There was an error published in J. Exp. Biol. 218,[440][441][442][443][444][445][446][447][448][449][450] In Fig. 5 of this paper, there was a mistake in the theoretical calculation of the average slip angle, b s . The corrected calculations show that b s rises more sharply for low mean relative curvatures and approaches 90 deg. This change does not affect the conclusions set forth in the original paper, and shows better agreement with the experimental findings.The corrected figure is reproduced below:The authors apologise for any inconvenience this may have caused. b s ) with changing mean relative curvature (kl s ), average number of undulations along the body ( j), body length to width ratio (L/w) and effective friction. Force relationships were established from empirical drag data in loosely packed (LP) media. The color of the curves and experimental data correspond to different j (where dark blue is j=0, dark red is j=5). Solid curves are predictions for an undulatory swimmer with a L/w=7.1 and body-particle friction, m s , of 0.17. The dotted-dashed curves show the theoretical predictions with L/w=33.5 and m s =0.17. The dashed curves are the predictions for L/w=33.5 and half the tangential force (which occurs when there is a 50% decrease in body-particle friction, i.e. ABSTRACT Squamates classified as 'subarenaceous' possess the ability to move long distances within dry sand; body elongation among sand and soil burrowers has been hypothesized to enhance subsurface performance. Using X-ray imaging, we performed the first kinematic investigation of the subsurface locomotion of the long, slender shovel-nosed snake (Chionactis occipitalis) and compared its biomechanics with those of the shorter, limbed sandfish lizard (Scincus scincus). The sandfish was previously shown to maximize swimming speed and minimize the mechanical cost of transport during burial. Our measurements revealed that the snake also swims through sand by propagating traveling waves down the body, head to tail. Unlike the sandfish, the snake nearly followed its own tracks, thus swimming in an approximate tube of self-fluidized granular media. We measured deviations from tube movement by introducing a parameter, the local slip angle, β s , which measures the angle between the direction of movement of each segment and body orientation. The average β s was smaller for the snake than for the sandfish; granular resistive force theory (RFT) revealed that the curvature utilized by each animal optimized its performance. The snake benefits from its slender body shape (and increased vertebral number), which allows propagation of a higher number of optimal curvature body undulations. The snake's low skin friction also increases performance. The agreement between experiment and RFT combined with the relatively simple properties of the granular 'frictional fluid' make subarenaceous swimming an attractive system to study functional morphology and bauplan evolution.