Snakes use scales to slither
Mathematical model suggests 'sideways' friction is key.
Snakes rely on the frictional properties of their scales to slither, a new study suggests. The work could explain how snakes move across bare terrain such as sand and roads, where they can't push off rocks and branches.
The researchers, led by mechanical engineer David Hu of the Georgia Institute of Technology in Atlanta, found that the resistance of a snake's belly scales is highest when its body is sliding sideways, rather than forwards or backwards. Snakes also seem to lift the parts of their bodies where friction is slowing movement the most, enabling them to slither faster.
Although scientists knew that snakes slide forwards more easily than backwards, "no one had measured the sideways friction", says Hu. "That's the key to their motion."
Snakes can move by folding themselves into pleats, contracting their bellies, contorting into helices or slithering in an S-shape. Previous experiments showed that snakes perform this last type of motion, called lateral undulation, by pushing their flanks off obstacles in their path1. Scientists suggested that the snakes' belly scales, which can catch on small bumps in the ground, might also aid movement1,2. But exactly how the scales' frictional properties contributed to slithering hadn't been fully explored.
Hu and his colleagues measured friction by sliding 10 juvenile Pueblan milk snakes head-first, backwards and sideways over two surfaces — rough cloth and smooth fibreboard. On the rough surface, friction was highest when the snake was sliding sideways, whereas on the smooth surface, friction was fairly evenly matched in all directions. The increased sideways friction seemed to be necessary for movement, because the snakes could slither successfully on the cloth but not on the fibreboard.
When the team plugged these friction values into a mathematical model, the theoretical snake followed roughly the same path as the real snakes. However, the speeds predicted by the model were lower than those the researchers observed in the snake experiments. So they factored in the snakes' tendency to concentrate weight on certain parts of their bodies — behaviour that seemed to reduce contact in areas where friction impeded movement the most. This addition made the theoretical snake move 35% faster, a rate that agreed more closely with real snake speeds, the team reports in the Proceedings of the National Academy of Sciences3.
The study illustrates how varying frictional resistance along the snake's body sets up the forces for forward motion, says Bruce Jayne, a snake locomotion expert at the University of Cincinnati in Ohio who was not involved in the work. "They're accounting for everything that happens from the head to the tail of the snake," he says.
The model also suggests that slithering depends much more on frictional than inertial forces, with frictional forces estimated to be an order of magnitude greater. This is intriguing, because for many other forms of land-based movement such as running and hopping, "body inertia is everything", says Bruce Young, a vertebrate anatomist at the University of Massachusetts Lowell. However, he notes that other forms of snake locomotion may not follow the same model.
Hu says the study might help engineers to design better snake robots, which can be used to manoeuvre into tight spaces. Some snake robots have wheels that resist sideways motion, but roboticists might be able to replicate slithering without using wheels if they can find a material with frictional properties similar to scales, he says.
Howie Choset, a roboticist at Carnegie Mellon University in Pittsburgh, Pennsylvania, says that most snake robots rely on movements other than lateral undulation, such as rolling. Still, Hu's work might inspire a closer look at the frictional properties of snake skins on future robots, he says.
- Gray, J. & Lissmann, H. W. J. Exp. Biol. 26, 354?367 (1950).
- Hazel, J. et al. J. Biomechanics 32, 477?484 (1999).
- Hu, D., Nirody, J., Scott, T. & Shelley, M. J. Proc. Natl Acad. Sci. USA doi:10.1073/pnas.0812533106 (2009).