Well, that didn’t take long! Jaap Vogel sent his photos to Dr. David Britton of the Australian Museum, who recognized the cocoons as belonging to Cyana meyricki, or at least something very closely related*. This species is one of the lichen moths, which under the latest classification are tribe Lithosiini within subfamily Arctiinae (tiger moths), which in turn is in the family Erebidae (no common name that I know of). The adult is thought to be a mimic of net-winged beetles (Lycidae), which are distasteful to predators.
When I contacted Dr. Britton for permission to use the above photo, he passed along an article devoted to these cocoons**. It is well illustrated with photos of the cocoon, pupa, adult, and the presumed caterpillar. In it, Geoff Monteith describes some interesting details of the cocoon, which measures about 3 cm by 2 cm:
High power magnification reveals that almost every cross-intersection of setae [hairs] is bound together by a thread of silk. It also shows that individual setae are too short to reach from the substrate to the upper midline of the cage, so the vertical side bars of the cage actually consist of two more setae laid precisely end to end to reach the full length of the cage. The completed cage is remarkably strong and rigid, and is under some tension due to the setae being curved and held in place by the silken hitches. If partly depressed by a finger tip, the cage will spring back into shape when released.
Inside the cage, the pupa hangs in space, perfectly in the centre and thus equidistant on all sides from the presumed probings of predators and parasites. The pupa is enclosed in a loose harness of several silken loops, like a hammock, and from this several suspension lines lead to the sides of the cage. . . Mysteriously, the shed skin of the last larval instar is always outside the cage.
We have not seen the adult moth actually emerge, but in those that have hatched for us in the lab the moth magically appears outside the apparently intact cage leaving the empty pupal shell still swinging in its hammock inside. But close examination shows that the overlapping setae at the two pointed ends of the cage are not silked together like the rest of the cage. This means that a creature can push them apart from inside, and they close behind it as it emerges.
He goes on to suggest that the pupa pushes out the skin of the caterpillar through the rear one-way exit as it molts. A live caterpillar has never been found and raised, so it’s still a mystery how exactly it puts this structure together. Being a lichen moth, the caterpillar is presumed to feed on lichens and algae on dead wood as its relatives do.
If you’re wondering how that tachinid fly got inside the Singapore cocoon, many tachinids lay their eggs on caterpillars rather than on pupae, and the larvae wait until their host has completed its development before completing their own. Other tachinids lay tiny eggs on foliage and hope that a foraging caterpillar will swallow them. So this fancy cocoon is of little use in preventing attacks from tachinids, and it ends up instead protecting the parasitoid–which has cleverly waited to kill the caterpillar until the cocoon has been completed–from its own would-be parasitoids.
* Added 4/30/2012: I should point out that there are many species of Cyana, all of which may make similar cocoons. At least five occur in Singapore, according to this paper, so the parasitized cocoon in my first post could have belonged to any of those.
** Monteith, Geoff. 2008. The mystery of the arctiid moth, Cyana meyricki Rothschild – An Insect Houdini. Entomological Society of Queensland 36(9):237-243.