next previous
Up: Long-term OH variability of

Subsections

   
4 Results

The 7 Miras presented here cover the infrared characteristics and period range of Miras. Table 1 presents these characteristics as follows: name of the source, mean period of the optical light curve, type of the maser emission, 12, 25 and 60 $\mu $m IRAS fluxes from Cols. 1 to 6 and finally in Cols. 7 and 8, the [25 - 12] and [60 - 25] color indexes calculated from the IRAS fluxes. Infrared fluxes are the 1987 corrected IRAS fluxes.

 

 
Table 1: Main characteristics of the 7 selected Miras
Source Period1 type $S_{12 \mu {\rm m}}^2$ $S_{25 \mu {\rm m}}^2$ $S_{60 \mu {\rm m}}^2$ [25 - 12] [60 - 25]
  days   Jy Jy Jy    
R Aql 291 II 401.69 244.31 <139.49 -0.535 <-0.624
RS Vir 353 II 108.63 65.25 11.72 -0.540 -1.126
S CrB 360 I 200.68 125.52 19.00 -0.523 -1.200
R LMi 372 I 425.85 175.71 25.68 -0.703 -1.215
RR Aql 394 II 332.18 150.95 27.20 -0.661 -1.124
U Her 406 I 499.73 179.54 26.92 -0.763 -1.204
UX Cyg 561 II 171.55 101.39 43.60 -0.547 -0.747

1: Campbell L. 1985, Studies of Long Period Variables.
2: 1987 corrected IRAS flux.

Figure 1 shows the distribution of these stars in the color-color diagram of the nearest (distance < 1 kpc) Miras.


  \begin{figure} \par\includegraphics[angle=-90,width=8.8cm,clip]{ds1680f1.ps} \end{figure} Figure 1: Distribution of the 7 stars of the study in the color-color diagram of the nearby Miras (i.e., distant of less than 1 kpc from the Sun). The "?'' is due to the incertainty concerning the 60 $\mu $m IRAS flux of R Aql and thus in its [60 - 25] value

In this section the main characteristics of the OH variations are presented separately for each star. Corresponding optical light curves obtained from the AAVSO (American Association of Variable Star Observers) and the AFOEV (Association Française des Observateurs d'Étoiles Variables) are given with the OH variability curves for comparison.

For simplicity, we define the "amplitude of minimum-maximum variations'' by $\Delta I_{\rm min-max}$. This quantity is the relative difference in intensity or integrated flux which will be given in percent (i.e., ${(I_{\rm max} - I_{\rm min}) \times 100\% /(I_{\rm max} + I_{\rm min})}$), depending on the case, between a consecutive minimum and maximum in the variation curves. Moreover, we used the modified Julian day defined as follows: Julian day -2400000 which will be called "Julian day''.

   
4.1 R Aql

This star was observed in linear polarization between February 1980 and March 1983 and in circular polarization between March 1982 and March 1983 and from July 1993 to May 1995. For the linear polarization data set the four available banks were divided such that we observed both horizontal and vertical polarization only at 1667 MHz. Observations were performed in horizontal polarization at 1612 MHz while they were performed in vertical polarization at 1665 MHz. Because of the weakness of the 1665 MHz signal, observations in circular polarization were mainly performed at 1612 and 1667 MHz. Even though the velocity resolution of the linear polarization data is rather poor, the sampling is very good, with a separation between observations of 10 days to less than one month. The mean sampling in circular polarization is about 1 to 2 months.

4.1.1 Spectra

Figure 2 displays the spectra of this source in the 3 OH maser lines in both circular polarizations.
  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f2.ps}\end{figure} Figure 2: Spectra of R Aql in the 3 OH maser lines in both circular polarizations. The red and blue peaks are displayed separately because of the great flux difference between them

The red and blue peaks are presented separately because of the great flux difference between them. One can see that the blue peak is wider than the red one for the 3 OH maser lines. The general spectral shapes at 1612 and 1667 MHz are roughly similar for the two peaks except for the intensity ratio between the red and blue peaks ( $(\frac{I_{\rm red}}{I_{\rm blue}})_{1667} = 0.25$ and $(\frac{I_{\rm red}}{I_{\rm blue}})_{1612} = 20.0$ for the period presented in Fig. 2). However, the red part of the emission at 1612 MHz goes beyond the 1667 MHz emission by 1 km s-1. On the other hand, the 1665 MHz profile is quite different from the two other profiles. Moreover, the 1665 MHz emission covers a smaller velocity range than the 1667 MHz emission. Indeed, both peaks at 1667 MHz go beyond those at 1665 MHz by more than 3 km s-1.


