Hypodermic

The enclosure is not terribly large, but we would not call it small, either; it is the sort of average size that always grows when one feels small and shrinks when one feels large and restless. Erring on the quaint side, perhaps. You must understand how astounding it was, then, that in such a place, we should find a network of tunnels. They took her five days to make. The dirt is packed tight, but even still, they are so close to the surface it is a wonder they do not cave in. Some of them must go quite deep, and we suspect some empty into small rooms, for we have witnessed the disappearance of a good portion of the enclosure's objects and furnishings and where could they have gone but into the tunnels? We do not consider any of this too peculiar in light of the arid climate here.

But here is the strange thing:

As far as can be told, the tunnels do not seem to extend further than the walls of the enclosure. The walls are constructed above ground, so there is no reason they should hinder the underground network's extension; indeed, it would be difficult to ascertain these boundaries once one began digging, I feel. But we have only ever seen entry points within the enclosure (in the interest of variable control, we have researchers posted quite far out along the landscape), an sometimes, when it has not rained for some time and the soil grows brittle and dry, we hear her scrapping about the perimeter, as she often does, though these times, unseen.

We cannot say whether there is something that inhibits her leaving or if she chooses not to leave. These are perhaps not mutually exclusive.

Further, we cannot yet discern the purpose of the tunnels, unless it is something related to the removal of the sight of her (but we are certain she does not know she is being watched), or the shifting and moving of things in the enclosure.