today I aimed at a tree trunk about 3 or 4 meters away. I never hit it, but very often I missed it by about 20cm to the right. I widened my feet position which gave me more stability, I think. And I tried beginning to turn my hip when my sling arm was nearly horizontal, but my weight shifting to my front foot began from the start.

(For linking: I have put in each paragraph with videos an anchor for easier linking to the videos. That way you have all video formats available together in one link.)

Making a braided sling

The article about sling braiding is a very good step by step guide for braiding a sling yourself. At the moment I want to add here some points which I learned and might be interesting for beginners:



The sling in Roman times

Second century C.E. at the German Limes

This is the time the COH IIII VIND has set their frame of time. Therefore I want to write something about the situation there and then. The following is the impression I got from the article from Völling.

But first some information on the military background at that time. In preceeding times there were big battles with big armies on open ground. The battle ranks were very close and the mobility of the soldier had not to be high. That was the age of the short gladius and big square scutum. In the second century battle of that kind happened no more. Smaller units had to be brought quickly into action and fought more openly on less open ground, maybe in the woods or in villages. There the gladius was to short and the scutum was to heavy. The result is the soldier with a long sword, smaller oval shield and chain mail.

This change of combat scenery had certainly also an effect on the usability of the sling. The advantage of the sling lies in the distance. It probably could outrange the bow. I am sure slinging distance of good units was beyond 300 meters, maybe even much more. (The modern world record using a stone is about 440 meters. At www.slinging.org are normal people with distances of beyond 200 or even 300 meters. Acient slingers certainly had better training with the same equipment.) Sending a large quantity of stones or lead glandes to an enemy unit in close formation probably gave certain hits. In the second century a soldier had the chance to sling once at a single enemy soldier before they got into close combat.

The information in the article supports in my opinion the decline of the importance of the sling. In earlier Roman times the historians mentioned the utilization of slingers in battles. This is no more the case in the later Roman times. This does not mean that the sling was not used anymore. There is still archaeolgical evidence, that the sling has been used, found slinging ammunition. But what we do not know is how much the sling was used in combat, as the sling was also often used for hunting. (The ancient Aegyptians probably used it for hunting, certainly the Apache Indians and on the carpet of Bajeux there is a depiction of a slinger hunting birds.) From the archaeological information I think there is as a result of the reduced important a decay of the art of slinging. In earlier times slings are shown to be strings and the projectiles were often almond shaped and made from lead. In later times only slings made from wider straps are shown. These certainly were less effective because of air drag. Also the projectiles seem to be more and more made from clay and made more spherical shaped, which has a higher air drag than almonds. (The reason for less might also be that the slings were more used for hunting.)

So what could have been the situation in the COH III VIND? Other Celts are known to have known and used the sling, e.g. Caesar's de bello gallico (aka 'the gallic dog' or in German 'der Gallische Hund'). So probably at least some people of that Celtic tribe knew how to handle the sling, e.g. because they hunted with it. These people certainly would have continued to use the sling as auxiliars. It was a free additional weapon in battle. And probably the officers have supported these people by providing supplies.

Thomas Völling's article in Saalburg-Jahrbuch 45

This is a very good article about the sling in the Roman army which looks at archaeolgical and textual sources for analyzation of used slings, slinging technic, projectiles, employment in the army, distribution, projectile manufacturing. This is definitely a must have for all people interested in that topic. Unfortunately it is in German and the Saalburg-Jahrbuch is not widely available.

Dietwulf Baatz's article in Saalburg-Jahrbuch 45

This article is very interesting to get a feeling for the different performances of projectiles made from lead, stone and clay, but he does make to much or false conclusions from them. Here I want to comment on these.