  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f6.ps}\end{figure} Figure 3: Spectra of RS Vir in the 3 OH maser lines in both circular polarizations


  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f13.ps} \end{figure} Figure 4: Spectra of S CrB in the 3 OH maser lines in both circular polarizations


  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f20.ps} \end{figure} Figure 5: Spectra of R LMi at 1665 and 1667 MHz in both circular polarizations


  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f24.ps} \end{figure} Figure 6: Spectra of RR Aql in the 3 OH maser lines in both circular polarizations


  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f30.ps} \end{figure} Figure 7: Spectra of U Her in both main lines and circular polarizations


  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f36.ps} \end{figure} Figure 8: Spectra of UX Cyg in the 3 OH maser lines and both circular polarizations


  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f3.ps} \end{figure} Figure 9: Integrated flux variation curves of R Aql in the 3 OH maser lines in linear b, c, d, e, f and circular polarization g, h, i, j for the first period (i.e., from February 1980 to March 1983) and in circular polarization l, n, m, o only for the second period (i.e., from July 1993 to May 1995) for the two standard peaks (labelled Blue and Red in the figure)

4.1.2 Integrated flux

In Fig. 9 are presented the integrated flux variations of the red and blue peaks in linear and circular polarizations for the two epochs of observations.

The fitted curve of variations shows a delay of about 40-50 days with respect to the optical one (cf. Table 2).

   
Table 2: Results of the OH curve fiting
Source line1 Pol.2 Peak3 Period4 $f_{\rm0 \; OH}^{4,5}$ $f_{\rm0 \; optical}^5$ $t_{\rm OH \: max}^4$ $t_{\rm optical \; max}$ OH delay 6
  MHz     days     days days days
R Aql 1612 H R 285 0.57 0.43 44916 44872 $\pm$ 5 44 $\pm$ 10
RS Vir 1665 RHC B 353 0.48 0.37 49333 49320 $\pm$ 15 (*) 13 $\pm$ 20
S CrB 1665 RHC R 380 0.47 0.35 49327 49278 $\pm$ 5 49 $\pm$ 10
R LMi 1665 RHC B 392 0.57 0.41 49638 49586 $\pm$ 10 52 $\pm$ 15
RR Aql 1665 LHC B 394 0.56 0.31 49318 49249 $\pm$ 10 69 $\pm$ 15
U Her 1667 LHC B 393 0.63 0.40 48330 48253 $\pm$ 5 77 $\pm$ 10
UX Cyg 1612 LHC B 562 0.43 0.40 49617 49536 $\pm$ 5 81 $\pm$ 10

1: Line in which the curve fitting had been made.
2: Horizontal polarization (H), right- (RHC) and left-handed (LHC) polarizations.
3: Red (R) and blue (B) peaks.
4: Deduced from the OH curve fitting. The uncertainty for the calculated period and OH maximum date is of $\pm$ 5 days.
5: f0 = [rising time from a minimum toward a maximum]/period.
6: $t_{\rm OH \; max}-t_{\rm optical \; max}$.
*: This value was obtained by superimposing 2 optical cycles for a better definition of the maximum.


