“…From measurements taken some hours after the impacts at sites A and E, Kim et al (1996) reported "that the intensities of the H + 3 lines on the impact site are significantly reduced compared with those on the undisturbed areas, although the H + 3 emission on the undisturbed areas varies longitudinally at the same latitude." This is confirmed by Dinelli et al (1997), who submitted UKIRT observations of sites B and C to exhaustive analysis of spectra in the 3.2-4.1 µm range, but not by Encrenaz et al (1995) and Schulz et al (1995) who also detected enhanced H + 3 emission at, albeit different, sites G/D and F. Dinelli et al (1997) and Miller et al (1997b) were able to attribute H + 3 enhancements, in the case of site B, to the proximity of the southern auroral oval as well as to pointing difficulties in combination with the normal spatial variation of H + 3 and, in the case of site C, to ionospheric heating prior to impact. UKIRT also observed increased emission at site G/D which may be explained simply by limb brightening and H + 3 spatial variation rather than SL9-related increases in column density.…”