See algebra/factor_1 theory for more details of this.
Approach here to dealing with 0 in lemmas and definitions has been a bit ad-hoc. Is there something more principled that could be done?
IF YOU CAN SEE THIS go to /sfa/Nuprl/Shared/Xindentation_hack_doc.html