"We could not have scheduled this if we had tried," said Chris Done, Durham University, UK, the lead researcher on the study. "The system went from about half its maximum radiation output to something much more intense, creating a wind that was thicker than we'd ever seen before."

But mysteriously, the wind was not travelling at the speed that the XRISM scientists were expecting. It remained around 1 million km/h. While fast by any terrestrial standard, this is decidedly sluggish when compared to the cosmic winds produced near the Eddington limit around a supermassive black hole. In that situation, the winds can reach 20 to 30 percent the speed of light, more than 200 million km/h.

“It is still a surprise to me how ‘slow’ this wind is,” says Chris, “as well as how thick it is. It’s like looking at the Sun through a bank of fog rolling towards us. Everything goes dimmer when the fog is thick.”

It was not the only difference the team observed. XRISM had earlier revealed a wind from a supermassive black hole at the Eddington limit. There the wind was ultrafast and clumpy, whereas the wind in GX13+1 is slow and smooth flowing.

“The winds were utterly different but they're from systems which are about the same in terms of the Eddington limit. So if these winds really are just powered by radiation pressure, why are they different?” asks Chris.

The team has proposed that it comes down to the temperature of the accretion disc that forms around the central object. Counterintuitively, supermassive black holes tend to have accretion discs that are lower in temperature than those around stellar mass binary systems with black holes or neutron stars.