Baatz has some variables in his calculations which are only guessed. One of these is the drag coefficient. He writes that a sphere has a cd of 0.48, but comes esp. at the beginning in the area of critical velocity and therefore he assumes a cd of 0.4. Achenbachs graphs showing the cd value of spheres with rough surfaces show curves with cd = 0.5 at low velocities and cd sometimes smaller than 0.1 at medium/high velocities. If we now assume an average of cd = 0.3 the clay ball does not only fly 200 meters, but 233 meters. If we assume that most of the way the ball is at critical velocities (Speed does not drop far below critical.) and therefore cd = 0.2, than the clay ball even flies 284 meters. That is a possible error of over 40%. The slingers, esp. at the early Roman and Greek times had mostly not balls, but almond formed glandes like they are shown in some photos on this page. I have no reliable information about their cd, but it sure is even lower. I think the determination of the cd would be a nice job for a physicist. (Compared to the error in cd the correction of air drag according to air density he has in his calculations is only about 9%.)

Baatz assumes a starting velocity of 75 m/s. This is a possible value, but still there is a huge error possible. More lazy slinging to save your power for more slinging may have been 50 m/s (clayball: 133m, -34%). Yurek, a good slinger on slinging.org, measured about 100 m/s (clayball: 251m, +26%). They might have had even higher velocities for long range, e.g. 125 m/s (clayball: 291m, +46%). To get here good values is probably much more difficult than for the cd value as it is influenced a lot by training and intend.

When applying both the range of cd and the range of velocity the clay ball can fly in our calculation between 133m and 459m (-34% - +130%). Therefore the absolute data of distance and impact energy (-23% - +116%) seems to me not really usable.

Now to one of his conclusions, the effective range. He compares slinging to the bow sports. There is the largest distance at tournaments 90 meters, so he writes. (This is not correct as the Korean archers shoot at targets 140 meters away.) Hence he assumes that the maximum distance where it is possible to hit a target is 100 meters, which is therefore the maximum military effective range. Now this is the case - as he writes - unless you want to hit larger targets, e.g. large groups of people. The historical reports of using the sling in war is where tens of thousands of people fight each other with smallest unit sizes of about 100 people. I am sure these qualify for larger targets. So assuming a maximum military effective range of 100 meters according to marksmanship seems to me much to low.

Because of impact energy he further reduces the 100 meters maximum effective range to 30 - 60 meters, according to the author the most common range, which might drop even to a minimum range of 20 meters. Now would I as a for close combat (nearly) unarmed person go as close as 20 or 30 meters to armed infantry? That is nearly (?) in range of their pila and behind them might be archers or cavalry which might attack me soon. My answer is: No. I would want to use the advantage of the sling for me and shoot from a save distance where no arrows could reach me, even when I do not really kill the enemy soldiers but just wound them.

There is an ancient report of the Rhodians being able to use lead glandes, but the Persians not. The first reason Baatz gives is a good one. He writes that the Persians just were not trained in using the lead glandes. His second reason is that the Persian slings would break when using lead glandes. All the projectiles weigh the same so the pull on the cords of the sling is the same, but pressure (force per square-cm) is according to Baatz a lot higher on the pouch when using lead glandes as these are smaller. Now if you take a thread which is able to hold 100g, lay it onto the floor and step onto it, does it tear? No. So the pressure is not relevant. Some of that might transform to a pull on the cord on tiny places around the projectile itself, but the sling should have been constructed to have a safetymargin of at least 100%. So in my opinion this second reason is not valid. A possible reason the Persians did not use lead is that they just did not know about it until the lead glandes were flying towards them. (funda_iucunda from slinging.org has the same opinion. See his posting.)

Despite of all these issues in his article it is nice to see the effect of changes on the flight path of projectiles. What happens if material, speed, air drag, ... changes? For better understanding these effects yourself you can download on this page a Perl script which implements the physic Baatz used in his article.

Insinuating usage of the word 'funda'

This is not about real slinging and not even in Roman times, but I think this is worth mentioning. In the medieval comedy 'Babio' written in the twelfth century it is written (338):

captus non totus abibit
mecum deuenient funda petraeque simul

Here 'funda' is a metaphor for penis and 'petrae' for the testicles. (J.N. Adams; The Latin Sexual Vocabulary; 1982; p.22)

Recreated ancient projectiles


I try to recreate ancient slinging projectiles. The cheap form of ammunition is the projectile made from clay. These haven probably often used for hunting. A common form for these was the clay ball which has probably been hand rolled. So I digged behind the house to get some clay, formed balls and baked these in the oven at estimated temperature of about 900C. (I don't know the exact temperature because the thermometer obviously showed a wrong temperature.) The following photo shows an unbaked (left) and a baked clay ball.

