I am trying to find a gas that would be used as a weapon, which could also be made by someone in a medieval setting. The gas doesn’t have to be explicitly deadly, but anything that would incapacitate the enemy (knocking out, blinding, causing extreme sickness, etc) while being mass produced at a military scale would work. I have access to very skilled craftsmen who could make almost any tools needed for chemistry. The gas needs to linger long enough to take effect, and needs to take effect before the enemy is able to realize what is going on and leaves the area. I am not sure if there are even gasses like this that could be made without modern techniques, but if there are I would appreciate the info.

Edit: These would need to be usable on the battlefield, not just in enclosed spaces or pre prepared ambushes/sieges. They could be used be stored in barrels which are opened (such as how it was used early during WW1) or launched (similar to artillery)


5 Answers 5


Pre-medieval chemical warfare has an extensive and ignoble history. Since this question is less worldbuilding, and in fact, answerable through history, I recommend reading about the history of chemical warfare.

Your main conduit for creating heavier-than-air toxins is fire, which makes smoke. Despite being heavier than air, smoke rises because of the heat, but it has still been used in warfare, especially in Ancient China.

Regular wood is a good starting point: poison sumac, castor oil, and fig trees all produce very nasty smoke I don't recommend inhaling.

Moving up from there, burning arsenic produces toxic arsenic fumes. Sulfur, bitumen, and other naturally occurring petrochemicals all produce pretty foul smoke.

  • 11
    $\begingroup$ When I think of all the warnings I have had over the years to never burn poison ivy, I just have to say what a great answer this is. $\endgroup$
    – Dan
    Commented Feb 1, 2022 at 14:02
  • 3
    $\begingroup$ @Dan My brother once accidentally burned some poison oak with some other wood. It was probably even worse than you are imagining. $\endgroup$
    – Michael W.
    Commented Feb 1, 2022 at 17:05
  • 3
    $\begingroup$ Then there are the things that aren't technically gasses, but are close enough for warfare purposes. Burning shellfish shells makes quicklime, which easily crushes into a fine powder that will float on the breeze for quite a distance before settling. And if there's a nearby deposit of super-fine sand that you can head up red-hot before throwing it that will work pretty well too. $\endgroup$
    – Perkins
    Commented Feb 1, 2022 at 20:22
  • $\begingroup$ What would castor oil actually do if it was burned? I can not find anything specific about the effects of castor oil smoke inhalation. Same with the fig trees. I also don't think I have access to poison oak since that is a new world plant. $\endgroup$ Commented Feb 2, 2022 at 1:34
  • $\begingroup$ Not castor oil itself, just the wood of the castor oil tree. Likewise for fig trees. As for the effects, I don't know of any specific studies, but afaik the aerosolized sap of both causes eye, lung, and skin irritation. For obvious reasons, I've never tried it myself. $\endgroup$
    – Daniel B
    Commented Feb 2, 2022 at 3:07


In laboratories and high-school classrooms animals occasionally need to be put to sleep for a short period for examination and other purposes. One of the things used is the gas CO2 or carbon dioxide.

In the right circumstances this can be used on people. If your enemy enters a natural dip or something like a moat, cellar or dungeon, then the deployment of the gas can be effective as it's heavier than air.

The gas itself can be produced by mixing quantities of easily available materials - chalk or lime with vinegar. These would need to be made available in fairly large quantities, the liquid being tipped onto the solid mass within a great pit, the maw of which spills onto the area that you'd want affected. Wind permitting, this should provide anaesthesia for a few minutes, death if longer.

A cellar, or dungeon can have a casement opened by pulley to flush the asphyxiant out. It would then be safe to enter.

  • 2
    $\begingroup$ Hmm, a gas moat. Pretty sneaky... $\endgroup$
    – DrMcCleod
    Commented Feb 1, 2022 at 9:25
  • 4
    $\begingroup$ Are you sure it's not carbon monoxide? Humans get panicky when we have too much CO2 in our bloodstream, but we don't notice a buildup of carbon monoxide at all. $\endgroup$
    – alexgbelov
    Commented Feb 1, 2022 at 19:20
  • 2
    $\begingroup$ Good point, I've never tried on humans. CO2 works just fine for fruit flies, they never complained of panic. (Maybe if I'd listened closer, their tiny voices don't carry far). The Wiki states loss of consciousness at about 12% preceded by irrational behaviour and confusion - it could work just fine, but not in a "tidy" way, there'd be panic yes. @alexgbelov $\endgroup$ Commented Feb 1, 2022 at 19:27
  • 3
    $\begingroup$ Related CO2 disaster. $\endgroup$
    – user
    Commented Feb 1, 2022 at 20:37
  • 2
    $\begingroup$ @EveninginGethsemane: Unfortunately (for our purposes), CO is lighter than air. $\endgroup$
    – Vikki
    Commented Feb 2, 2022 at 0:59

Chlorine Gas

While this was not widely weaponized until WWI, it can be made using medieval technology.

There are many ways to make Chlorine Gas, but the earliest recorded method was probably Aqua regia: a mixture of nitric acid and hydrochloric acid invented in the 13th century that alchemists used to dissolve noble metals like Gold and Platinum. However, Aqua regia is very unstable and releases a constant emission of Chlorine Gas. Although the full dangers of Chlorine Gas were not documented until the 1600s, Chlorine Gas is very noxious and its effects would have certainly been obvious to the alchemists who worked with it in the 1200s.

So, while not historically weaponized in the medieval period, if a king ever bothered to ask an alchemist about weaponizing gases, this would have been an obvious choice.

