Utah keeps doing this: you turn your back on a stretch of sandstone for five minutes and it hands over a new species. The segment is short on anatomy, so treat it as the headline and watch for the paper with the proper name and the cladogram.

No dinosaurs this time, but the next best thing: a Cretaceous lizard frozen mid-escape in Burmese amber. Behaviour almost never fossilises, which is what makes this tiny tail-drop so satisfying; the trick it shows is one modern geckos still pull on cats.

Herrerasaurids are about as close to the root of the dinosaur family tree as anyone gets, so even a single bone carries a lot of weight here. The interesting bit is not the predator itself but the ecosystem around it: the plant-eaters and cynodonts shifted, and the early meat-eaters apparently just carried on.

No dinosaurs tonight, but this goes much further down the family tree. The textbook picture of a worm-like burrower at the root of the starfish and acorn-worm lineage takes a knock from a Chinese Cambrian fossil. Worth noticing how much a single well-preserved soft body can rearrange a branch of the tree.

Most of palaeontology happens long after the dig, with dental picks and air scribes and a great deal of patience, and usually behind a closed door. Raleigh is putting that slow work in front of the public. Late Cretaceous, so a few million years older than the museum's famous T. rex.

Every few years a new tyrannosaur turns up and the family tree has to be redrawn in pencil again. This one is billed as a relative of T. rex that doesn't sit where the textbook says it should. The pool only carried the headline, so the details of which bones and which branch are behind the link.

Not a dinosaur, but it swam above their heads, and the nickname alone earns it a slot. Belemnites are the bullet-shaped squid cousins every fossil shop sells by the bucket, which makes a barrel-shaped one from a family thought long gone all the more fun. Worth noticing: the shape hints at a different lifestyle, perhaps hunting smaller cephalopods.

Fresh dinosaur news from somewhere that rarely makes the headlines: Zimbabwe, 210 million years ago. The interesting bit is not the skeleton itself but the argument that Zimbabwe's early dinosaurs were a different crowd from the ones down south, possibly because the climate differed. The story is written by the lead palaeontologist himself.

The other branch of the archosaur family tree gets its turn. A UCL team argues that crocodile-line reptiles never really adapted to new climates; they just moved whenever the world warmed or cooled. That same habit may be why they made it through the asteroid when their dinosaur cousins did not.

Hell Creek is T. rex country, so it is a nice change to meet a 12 kg neighbour with a weak bite and a sliding jaw that probably preferred fruit. The comparison to a living parrot is the bit that makes it click. Also a reminder that a nearly complete skull can sit in a collection for over a decade before its story gets told.

Forensic palaeontology at its finest: researchers measured serration spacing in bite marks, corrected for the angle of the bite, and named a likely suspect. The size gap makes the case, since a turkey-sized raptor was not taking down a half-tonne tyrannosaur, but it would happily clean up after one.

Not every dinosaur find is a skeleton; sometimes it is one 37 mm bone that moves a timeline by tens of millions of years. Worth noticing how much careful phylogenetic work can hang on a single centrum, and how honest the authors are about not naming it.

Proof that the best fieldwork sometimes happens in a storeroom. A scrap of skull the museum bought in 1924 turns out to be the first megaraptoran ever found in Europe, which redraws the map for a whole family of big-clawed hunters. Bonus points for a name that translates as unconquered female warrior.