Thanks for the reply.
Your example works, and the discid is there in the XML as you say. However, as I'm writing in Python, the obvious place to start seemed to be the Python module "musicbrainzngs" which claims to provide bindings for the Musicbrainz NGS webservice. The particular line of Python code I'm using to fetch the data is:
Result = musicbrainzngs.get_releases_by_discid(DiscId, includes=["labels",'artists','recordings'])
and that works, in the sense that I get back a blast of data. However, that data does NOT include the discid inside medium > disc-list. I thought that perhaps I had to actually specify in the 'includes' argument that I wanted the discid (the module documentation mentions "discids" as one of the possible include values) but when I do that I get an error 400 for reasons that are not clear to me.
I can, of course, just do the Web Service call and then parse the XML myself but that is an extra level of complication that I would have avoided if the Python module had done what I wanted.