  • 1
    $\begingroup$ The problem with chlorine is that, although an alchemist can easily produce it, producing enough of it to be useful would strain the capabilities of medieval industry. Dissolving sal ammoniac in aqua fortis may produce chlorine, but it doesn't produce very much, and both sal ammoniac and aqua fortis are relatively expensive ingredients. $\endgroup$
    – Mark
    Commented Feb 1, 2022 at 23:22
  • $\begingroup$ What would you need to make nitric acid and hydrochloric acid? Are they easily obtained ingredients for someone in the medieval times or would it take time to procure enough of it to be actually useful? $\endgroup$ Commented Feb 2, 2022 at 1:14
  • $\begingroup$ to make nitric acid add saltpetre to concentrated sulphuric acid and disiolve the fumes in water, to make hydrochloric use salt instead of saltpetre. $\endgroup$
    – Jasen
    Commented Feb 2, 2022 at 10:43
  • 2
    $\begingroup$ @TheEmperorProtects To add to what Jasen and Mark said, sulfuric acid was made by heating a mineral called green vitriol, a common type of sulfate crystal. These chemicals were expensive in the Medieval period because there was not much demand for them. An alchemist might spend days making a batch of hydrochloric acid just to get the few ounces he needs to meet demand, but producing it by the barrel full would not be hard if the demand was there since the material cost is low. $\endgroup$
    – Nosajimiki
    Commented Feb 2, 2022 at 16:45
  • 1
    $\begingroup$ @Nosajimiki lye and hydrochloric acid makes salt and water, not chlorine. $\endgroup$ Commented Feb 2, 2022 at 16:49

Rather than CO2 a gas moat or asphyxiating pit trap would be better realised by SF6. Sulphur hexafluoride is far denser, meaning it hangs around far longer, and, like CO2 is odourless and colourless. I suggest you could stretch late-medieval alchemy to get to it. You could possibly even pour it down a hillside into an enemy encampment in a valley.

Unlike CO2, an increase in SF6 in the body isn't physiologically detectable. The urge to breathe when you hold your breath isn't due to lack of oxygen, but excess of CO2. This means that when it's used industrially e.g. (as an electrically insulating gas), real precautions have to be taken to vent, or perhaps I should say drain any leakage. The victim would pass out before realising anything was wrong, akin to shallow water blackout. Once they'd collapsed, lying on the ground the concentration would be still higher than standing. Anyone going to rescue the victim would exert themselves and breath deeply. The mild anaesthetic effect would only add to the hazards.

Potential contaminants in the manufacturing process and form environmental degradation are useful too:

  • S2F10, which is toxic to the point that it's been considered for chemical warfare. This is liquid at room temperature but will have a significant vapour pressure; its odour is like burnt matches, so would probably go unnoticed on a battlefield.
  • SF4 is a toxic and corrosive gas with a similar density to CO2

Sulphur has been known since antiquity, so the limit to production is fluorine. The earliest steps in fluorine chemistry go back to 1529, the very end of the middle ages, with possible 15th century alchemical mentions. Isolation took longer, partly because fluorine and in particular HF (hydrofluoric acid) killed people researching it.

Compressing it to store it would be tricky but not impossible (old answer of mine)

  • 1
    $\begingroup$ The problem with sulfur hexafluoride is that while it might not be physiologically detectable, it will certainly be acoustically detectable, what with everyone's voice dropping a few octaves. $\endgroup$ Commented Feb 1, 2022 at 19:34
  • 1
    $\begingroup$ @KeithMorrison So they'll all think they've been possessed by evil spirits for twenty seconds or so before they keel over... What's the down side? $\endgroup$
    – Perkins
    Commented Feb 1, 2022 at 20:25
  • $\begingroup$ "stress medieval alchemy to get it" lol guess if you gonna stress medieval alchemy, you can produce mercury vapors or mustard gas as well.. but nice thing about alchemy is producing gold and you won't need chemical weapons of any kind.. pay the cavalry, they'll conquer the land for you.. $\endgroup$
    – Goodies
    Commented Feb 1, 2022 at 20:44
  • 7
    $\begingroup$ No way. SF6 needs fluorine gas to be isolated first and that is both very difficult and extremely dangerous. It didn't happen in our society until well into the industrial revolution (1886), and at that after almost a century of attempts (and deaths) trying and failing. A medieval alchemist would have no hope of success here. The required technology simply does not exist. $\endgroup$
    – J...
    Commented Feb 1, 2022 at 21:03

Iron carbonyl? You get it by passing carbon monoxide over sufficiently finely divided iron metal. The carbon monoxide is only weakly attached, so inhaling it gives you carbon monoxide and free iron atoms in your lungs. It's also explosive in the right concentrations. It's (obviously) much heavier than air.

If you can plausibly get nickel (not available pure until the 18th century on Earth, but there are ancient coins with 20% nickel in copper alloy) then nickel carbonyl is substantially nastier along the same lines. Its LC50 for a 30 minute exposure is about 3 parts per million, according to Wikipedia

  • $\begingroup$ The synthesis isn't just "pass carbon monoxide over iron filings". If you read the paper describing the original synthesis, you'll see you need, among other things, high-purity carbon monoxide -- if any oxygen is present, you get rust instead. You also need your iron to be in nearly-monatomic form: the original synthesis reacted ferrous oxalate with high-purity hydrogen to get what they needed. Not something a medieval alchemist could do. $\endgroup$
    – Mark
    Commented Feb 2, 2022 at 21:01
  • $\begingroup$ Nickel carbonyl is nastier, true, but the same reactivity that makes it so highly poisonous also makes it very short-lived: ten minutes after opening a bottle of the stuff, you won't have any left. $\endgroup$
    – Mark
    Commented Feb 2, 2022 at 21:05

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .