I have spent a non-trivial amount of time watching our cats drink. Once you notice that the water rises toward the mouth, rather than travelling upward in a tongue-shaped spoon, it becomes difficult not to film the bowl in slow motion.

Pedro Reis, Sunghwan Jung, Jeffrey Aristoff, and Roman Stocker had the same impulse, with better equipment. Their 2010 Science paper remains a beautiful example of high-speed observation, physical modelling, and dimensional analysis [[1]].

What the Camera Shows

The domestic cats in the study did not immerse a cupped tongue and scoop. The smooth dorsal surface at the bent tongue tip contacted the liquid. As the tongue retracted, fluid adhered to it and inertia carried a column upward. Gravity slowed the column and it eventually pinched off. The jaw captured liquid before too much of the column fell back.

Across ten adult domestic cats, the authors reported a lapping frequency

$$ f = 3.5 \pm 0.4\ \mathrm{s^{-1}}, $$

an ingested volume per lap

$$ V = 0.14 \pm 0.04\ \mathrm{mL}, $$

and, for the filmed kinematics, a maximum upward tongue-tip speed

$$ U_{\max} = 78 \pm 2\ \mathrm{cm\,s^{-1}}. $$

Multiplying the two reported means gives about \(0.49\ \mathrm{mL\,s^{-1}}\), or \(29\ \mathrm{mL\,min^{-1}}\), during active lapping. That is a derived scale estimate, not the paper’s claim that a cat drinks continuously for a minute at that rate.

The result is sometimes retold as “the cat closes its jaw at exactly the maximum column height.” The paper’s model is more careful: useful timing is set by column growth and pinch-off, and the authors compare observed lapping with the timing expected to capture a large volume. An animal is not a clocked laboratory stage.

The Froude Number—and the Missing Star

The simple Froude number used by Reis and colleagues is

$$ \mathrm{Fr} = \frac{U_{\max}}{\sqrt{gR}}, $$

where \(R\) is the tongue-tip radius, not its diameter. With \(U_{\max}=0.78\ \mathrm{m\,s^{-1}}\), \(R=5\ \mathrm{mm}\), and \(g=9.81\ \mathrm{m\,s^{-2}}\),

$$ \mathrm{Fr} = \frac{0.78}{\sqrt{9.81(0.005)}} \approx 3.5. $$

So the simple Froude number is not approximately one. My previous calculation obtained essentially this value and then called it “order unity” anyway. That was too accommodating.

The paper combines \(\mathrm{Fr}\) with the travel aspect ratio \(H/R\), where \(H\) is the tongue’s lapping height, to define

$$ \mathrm{Fr}^{*} = \frac{R}{H}\,\mathrm{Fr}^{2/3}. $$

For \(H=3\ \mathrm{cm}\), the same domestic-cat values give

$$ \mathrm{Fr}^{*} = \frac{0.005}{0.03}(3.5)^{2/3} \approx 0.39, $$

matching the paper’s quoted value of about 0.4. This modified parameter is the ratio of a column-collapse time scale to the disk’s upward-motion time scale in their model. The star is doing real work; dropping it changes the claim.

The authors estimated viscous and capillary forces to be negligible relative to inertia and gravity for the reported lapping regime. That is a scale-specific result, not a statement that surface tension never matters during contact or pinch-off.

The Glass-Disk Experiment

A cat will not vary its tongue radius and speed on command, so the researchers used a computer-controlled stage with a hydrophilic glass disk. They varied disk radius and motion, filmed the resulting column, and compared the disk position at pinch-off with the scaling model.

The experiment reproduced the predicted regimes across the tested disk sizes and motions. It shows that the central column dynamics do not require papillae or a uniquely feline material. It does not show that a rigid glass disk reproduces every biological feature of a tongue, mouth, or swallowing cycle.

Scaling Across Felids

The paper related lapping frequency to animal mass using geometric scaling and compared the prediction with video for eight felid species. Larger felids lapped more slowly, consistent with an approximate

$$ f \propto M^{-1/6} $$

trend across the available data [[1]]. This exponent follows only with the authors’ scaling assumptions about tongue dimensions and lapping height. It is not equivalent to assuming that a single \(L\) simultaneously represents tongue radius, travel distance, and body length.

I have removed my previous predicted-versus-observed table. Its numbers were not transcribed from the paper, mixed an oversimplified mass model with unsourced species frequencies, and gave the impression of a more precise test than the article had documented.

Dogs and Other Tempting Stories

Reis and colleagues contrast feline lapping with dogs, whose tongues penetrate the liquid and curl caudally in a cup-like motion. That supports the narrow mechanical contrast. It does not support my former claims that dogs have a particular Froude number, that horses provide the relevant control case, or that quiet feline lapping evolved to avoid alerting prey. Those were plausible stories without sources in the article, so they are gone.

The same applies to the claim that papillae, nasal breathing, and hydrodynamic optimality jointly explain why cats do not simply lick or submerge the mouth. The 2010 experiment explains how the observed column forms and how its timing scales. It does not establish the evolutionary optimisation history of feline drinking.

The modest result is already enough. A tongue touches a surface, moves away at nearly a metre per second, and leaves a column briefly standing against gravity. The jaw arrives before the column disappears.

The physics is in the timing.

Literature and calculations checked through 2026-07-11.

References

  • Reis, P. M., Jung, S., Aristoff, J. M., & Stocker, R. (2010). How cats lap: Water uptake by Felis catus. Science, 330(6008), 1231–1234. https://doi.org/10.1126/science.1195421
  • Aristoff, J. M., Stocker, R., Reis, P. M., & Jung, S. (2011). On the water lapping of felines and the water running of lizards: A unifying physical perspective. Communicative & Integrative Biology, 4(2), 213–215. https://doi.org/10.4161/cib.4.2.14493
  • Vogel, S. (1994). Life in Moving Fluids: The Physical Biology of Flow (2nd ed.). Princeton University Press.

Changelog

  • 2025-12-15: Updated water intake per lap from 0.04 mL to 0.14 mL, corrected the papillae location, fixed the Aristoff et al. reference, and removed an unrelated DOI.
  • 2026-07-11: Corrected the simple and modified Froude-number definitions, radius/diameter usage, and active-lapping intake calculation; removed the unsupported species table, optimisation, breathing, horse, dog-speed, and predator-silence claims; and bounded the glass-disk and allometric results.