a1) Red peak at 1612 MHz
 

 
Table 3: Central velocity and full width at half-maximum (FWHM) of the fitted components which have shown a longevity greater than 1 stellar period of R Aql at 1612 and 1667 MHz for the 2 intervals of observations in circular polarization. For comparison are given too in the last column the velocity of the components determined by Herman & Habing at 1612 MHz (1985, i.e., column labelled H&H)
  LHC FWHM RHC FWHM H&H
  km s-1 km s-1 km s-1 km s-1 km s-1
interval 11 V1=54.00 0.814 V1=54.00 0.810 54.04
  V2=52.91 0.752 V2=52.92 0.800  
interval 22 V1=54.00 0.843 V1=54.00 0.835  
  V2=52.95 0.849 V2=52.86 0.700  

a2) Blue peak at 1612 MHz



  LHC FWHM RHC FWHM H&H
  km s-1 km s-1 km s-1 km s-1 km s-1
interval 11 V1=43.54 0.727 V1=43.53 0.708 43.45
      V2=42.88 0.504  
  V3=42.40 0.420 V3=42.28 0.475  
  V4=41.86 0.334 V4=41.90 0.336  
  V5=41.51 0.404 V5=41.46 0.372  
  V6=41.07 0.402 V6=41.03 0.345 41.01
  V7=40.58 0.525 V7=40.55 0.370  
  V8=39.98 0.529 V8=40.11 0.480  
interval 22 V1=43.54 0.733 V1=43.54 0.689  
  V2=42.85 0.423 V2=42.84 0.463  
  V3=42.37 0.341 V3=42.36 0.327  
  V4=41.92 0.385 V4=41.93 0.353  
  V5=41.47 0.333 V5=41.50 0.340  
  V6=41.04 0.393 V6=41.05 0.408  
  V7=40.61 0.390 V7=40.60 0.440  
  V8=40.19 0.468 V8=40.13 0.451  
  V9=39.58 0.510 V9=39.53 0.558  


b1) Red peak at 1667 MHz
  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
interval 11 V1=53.48 0.683 V1=53.47 0.706
  V2=52.82 0.603 V2=52.89 0.616
  V3=52.46 0.452 V3=52.44 0.633
  V4=51.82 0.652 V4=51.77 0.565
interval 22 V1=53.43 0.762 V1=53.47 0.731
  V2=52.69 0.587 V2=52.59 0.511
  V3=51.78 0.573 V3=52.01 0.524
      V4=50.91 0.813
b2) Blue peak at 1667 MHz



  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
interval 11     V1=44.98 0.465
  V2=44.45 0.569 V2=44.39 0.485
  V3=43.55 0.819 V3=43.56 0.896
  V4=42.52 0.654 V4=42.57 0.637
  V5=41.84 0.681 V5=41.87 0.587
  V6=41.17 0.760 V6=41.20 0.590
  V7=40.13 0.654    
interval 22     V1=45.31 0.647
  V2=44.41 0.489 V2=44.40 0.516
  V3=43.44 0.942 V3=43.54 0.839
      V4=42.80 0.507
  V5=42.40 0.540 V5=42.29 0.453
  V6=41.85 0.544 V6=41.83 0.572
  V7=41.33 0.509 V7=41.24 0.474
  V8=40.76 0.798 V8=40.79 0.608

1: Interval of Julian days [45000 - 45400].
2: Interval of Julian days [49150 - 49850].

The general shape of the optical and OH variation curves is similar: a secondary maximum can be seen in both of them (cf. Figs. 9a,e).

This source is rather weakly circularly polarized at 1612 as well as at 1667 MHz (-0.12<[RHC-LHC] 1667<0.0 and -0.15<[RHC-LHC] 1612<0.15). Moreover, the degree of polarization is presumably not significant for the first set of data (i.e., between 1980 and 1983) because of expected calibration errors as previously mentioned in the Sect. 2.

The strongest line is clearly the satellite 1612 MHz line which can show integrated flux up to four times greater than the 1667 MHz line. The 1665 MHz flux is the faintest, often below the detection threshold. Nevertheless, the largest $\Delta I_{\rm min-max}$ are reached for the 1665 and 1667 MHz integrated fluxes (cf. Table 4a). Moreover, we can clearly see from the first set of observations in linear polarization (i.e., from Figs. 9b to 9f) that the differences in $\Delta I_{\rm min-max}$ from one cycle to another at 1612 MHz are about the same while those differences are rather strong in the main lines.