The following photo shows an unbaked (left) and a baked clay ball.

[ Comparision of unbaked and baked clay ball ]

I also opened a defective clay ball after baking (following photo). You can see there a little dark stone and a larger lighter gray area. Probably the iron in the clay has not been fully oxidized. Maybe the baking time of an hour was to short, but longer baking times do not improve the durability of the clay.

[ Opened baked clay ball ]


More famous than the clay projectiles are the lead glandes. Here is an original ancient glans of unknown origin.

[ Ancient lead glans ]

I have produced some new lead glandes using plaster molds. For the future I want to make a clay mold like it was used in Roman times. The glandes I made so far weigh 51-56 g. On the next image one of these is shown in detail. It has an inscription which has been scratched into the mold. I asume it was done the same way in the classics, although inscriptions were not common. (The inscription is 'CAEPIO F'. Caepio is a real Roman cognomen meaning 'onion'. F is the abbreviation of 'FECIT' meaning 'he has made (this)'.)

[ Selfmade lead glans ]

On the following photo you see a glans and a baked clay ball. When people see them they think the clay ball is heavier. When they take them in their hands they think the lead glans is heavier. In fact both weigh about 52 grams.

[ New lead glans and baked clay ball ]


Romans had also sharpened stone projectiles formed like glandes, but I am really not interested in fiddling with stones.

Reconstruction of clay molds for lead glandes

Two Roman molds for lead glandes are known to me from Völling, both are made from clay. The one I want to reconstruct is the simpler mold where the glandes are connected on their side togehter.

I again collected clay from my favorite clay source behind the house. I did clean two batches by different methods. I kneaded the first batch thoroughly and removed in that process the noticable stones. To the second batch I added a lot of water and sieved it. For the production of the mold I used the first batch because sieving is a lot of for and for the mold I need larger quantities.

For the glandes found at Tauromenium in Sicily (Voelling, Abb. 27) only one-halve molds have been used, i.e. there was one mold with a cavity in which the lead has been poured. The other halve of the glans was the open surface of the poured lead. For making such a mold I made a male form of the bottom of such a glans by carving from wood. I pressed this male form multiple times into a prepared flat clay chunk thereby resulting in multiple identical cavities of the correct form. After drying of this mold glandes like these found at Tauromenium could be easily casted. The only difficulty was getting the right amount of lead into the mold, but after some tries this went well too. The optimum amount for that mold is in the range of 45 - 50g per glans, but 30g still results in an acceptable glans.

[ One-part clay mold ]

I have read that clay is shrinking when drying. So first making one halve and after drying forming and fitting the second mold is not possible as then after drying of the second mold it does not fit anymore. When doing both halves at the same time there is the problem of forming them properly to the glandes and them sticking together. One possibility is to go with intermediate plaster forms where the clay is fitted to. Some kind of fitting problems seem to have also been present in ancient times, because my ancient lead glans has been made in a mold where the halves were shifted by a few milimeters.

So I went the way with the intermediate plaster forms. I made first one form with three alread cast lead glandes pressed halve in, but without the casting channels. After hardening I made some holes at the corners to assist correct fitting of the mold halves. Something like that can be see in Voelling, Abb. 26. After that I did cast the second plaster form fitting to the first halve with inserted glandes. (Some corrections of the forms have been done at various steps to fix errors in making them, e.g. at the edges of the glandes or to flatten the contact point between the molds.) So now I had two halves of a mold without the casting channels. I fitted to both forms with glandes inserted one after the other my cleaned clay. Drying had to be done very slowly and possibly under some pressure, otherwise the mold halves would bend. After carefully drying the halves still fitted good, but there were some gaps between them. Possibly because of the shrinking (6% of the long side) the heights of the uneveness of the connecting sides were reduced resulting in these gaps. The casting channels were cut into the mold when it was halve dry. The first cast was not successful. The main reason is that the pan I use for melting the lead is with the estimated required 200g lead nearly full. Because of that the lead came out of the pan at the beginning not at the desired point but nearly every else. When I succeeded really getting lead into the mold the lead inside the pan had cooled down and also the little lead I got in before. The result was that only the top two glans cavities had been filled with lead. Additionally the gaps had filled. This stuff is about 1mm thick, but because of the softness of the lead it surely can be cut away easyly. Some casts later with better aiming and last humidity evaporated the result was much better and all cavities had been filled. The weight of the whole raw cast is 170g.

