The number I want is not the 900. It is how those embryos divide across developmental stages, and none of the published accounts give it. More than 900 embryos from 39 species works out to roughly 23 per species on average [1], and the images were harvested from published papers and museum collection databases [2], which makes this a found sample rather than a designed one. If the earliest stages are well covered in three species and thin across the other 36, the claim about early handedness is resting on a much smaller denominator than 900.
The pattern is stage-specific, and that is the strongest thing about it. Hold the two ends of development together, exclusively right-handed in the first weeks [3] and about even near hatching [8], and it follows that roughly half of all embryos reverse their coil somewhere in between [2]. The team's account of the switch is unforced: the yolk is consumed, the embryo has more room, muscle matures, and some animals recoil to the left [8].
Miyashita's mechanism is a strap. "It's like when you adjust the length of a strap and the longer, buckling side of the loop twists," he says [9]. The body accelerates its elongation, the gut does not keep pace, the slow gut detaches and tethers the lengthening body, and the body buckles [6]. Buckling accounts for the coil but not for its direction. The direction comes from the yolk, which lies to the left, so the coiling force drives the growing body to the opposite side [7]. That is worth stating plainly: the explanation moves the handedness question one step earlier in development, from the coil to the yolk, and the yolk's position is not what this paper explains.
What these accounts do not contain is a manipulation. No gut is released or shortened to see whether the coil follows, and no side-by-side measurement of body and gut elongation rates appears in the summaries, so the growth mismatch is inferred from the anatomy and from the handedness pattern. The anatomy is genuinely new: a pillar of intestine detached from the body and sheathed in blood vessels from the yolk [5], described in two of the three accounts as the scan of a snake embryo, singular [5]. Discover's write-up keeps the functional claim at the right temperature, reporting that the researchers think the interaction of the lengthening body, the slower gut and the yolk may have helped make the elongated snake body possible [15].
My reading is that the early dextral bias is real and the tether is the most economical explanation anyone has offered for it, which is more than a chirality usually gets. The wider claim, that this is how snakes accommodate the longest bodies of any vertebrate [10], is a plausible story about a physical constraint and not yet a demonstration. It is also worth noticing where the result came from. Work on the snake body form has generally gone through Hox genes and enhancers [12]; this came from a lockdown project in which undergraduates scrolled an album of embryo photographs and wrote down which way each one curled [11][12].