Clearly, from one cycle to another the variations of the blue peak OH curve maxima are not correlated with the red one in any line (Figs. 9b-f). Nevertheless, variations at 1665 and 1667 MHz are similar for the same peak. This mimic behaviour between the 1665 and 1667 MHz OH variations may be explained by an overlap in the location of the maser emitting region in these two lines and shows moreover that the degree of non-saturation is comparable for both main lines. In contrast, the 1612 MHz variations are rather similar in both standard peaks and changes in the amplitude of variations from a cycle to another are very small. This confirms a higher level of saturation in this latter line in comparison to the main lines.

 

 
Table 4: $\Delta I_{\rm min-max}$ for R Aql
a) $\Delta I_{\rm min-max}$ for the integrated flux
Frequency peak $\Delta I_{\rm min-max}$
(MHz)   interval 1* interval 2*
    polar. polar.
    lin. cir. cir.
1612 blue 70% 31% 50%
  red 58% 24% 36%
1667 blue 55, 70%** 71% 48%
  red 60, 82%** 79, 94%** 72%
1665 blue 89, 80%**    
  red 95, 78%**    

b) Range and mean value of $\Delta I_{\rm min-max}$ for the components of great longevity

Freq. Peak $\Delta I_{\rm min-max}$
(MHz)   Range Mean Value
    interval interval interval interval
    1* 2* 1 2
1612 blue 10 to 48% 35 to 72% 29% 51%
  red 21 to 31% 25 to 46% 27% 34%
1667 blue 37 to 81% 32 to 58% 62% 43%
  red 45 to 92% 46 to 67% 69% 59%

*: Same intervals as given Table 3.
**: Respectively for the first and the second cycle observed in this interval of observations.

4.1.3 Spectral components

Figures 10 and 11 display the variations with time of the fitted components at 1612 and 1667 MHz respectively in both circular polarizations for the 2 sets of observations, which correspond in Julian dates to [45000 - 45400] and [49150 - 49850].


  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f4.ps} \end{figure} Figure 10: Intensity variation curves of the Gaussian fitted components of R Aql at 1612 MHz in left- (LHC) and right-handed (RHC) polarizations for the red (Red) and blue (Blue) peaks from March 1982 to March 1983 and from July 1993 toMay 1995


  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f5.ps} \end{figure} Figure 11: Intensity variation curves of the Gaussian fitted components of R Aql at 1667 MHz in left- (LHC) and right-handed (RHC) polarizations for the red (Red) and blue (Blue) peak from March 1982 to March 1983 and from July 1993 toMay 1995

The Gaussian fitting of the components brings out the complexity of the blue peak at 1667 as well as at 1612 MHz. One can distinguish 8 and 9 components respectively with a lifetime greater than one stellar period. On the other hand, the red peak shows only 5 and 2 components respectively at 1667 and 1612 MHz (cf. Fig. 2).

Table 4b gives the range and mean value of the $\Delta I_{\rm min-max}$ for the intensity of the components exhibiting a longevity greater than one cycle at 1612 and 1667 MHz.

At 1612 MHz the various components are highly stable: they were observed during the whole 15 year span. At this frequency, the blue peak components show a degree of polarization less than 10% except for the one located at V=+39.55 km s-1 during the second set of observations which shows quite a strong degree of left-handed polarization ([RHC-LHC] =-0.27). In contrast both components of the red peak (at V=+52.91 km s-1 and V=+54.00 km s-1) show a degree of right-handed polarization as high as [RHC-LHC] =0.17 in the first set of observations and less than 10% in the second set.

At 1667 MHz the stability of components is shorter, since some of them disappeared between the first and second set of observations (less than 10 years) while others appeared. At this frequency, the degree of polarization of the whole set of red and blue components is less than 11% for both sets of observations except for the blue component located at V=+40.78 km s-1 which had a degree of circular polarization as high as [RHC-LHC] =0.24 during the second set of observations.

One may note that the peak exhibiting the strongest integrated flux (the red peak at 1612 MHz and the blue peak at 1667 MHz) shows a fainter mean value for $\Delta I_{\rm min-max}$ of the components in comparison with the companion peak for the same line (Table 4).

Thus, there is a correlation between the observed difference in the emission strength between the front and the back part and the degree of saturation in those regions such that the strongest emission shows the strongest saturation.