[ Two-part clay mold ]

The two-mold glandes have been found much more often than the one-mold glandes. In fact the depiction in Voelling, Abb. 27 is the only one-mold glans I know of. Every other glans, e.g. all these I have seen so far on ebay, must have been made with two molds. The two-halve forms were for me much more complicated to manufacture. So either the ancients valued the form so much that they accepted the additional work or they had simpler production. Mayhap they had additives which reduced the shrinking of the clay to a managable amount. Another possibility is that the produced the two halve like I did with the one-halve mold. This could also explain the non-fitting of the form which was used for casting my ancient glans.

After some thinking I would make the mold for large scale production as following: A mold where the glandes are cast inline is symmetric. So it is possible to make a halve positive form. This made e.g. of wood. It consists of a board where halves of the glandes and casting channels are mounted. Additionally on one side (left or right) of the glandes are two pins and on the other side are matching holes in the board. With meassuring instruments that should be possible. Then clay can be pressed onto that form to get the mold halves. Because of the symmetry of the form any two of the resulting mold halves will fit together.

Reconstruction of Etruscan (possibly Archaean) sling

[ Etruscan and late Roman slings ]

On Etruscan cut stones are two detailed depictions of slings which are known to me. These show braided slings like my braided sling, but with the pouch consisting of three strings. The middle string looks to be thinner than the release/retention cords while the outer strings look to be as thick as them. (See a above.)

My design goals where (in the given order):

The design is based on the braided sling described by Dan Bollinger. It is probably helpful to read that first. If you have some experience with braiding the whole procedure will take about two hours for a 60cm sling, if this is your first sling mayhap three hours.

The following photo shows the result:

[ Etruscan sling ]

The finger loop:

[ Etruscan sling - finger loop ]

This shows the part where additional strings are braided in and the cord gets thicker:

[ Etruscan sling - split ]

And finally a photo of the pouch:

[ Etruscan sling - pouch ]

I think this fulfills all my design goals.

Reconstruction of later Roman (possibly Balearic) sling

Depictions of slings on later roman works show slings to be build of a single strap which is much wider than the braided sling and which has an at most little defined pouch. (See b above.)

Voelling asumes that because Livius writes that Archaean slings are "three-straped" and Balearic slings are "single-straped" the former is the Etrsucan sling and the later this simpler Roman sling. This is a possible interpretation and so I will combine knowledge about Balearic slings with the Roman depictions.

The depictions and Livius's description give us following form of the sling:

Regarding the material Strabo writes in his Geography Book III Chapter 5:

"...of black-tufted rush, I say, or of hair or of sinews..."

So regarding the material we have then following items:

When I combine this information together with available material and practical considerations, I get following properties:

So overall the reconstruction would for me be a tablet woven strap of e.g. 1.2 meters length. This is mostly about 2-3cm wide and increases width by 1-2cm where the pouch is. At one side about 2cm before the end a slit of about 4cm is woven into the sling.

Tablet weaving is not easy. My first test strap looked like a worm on psychedelics while being on a ship in an huricane, well, nearly. For my next test strap I will use alternating S and Z card configuration.

