That's what I meant, they all have variations of this, but that account doesn't suggest how it ignited, whether by contact, time, pressure, catalysis, or what.

There's a contemporary picture that shows fire spurting out of the barrel, i.e. it was on fire as soon as it entered the air. I don't suppose we need rely on that picture, but it doesn't look like twin tubes. A binary would be easier in grenades, but it doesn't look like it was fired as a binary.

Could a mixture be flammable in the open air and non-flammable under confinement, so that they could light it at the nozzle as it sprayed out and not have it blow up the barrel?