   
4.2 RS Vir

The first set of observations for this star was performed between January 1982 and March 1983 and the second set between May 1993 and June 1994. According to its cyclic periodicity of about 350 days (cf. Table 1), both time spans cover a bit more than a cycle. Unfortunately, the sampling at 1612 MHz for the first set of data is not regular with a gap around the OH minimum. It is regular for the second set, with an observation performed every 1.5 months. The sampling in the main lines for both sets of data is good with, on the average, an observation per month.


 

 
Table 5: Central velocity and full width at half-maximum (FWHM) of the fitted components which have shown a longevity greater than 1 stellar period of RS Vir at 1612 MHz for the 2 intervals of observations. For comparison are given too in the last column the velocity of the components determined by Herman & Habing (1985, i.e., column labelled H&H)
a) Red peak
  LHC FWHM RHC FWHM H&H
  km s-1 km s-1 km s-1 km s-1 km s-1
interval 11 V1=-10.37 0.859 V1=-10.36 0.845 -10.25
  V2=-11.31 0.840 V2=-11.26 0.771  
interval 22 V1=-10.32 0.855 V1=-10.31 0.850  
  V2=-11.25 0.793 V2=-11.26 0.735  

b) Blue peak

  LHC FWHM RHC FWHM H&H
  km s-1 km s-1 km s-1 km s-1 km s-1
interval 11 V1=-15.41 0.451 V1=-15.40 0.453 -15.29
  V2=-16.52 0.870 V2=-16.52 0.975 -16.48
  V3=-17.51 0.527 V3=-17.51 0.524 -17.53
  V4=-18.27 0.627 V4=-18.27 0.606 -18.43!
  V5=-18.86 0.697 V5=-18.82 0.689  
  V6=-19.75 0.587 V6=-19.76 0.525 -19.71
interval 22 V1=-15.37 0.464 V1=-15.40 0.467  
  V2=-16.50 0.900 V2=-16.50 0.889  
  V3=-17.56 0.579 V3=-17.54 0.655  
  V4=-18.25 0.501 V4=-18.27 0.505  
  V5=-18.87 0.750 V5=-18.88 0.705  
  V6=-19.86 0.510 V6=-19.85 0.552  
1: Interval of Julian days [45000 - 45400].
2: Interval of Julian days [49100 - 49550].
!: Corresponding to the mean velocity of components 4 and 5 when there are tightly blended due to lower velocity resolution
V4+ V5)/2 = -18.56 km s-1.


4.2.1 Spectra

Like the previous star, RS Vir is a type II emitter (i.e., strongest at 1612 MHz). Its spectra are displayed in Fig. 3. As can be clearly seen, this star exhibits an intermediate group of components in the main lines which can reach relatively high intensities with respect to the red peak at 1665 MHz.

4.2.2 Integrated flux

In Figs. 12 and 13 are displayed the OH integrated flux variations of the blue, red and intermediate peaks for the 3 maser lines in both circular polarizations.


  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f7.ps} \end{figure} Figure 12: Integrated flux variation curves of RS Vir in the 3 OH maser lines in circular polarization (LHC: circles/RHC: triangles) for the first set of observations (i.e., from January 1982 to March 1983) for the blue (left column i.e., "b,d,f" schemes) and red (right column i.e., "c,e,g" schemes) standard peaks as well as for the intermediate one ( "h,i'' schemes)


  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f8.ps} \end{figure} Figure 13: Source: RS Vir. The same as for the previous figure, but for the second set of observations (i.e., from May 1993 to June 1994)

A general increase of the integrated flux in both main lines is seen, the greatest increase being at 1665 MHz. At 1612 MHz, one can note a rather small increase in the integrated flux of the blue peak and a small decrease for the red peak with approximately the same ratio.