Now I weaved 12 threads in 6 tablets (so 2 per tablet). The center 4 tablets with two halve turns forward/backward (each weaving step one halve turn) and the border tablets with different shearing(?) with two halve turns backward/forward. The result is the most even woven piece I ever got:

[ Weaving test ]

First try

So my plan is:

I started weaving and it's hell. The problem is that turning of the tablets is very difficult. Turning them all together is impossible. I have fixed now the start of the warp with a rope to my belly. ("Waist" would not have been the correct word.) This freed my second hand and turning the tablets four or five at a time is now easier. Waxing the warp has helped making the resulting woven stuff more even. (Thanks for this tip, Willeke!) Reason for both problems is the rough surface of the yarn.

Done! Done! Done! I have it finished! A 155cm long strap. This will probably be also my last try. I can't imagine why anybody would tablet weave a sling instead of braiding it. It neads much more thread and much more time.

First little tests showed this sling to have bad performance as expected. Luckyly this is a dual use item. I can it use as a sling and as a girdle.

Second a bit more extensive tests showed that this sling might be good for larger stones because the release cord is so wide that it is easy to hold. I have also noticed that the release is much later than with my 3-knot sling. The first couple of stones went into the ground a few meters befor me. And finally usage of this sling requires more force than my light slings.

And some photos:

[ Tablet woven sling ]

[ Tablet woven sling - fingerloop ]

[ Tablet woven sling - end ]


I wrote a simple program in Perl implementing the ballistics described in the article from Baatz. This software does do many little steps to calculate the flight path of the projectile. It implements air drag (constant and curves similar to Achenbach), atmospheric density changes according to height (is hopefully good enough for slinging heights) and gravity. I am not an expert in these things, but the results are very close to the data given by Baatz in his article, so I think they are at least good enough for analyzing slinging physics. You can try how light clay bullets fly or heavy stones. You can try how far the things fly when you get in the low area shown by Achenbach. And there are many other things to try. A description of how to use and configure this script is at its beginning, but you should be at least a bit familiar with Perl.

Example plot (made with gnuplot) of the three cases shown in Baatz's article. For comparing the lead values in the article are shown as little crosses.

[ Ballistic curves ]

And the script itself: ballistics.pl

Not implemented and certainly important for e.g. the flight of a golf ball is lift because of rotation. Some can control the rotation with the sling too, but probably most - at least I - can't control it. Sometimes if I sling a flat stone it flys an arc because of the rotation and some make a buzzing sound.

It would be interesting to see cd values of arrows.

Impact tests

Now this is really a long term plan. Testing impact is not really easy. I am not satisfied with the tests I have so far seen with whatever long ranged ancient or medieval weapon. To get a full understanding of the effect of a weapon we have to consider and analyze following:

So after the environment is set all the things happening from the slinger up to the hit person can be tracked. E.g. the slinger manages to get the projectile flying at 80 m/s. This projectile is a 60 g lead glans. This has after the determined distance of 200 m at the calculated air drag a remaining impact speed of 50 m/s and hits at an angle of 30°. When a 60 g lead glans at 50 m/s with an angle of 30° hits a lorica hamata on a person on the middle of the thorax it does do that damage to the armor and this ramaining damage to the body.

All tests I have seen did something different. They use shooting distances of a few meters. They consider only direct frontal hits. Their body model is not good for the armor because the armor behaves differently depending on what's inside. Or it is not asked what damage is done to the body itself.

This is not easy, but nonetheless things should be done properly. To make the problem simpler it should be broken down to following phases:

These three phases can be analyzed in that order. Each has a result which others can base their research on. So first step is to get data on the sling, the starting velocity and rotation of the projectile. The next step would ask how this projectile would behave in flight, i.e. is it in critical speed, what is the air drag, is there a lift because of rotation. The last step would then be the impact on the armor + body simulation. For this last step and also for the second step there is no sling required. A mechanical device which gets the glans up to speed would be enough and I think that would make consistent testing much easier. It would also make the scenario possible that a person has a kind of armor and wants to know on its properties against lead glandes. He would use the data of the analysis of the first two phases from somebody else, sets up his projectile acceleration machine and starts testing his armor.



Copyright 2006 by Matthias Brüstle (web@mbsks.franken.de).
Last changes: 2006-07-10.

[ Valid HTML 4.01! ] [ vi powered ]