We present the results object by object, describing the photometric variations and illustrating with figures only the profiles where variations have been detected. No variations have been seen in the available spectra within the timescale of hours. Given the duration of the exposures, it is however evident that only markedly large changes could have been revealed this way.

To better illustrate the flux variation in the spectra we present at the bottom of each figure the absolute difference between the pair of displayed spectra. Over this difference the sum of the standard deviations of the two spectra, in bins of 200 km/s is plotted. When the level of the flux difference is above this line, we can say that the two spectra differ by more than

(provided, as it is generally the case, that the noise level is similar in the two spectra). In this way one obtains a clear quantitative representation of the zones where the two spectra differ.

In Table 2 (click here) for each pair of considered spectra the presence or not of variability is reported. In the positive case the observed velocity range is given. These ranges in some cases indicate red-shifted motions. Most probably these variations are due to changes in the receding part of the envelope contributing to the emission peak of the profile. This might indicate an asymmetry in the CSPN's wind.

**Table 2:** Observed variability in wind lines^*
Spectra	NV	OIV	OV	SiIV	CIV	NIV
NGC 40
SWP 51880/42188	no	-	no	no	+500/+1000	-
SWP 51880/52774	no	-	no	no	+500/+900	-
NGC 6543
SWP 3324/47852	no	no	see text	no	+200/+1000	no
					-1900/-1730
SWP 47852/51881	no	no	no	no	-1900/-1730	no
SWP 51881/55982	no	no	no	no	-1900/-1730	no
NGC 6826
SWP 20869/55981	no	no	-400/-1000	-	no	no
SWP 51871/55981	no	no	no	-	no	-600/0
BD+30 3639
SWP 13333/51870	-	-	no	no	+500/+800	-

In Paper I we did not see variations between the two available spectra (January 1983 and August 1991). With the addition of the new observations (August 1994 and November 1995) we see that changes did occur in the CIV line profile between the last two epochs (Fig. 2 (click here)) and also relatively to the previous data (Fig. 1 (click here)). Between August 1994 and November 1995 the emission peak of the CIV profile increased by 15%, while between August 1991 and August 1994 the same peak decreased by a 25%.

In conclusion in this star we observe P Cygni profiles of the following lines: NV, OV, CIV and SiIV and changes have been seen only in CIV.

Figure 1: NGC 40. Comparison of SWP 51880 vs. SWP 42188. The older spectrum (lower number) is plotted with the thick line. 0 km/s coresponds to the rest wavelength of the blue component of the CIV resonance doublet. In this and in all the following figures the unit of Flux is

. For the meaning of the lines at the bottom see text

Figure 2: NGC 40. Comparison of SWP 51880 vs. SWP 52774. The older spectrum (lower number) is plotted with the thick line

In Paper I we reported variations in the regions of all the P Cygni line profiles seen in the object, i.e. the lines listed in Sect. 1 (click here). But, due to evident changes in the level of the continuum and of the technique used in that article, we had not been able to disentangle the two types of variations, i.e. the photometric ones and those affecting the line profile. Now we can separate the two kinds of variations.

Between 1978 and 1993 we see a photometric decrease by a 15% followed by an increase of 12% between 1993 and 1994.

As with the variations in the line profiles, between 1978 and 1993 a decrease of 10% occurs in the emission peak of the CIV profile, as well as a decrease also of

of 170 km/s (Fig. 3 (click here)). Between 1993 and 1994 the CIV profile retains the same shape except that its edge velocity did increase by the same amount (Fig. 4 (click here)). Between 1994 and 1995 the only change in the CIV profile consists again in a reduction of its

by 170 km/s (Fig. 5 (click here)). The last two figures can be seen as examples of profiles that do not change (neglecting the portion around

As with the OV profile, we note variations only between 1993 and 1994 (Fig. 6 (click here)). These are however difficult to interpret even qualitatively.

Figure 3: NGC 6543. Comparison of SWP 3324 vs. SWP 47852. The older spectrum (lower number) is plotted with the thick line

Figure 4: NGC 6543. Comparison of SWP 51881 vs. SWP 47852. The older spectrum (lower number) is plotted with the thick line

Figure 5: NGC 6543. Comparison of SWP 51881 vs. SWP 55982. The older spectrum (lower number) is plotted with the thick line

Figure 6: NGC 6543. Comparison of SWP 3324 vs. SWP 47852. The older spectrum (lower number) is plotted with the thick line

Figure 7: NGC 6826. Comparison of SWP 20869 vs. SWP 55981. The older spectrum (lower number) is plotted with the thick line

Figure 8: NGC 6826. Comparison of SWP 51871 vs. SWP 55981. The older spectrum (lower number) is plotted with the thick line

No changes have been noted between the new spectra (August 1994 and September 1995), except for the NIV profile (Fig. 8 (click here)) where differences are seen in the absorption component which became deeper by a 15%. By comparing SWP 20869 (see Paper I) vs. SWP 55981 we have between August 1983 and August 1994 a decrease in edge velocity (120 km/s) in the OV line (Fig. 7 (click here)). We note that in this star changes are seen also in the subordinated lines.

This WC star is quite rich in emission and absorption features across the whole observed IUE spectrum. P Cygni lines are seen in OV, CIV and SiIV while the wealth of the features mimics the existence of the NV and NIV lines when observed at IUE low resolution (cf. Patriarchi & Perinotto 1991).

We have compared the images SWP 13333 (see Paper I) and SWP 51870. In CIV line (Fig. 9 (click here)) we observe a reduction of the emission component of the P Cygni profile of 20%.

Figure 9: BD +30 3639. Comparison of SWP 13333 vs. SWP 51870. The older spectrum (lower number) is plotted with the thick line

3. Results