The $\Delta I_{\rm min-max}$ are the weakest at 1612 MHz, the greatest value being only 46% (i.e., $I_{\rm max}/I_{\rm min} \simeq$ 2.7). Here again the greatest $\Delta I_{\rm min-max}$ are observed in the 1665 and 1667 MHz integrated fluxes with respective values of 91% (i.e., $I_{\rm max}/I_{\rm min} \simeq$ 20) and 89% (i.e., $I_{\rm max}/I_{\rm min} \simeq$ 17). We can note smaller values of $\Delta I_{\rm min-max}$ (about 10-15% less) for the second set of observations in comparison with the first set, for the three standard maser line emissions (cf. Table 8).

Two hypotheses can explain this general decrease of $\Delta I_{\rm min-max}$: (1) a change in the degree of saturation, which in the present case must involve a mechanism affecting the 3 lines simultaneously. The most probable cause is a change in the OH density. (2) A weakening in the radiation of the pumping source shared by the 3 maser lines.

The measurement of the delay between the OH and optical maxima was determined using the second set of data at 1665 MHz. Because of the small amount of data, the optical maximum was measured by the superposition of 2 consecutive cycles for a better definition. The calculated delay is $13 \pm 20$ days (cf. Table 2). Nevertheless, for first set of data (i.e., Figs. 12d,e,f,h, and i) the delay is about 50-70 days. Thus, the most realistic value for the delay is certainly between the two values, about 30 to 50 days.


 

 
Table 6: Central velocity and full width at half-maximum (FWHM) of the fitted components which have shown a longevity greater than 1 stellar period of RS Vir at 1667 MHz for the 2 intervals of observations
a) Red peak
  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
interval 11 V1=-10.18 1.251 V1=-10.18 1.265
  V2=-11.38 1.177 V2=-11.45 1.111
interval 22 V1=-10.07 0.944 V1=-10.08 0.947
  V2=-10.99 0.847 V2=-10.92 0.822
  V3=-11.84 0.896 V3=-11.80 0.846

b) Intermediate velocity peak

  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
interval 11 V1=-13.91 1.390 V1=-13.91 1.502
interval 22 V1=-13.47 1.311 V1=-13.03 1.022
      V2=-14.14 1.233

c) Blue peak

  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
interval 11 V1=-16.36 0.899 V1=-16.36 0.927
  V2=-17.52 1.016 V2=-17.52 1.025
      V3=-18.45 1.021
interval 22 V1=-16.50 1.061 V1=-16.40 1.019
  V2=-17.63 0.928 V2=-17.49 0.950
  V3=-18.44 0.876 V3=-18.29 0.943
  V4=-19.49 0.967 V4=-19.48 1.035

1 and 2: Same intervals as given in the previous table.



 

 
Table 7: Source: RS Vir. The same as for the previous table but for the 1665 MHz
a) Red peak
  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
interval 11 V1=-10.25 0.677 V1=-10.06 0.677
  V2=-11.11 0.769 V2=-11.10 0.956
interval 22 V1=-10.82 1.111 V1=-10.75 1.009
  V2=-11.89 0.933 V2=-11.84 0.927

b) Intermediate velocity peak

  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
interval 11 V1=-13.40 0.674 V1=-13.23 0.690
  V2=-14.53 0.702 V2=-14.64 0.777
interval 22 V1=-13.26 0.925 V1=-13.18 0.854
  V2=-14.51 0.988 V2=-14.69 1.186

c) Blue peak

  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
interval 11 V1=-16.26 1.007 V1=-16.34 0.975
  V2=-17.62 0.589 V2=-17.65 0.780
interval 22 V1=-16.41 1.055 V1=-16.47 1.037
  V2=-17.69 0.649 V2=-17.72 0.679

1 and 2: Same intervals as given in Table 5.



 

 
Table 8: $\Delta I_{\rm min-max}$ for the integrated flux of RS Vir
Frequency peak $\Delta I_{\rm min-max}$
(MHz)   interval interval
    1* 2*
1612 blue 40% 29%
  red 46% 31%
1667 blue 82% 73%
  red 79% 62%
  intermediate 85% 89%
1665 blue 85% 58%
  red 91% 83%
  intermediate 89% 78%

*: Same intervals as given in Table 5.


