Five unblended lines from the 1318.5 Å spectral window were taken to represent different formation temperature regions: N I 1318.98 (10 000 K or below), O I 1306.03 (10 000 K or above, potentially influenced by O II), Si II 1309.3 (14 000 K), C II 1334.54 (23 000 K), S IV 1322.83/2, i.e. in second order (138 000 K). Line intensities were averaged over 24 exposures, respectively. Continuum was subtracted. The limb was taken to be the pixel with the strongest limb brightening of the continuum (pixel #4 for 08/10 Jul 2000, pixel #42 for 04 Jul 2000). The different angle of the SUMER slit to the limb was neglected so far.
The following five images show the line intensities vs. distance from the limb for the reference situation and the two prominences (dotted lines). Here, the prominence of 04 Jul 2000 starts at a lower altitude (except for S IV) and eventually reaches no-prominence values; the prominence of 08 Jul 2000 generally starts at a higher altitude and stays on a higher intensity level.
The next two images show the relative intensity increase (prominence value minus reference value, divided by reference value) vs. distance from limb. For clarity, all intensities were smoothed with a moving average of five pixels. The results are: Except for S IV, the prominence of 08 Jul 2000 is brighter by a factor of three; in S IV it is less bright. Both prominences have very little (but measurable) influence on N I, more on O I, Si II and S IV, but the most on C II (23 000 K). In the case of 04 Jul 2000, where the prominence starts at a lower altitude above the limb, the relative intensity increase for N I and O I peaks around 20" to 30" above the limb, followed by 30" to 40" for the peaks in Si II and C II, followed by 40" to 50" for the peak in S IV. Contrarily, in the case of 08 Jul 2000, where the prominence starts higher above the limb and the SUMER slit crosses a darker region first, the relative intensity increase peaks around 70" above the limb for N I, around 60" to 70" for O I, around 50" to 60" for Si II and C II, and around 40" for S IV.
Therefore, two types of prominences were observed, both show enhanced emission in the ultraviolet within 4<log(Te)<5 with a peak in the middle of this interval, and both show a temperature structure from the inside (cooler) to the outside (hotter).
Due to uncertainties in the exact limb position and in the straylight fitting, the interpretations should be very cautious. Nevertheless, it still appears that in the prominence of 04 Jul 2000, which is cut in its northern part, directly at the solar limb, the colder lines are relatively stronger in the southern (inner) region, while the hotter lines are relatively stronger in the northern (outer) region. In the prominence of 08 Jul 2000, which is cut in its southern part, clearly above the solar limb, the colder lines are relatively stronger in the northern (inner) region, while the hotter lines are relatively stronger in the southern (outer) region.