The H II region created by the progenitor of SN 1987A was further heated and ionized by the supernova flash .
Prior to the flash , the temperature of the gas was \sim 4000 - 5000 K , and helium was neutral , while the post-flash temperature was only slightly less than \sim 10 ^ { 5 } K , with the gas being ionized to helium-like ionization stages of C , N and O .
We have followed the slow post-flash cooling and recombination of the gas , as well as its line emission , and find that the strongest lines should be N V \lambda 1240 and O VI \lambda 1034 .
Both these lines are good probes for the density of the gas , and suitable instruments to detect the lines are STIS on HST and FUSE , respectively .
Other lines which may be detectable are N IV ] \lambda 1486 and [ O III ] \lambda 5007 , though they are expected to be substantially weaker .
The relative strength of the oxygen lines is found to be a good tracer of the color temperature of the supernova flash .
From previous observations , we put limits on the hydrogen density , n _ { H } , of the H II region .
The early N V \lambda 1240 flux measured by IUE gives an upper limit which is n _ { H } \sim 180 ~ { } \eta ^ { -0.40 } ~ { } cm ^ { -3 } , where \eta is the filling factor of the gas .
The recently reported emission in [ O III ] \lambda 5007 at 2500 days requires n _ { H } = ( 160 \pm 12 ) ~ { } \eta ^ { -0.19 } ~ { } cm ^ { -3 } , for a supernova burst similar to that in the 500full1 model of Ensman & Burrows ( 1992 ) .
For the more energetic 500full2 burst the density is n _ { H } = ( 215 \pm 15 ) ~ { } \eta ^ { -0.19 } ~ { } cm ^ { -3 } .
These values are much higher than in models of the X-ray emission from the supernova ( n _ { H } \sim 75 ~ { } cm ^ { -3 } ) , and it seems plausible that the observed [ O III ] emission is produced primarily elsewhere than in the H II region .
We also discuss the type of progenitor consistent with the H II region .
In particular , it seems unlikely that its spectral type was much earlier than B2 Ia .