The most polarized maser emission is at 1612 MHz (cf. Table 9). The emission at this frequency is right-hand polarized in both data sets. The 1667 MHz maser emission of both standard peaks was also right-hand polarized in the first set, showing a smaller mean value, while the degree of polarization is zero for the whole of second set. The polarization of the standard peaks at 1665 MHz is very weak (<5%) for both data sets.

The signal lying between the two standard peaks shows no polarization at all at 1667 MHz for all observations. On the other hand, at 1665 MHz it shows a weak left-handed polarization with a value of [RHC-LHC] 1665 = -0.1 during the OH maximum of the first data set and a very faint right-handed polarization on the second set. This latter behaviour is inverted compared with the standard 1665 MHz emission. Thus, we do not have a correlation between the behaviour of the intermediate and standard peak emission with regard to the polarization, suggesting that the zones concerned with the two emissions possess different characteristics.

4.2.3 Spectral components

Figures 14-17 show the variations during the two epochs of observations of the fitted components respectively at 1665, 1667 and 1612 MHz for the 3 peaks (i.e., the blue, the red and the intermediate one) and in both circular polarizations.
  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f9.ps} \end{figure} Figure 14: Intensity variation curves of the Gaussian fitted components of RS Vir at 1665 MHz in left- (LHC) and right-handed (RHC) polarizations for the red, blue and intermediade (Int.) peaks from January 1982 to March 1983 and from May 1993 to June 1994


  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f10.ps} \end{figure} Figure 15: Source: RS Vir. The same as for the previous figure but for the 1667 MHz


  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f11.ps} \end{figure} Figure 16: Intensity variation curves of the Gaussian fitted components of RS Vir of the 1612 MHz blue peak in left- (LHC) and right-handed (RHC) polarizations from January 1982 to March 1983 and from May 1993 to June 1994


  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f12.ps}\end{figure} Figure 17: Source: RS Vir. The same as for the previous figure but for the 1612 MHz red peak


 

 
Table 9: Mean value of [RHC-LHC] for the standard peak integrated flux of RS Vir
Frequency [RHC-LHC]
(MHz) interval interval
  1* 2*
1612 0.2 0.05
1667 0.1 0
1665 $-0.05<[{\rm RHC-LHC}]<0.05$ $-0.05<[{\rm RHC-LHC}]<0.05$

*: Same intervals as given in Table 5.


The complexity of both standard peaks is of the same order in the main lines while a large difference in complexity can be seen between the red and blue peaks at 1612 MHz, the latter exhibiting a great number of components (cf. Fig. 3). Here again component longevity in the main lines is much shorter than in the satellite line. The range and mean value of the $\Delta I_{\rm min-max}$ for the intensity of the components with a longevity greater than one cycle for the three OH maser lines are given Table 10a.

For the main lines, a few components (<4) in both standard peaks show a longevity greater than one cycle. Otherwise, components appear and disappear from one cycle to the next. At 1665 MHz, degrees of polarization exceeding 10% are only observed for 2 components in the first set of observations (cf. Table 10b). At 1667 MHz, none of the components exhibits a degree of polarization greater than 10%. The main line emission of the intermediate peak of this source seems not to have any peculiarity in comparison with the blue and red standard peaks. Indeed, its temporal variations are quite similar to both standard peaks and its polarization is about the same.

At 1612 MHz, 6 components in the blue peak and 2 in the red show a long lifetime. At this frequency, the $\Delta I_{\rm min-max}$ are quite a bit smaller than in the main lines. Except the component located at V=-18.84 km s-1 (with [RHC-LHC] =0.08) all the components with a longevity greater than one stellar cycle contribute to the right-handed polarization observed for the 1612 MHz integrated flux during the first set of observations (cf. Table 10b). During the second set of observations all components show degrees of polarization smaller than 10% for both peaks.

