In a comment on Pharyngula, Jerry Coyne notes:
I suspect that this snake ingested a lizard, and that the lizard's limb simply burst through the side of the snake. I may be wrong, and I hope so, because this is great evidence for evolution.
Some graphic images below the fold illustrate why this is not unreasonable speculation.
Snakes sometimes consider their prey choices poorly. Here's a snake with legs and two tails:
(Hat tip to Febble)
Oops! I sort of skipped the first comment at Pharyngula. This commenter noted first that it was probably something the snake ate. Moreover, they note a fact I forgot to mention in my haste: the limb is quite far from where we'd expect the hindlimb to be, if one were to show up. It would be much closer to the tail, not at mid-length of the body. It should be at approximately the same level as the cloaca. There's the unlikely case that it's an atavistic forelimb however, which would raise the issue of where a snake's neck begins or ends.