Many references state that pumice is formed by gases in lava, but I have yet to find a reference of which gases they are. If a piece of pumice floats, then conceivably it still contains the original gases from the magma, right?

Breaking the pumice would release the gas trapped in the vesicles.

According to https://www.tulane.edu/~sanelson/Natural_Disasters/volcan&magma.htm, the gases in magma are

  • Mostly H2O (water vapor) & some CO2 (carbon dioxide)
  • Minor amounts of Sulfur, Chlorine, and Fluorine gases

But is this true for pumice? If it's mostly water vapor in the gas, then when it cools, the pumice holes would contain liquid water, albeit small amounts compared to the size of the vesicles.

Otherwise, does pumice typically have other gases?

Edit: Here's some background for this question. My 7yo son asked me where the bubbles in pumice came from, when I showed him that this "rock" can float. Google didn't give a good answer, so after he went to bed I started searching... Kids ask the best questions!


3 Answers 3


Pumice does not contain any gas, the vesicles are empty1. The magma which generated the pumice used to contain gas, mostly water and CO2 like you said. But this gas was lost during the eruption. Most of the vesicles are actually connected, allowing the gas to escape. The connectivity $C$ is defined as:

$C = \frac{\phi'}{\phi}$

where $\phi'$ is the connected porosity, and $\phi$ is the bulk porosity. So $C$ varies from 0 (no connectivity, all vesicles are isolated) to 1 (all vesicles are connected).

Colombier et al. (2017) compiled a database of connectivity measurements of volcanic rocks, the results are summarized by this figure:

enter image description here

The covariance of $C$ with $ϕ$ for all natural rocks classified by bulk rock composition

You can see that connectivity in pumice (panels a, c, d, and e) is very high, usually between 0.8 and 1.0, meaning that most of the porosity is connected, allowing gas to escape through the pore network. As for the causes of such high connectivity, the study states:

Pore connectivity develops by vesiculation and bubble growth-driven coalescence or by cracking or combinations of both processes.

1Well, as pointed out in a comment, connected vesicles are actually filled with atmospheric gases.

  • 7
    $\begingroup$ Nitpick - that sounds like they contain N2, O2 a little Ar, H2O, CO2, boring old air $\endgroup$
    – Chris H
    Commented Nov 17, 2021 at 11:00
  • 1
    $\begingroup$ @ChrisH Good point, I added a sentence to address your comment, thanks! $\endgroup$ Commented Nov 17, 2021 at 14:31
  • $\begingroup$ It's probably overly pedantic of me, but I'd just been working on a vacuum system so has an instant of confusion. Thanks $\endgroup$
    – Chris H
    Commented Nov 17, 2021 at 14:43
  • 4
    $\begingroup$ If the vesicles are connected, how does it float? Wouldn't water fill the vesicles? $\endgroup$
    – Austin
    Commented Nov 17, 2021 at 20:58
  • 7
    $\begingroup$ @Austin - Capillary action won't work until it's saturated due to air lock. And the surface tension of water doesn't easily lend itself to the wetting of passages that small. $\endgroup$
    – Mazura
    Commented Nov 18, 2021 at 0:37

The work of Formenti et al. in Earth and Planetary Science Letters 214 (2003) 561 disagrees with the idea that pumice can never retain original gases (emphasis added):

Our observations show that dense blocks derived from lava domes are commonly not able to retain gas because essentially all the vesicles are interconnected, as also found by Le Pennec et al. [18] in an electrical conductivity study of samples of Merapi lava dome. On the other hand, typically a few volume percent of vesicles in pumices from fountain-collapse pyroclastic flows at Soufrière Hills and Lascar are isolated, showing that gas can be retained in isolated vesicles of pumices even after cooling and deposition.

Unfortunately, they don't have any information on what the contents of the vesicles would be. It was discovered in the 1920s that pumice did not retain sufficient original gases to be useful as a sample of volcanic gas. Scientists decided it was easier to catch the gases emerging from the volcano than to try to sample the highly contaminated contents of crushed pumice. Even with modern techniques, the difficulty of determining the contents of isolated vesicles is simply too high, since they presumably contain the same gases as emitted from volcanic vents.

  • 1
    $\begingroup$ Nice addition! Also, we are now able to measure pre-eruptive, magmatic gas contents from bubbles trapped in melt inclusions, so we don't really need pumice to retain some gas to access this information. $\endgroup$ Commented Nov 17, 2021 at 14:58

Just to add to the other answers,

If it's mostly water vapor in the gas, then when it cools, the pumice holes would contain liquid water, albeit small amounts compared to the size of the vesicles.

This is actually a very important point.

When the magma erupted, the vesicle contained very low density H2O gas.

Assuming the vesicle remains closed somehow and isolated from the outside, cooling down of the vesicle would not necessarily lead to the formation of liquid water. The total amount of molecules inside the vesicle if fixed, and the drop in temperature leads to a very low inside pressure. The total density remains the same, which is much lower than the density of liquid water. Therefore, pumice floats.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.