Doug Miller writes:

Actually, it's doing *both*, isn't it? I think Nick's point was that
window glass isn't nearly as transparent to IR as it is to visible
light. Emissions from an IR source inside the building will be mostly
absorbed by the skylight, instead of passing through, no?

Yes and no. Yes, it absorbs some incident radiation instead of directly
transmitting it (depending on wavelength). But it re-radiates much of what
it absorbs, which is an indirect transmission.

There is a mythical version of "glass blocks IR" that some people think
makes cars hot on a sunny day. In fact it is the *transmission* of shorter
IR sunlight (many 1000s deg K blackbody) via the glass into the car
interior, where it is absorbed on the interior surfaces, and re-radiated at
longer IR spectra (several 100s deg K blackbody) that is blocked by the
glass.