   
4.3 S CrB

This source was observed at three different epochs: from January 1982 to March 1983, from January 1986 to July 1987 and from April 1993 to April 1995. During the first interval, the observations in the 1612 MHz satellite line were performed during a stellar period but unfortunately centered on an OH minimum; moreover the sampling was not regular. In the main lines, the observations covered almost 1.5 stellar cycles with good sampling: on average an observation every 1.5 month at 1665 MHz and once per month at 1667 MHz. The second epoch covered a bit more than 1.5 cycles. For these data, observations were performed in both circular polarizations only at 1667 MHz while they were performed in the left-handed polarization at 1612 MHz and in the right-handed polarization at 1665 MHz. Finally, during the third epoch the observations cover almost 2 cycles at 1612 MHz and exactly 2 cycles in the main lines. The sampling in the main lines is very good with a separation between two observations less than one month around minima and maxima. In the 1612 MHz satellite line, the sampling is on the average one observation every 1.5 months.

4.3.1 Spectra

Figure 4 shows the spectra of this source in the 3 OH lines and in both circular polarizations. One can see that the most extended spectrum here is at 1667 MHz which extends beyond the 1612 MHz profile by about 2 km s-1for the bluest part of the profile and 3 km s-1 for the reddest part. Moreover, the shape of the profiles is quite different from one line to another, especially between the main and the satellite lines. Thus, the red peak of the 1667 MHz clearly shows 2 main groups of components, centered respectively at 3.5 and 6.2 km s-1, with a flux ratio of about 2. This ratio is of more than 7 for the 1665 MHz red peak emission. This may be explained by a weaker saturation in the 1665 MHz line than in the 1667 MHz. Moreover, the 1667 MHz blue peak is quite large, with a total spread of about 7 km s-1, showing at least 3 distinct main groups of components of faint intensity. The 1665 MHz blue peak exhibits a flater shape and no predominent groups of components are seen. At 1612 MHz, 3 main groups of components are evident in the blue peak but they are narrower and do not correspond to those at 1667 MHz. This remark is also valid for the 2 main groups of components observed in the 1612 MHz red peak in comparison with the 1667 MHz ones.

4.3.2 Integrated flux

Variations of the integrated flux of the 3 OH maser lines for the 3 epochs of observations are presented Figs. 18 and 19 respectively for the red and the blue peaks.


 

 
Table 10: Results for the components of great longevity of RS Vir
a) Range and mean value of $\Delta I_{\rm min-max}$
Freq. Peak $\Delta I_{\rm min-max}$
(MHz)   Range Mean Value
    interval interval interval interval
    1* 2* 1 2
1612 blue 19 to 58% 16 to 43% 34% 29%
  red 30 to 64% 28 to 51% 49% 38%
1667 blue 69 to 86% 36 to 60% 80% 52%
  red 77 to 80% 45 to 67% 78% 57%
  inter. 78 to 81% 42 to 55% 79% 48%
1665 blue 78 to 87% 49 to 69% 84% 58%
  red 21 to 89% 60 to 79% 59% 71%
  inter. 45 to 89% 46 to 68% 74% 55%

*: Same intervals as given in Table 5.

b) |[RHC-LHC]$\vert\ge $0.10 for the interval of Julian day [45000-45400]

Freq. peak $V_{\rm comp.}$ [RHC-LHC]
(MHz)   (km s-1)  
1612 blue -15.40 0.25
  blue -16.52 0.26
  blue -17.51 0.18
  blue -18.27 0.24
  blue -19.75 0.30
  red -10.36 0.30
  red -11.26 0.20
1665 red -11.10 -0.18
  inter. -14.60 -0.21



  \begin{figure} \par\includegraphics[angle=-90,width=8.8cm,clip]{ds1680f14.ps}\end{figure} Figure 18: Integrated flux variation curves of S CrB in the 3 OH maser lines in circular polarization from January 1982 to March 1983, from January 1986 to July 1987 and from April 1993 to April 1995 for the red peak. First row (i.e., boxes  a,b,c): 1612 MHz, second row (i.e., boxes  d,e,f): 1667 MHz and third row (i.e., boxes  g,h,i): 1665 MHz


  \begin{figure} \par\includegraphics[angle=-90,width=8.8cm,clip]{ds1680f15.ps}\end{figure} Figure 19: Source: S CrB. The same as for the previous figure, but for blue peak

One can observe the usual phase delay between the OH and optical maxima, calculation gives a value of about $50 \pm 10$ days (cf. Table 2). Comparing the general shape of the OH and optical curves it is clear that the asymmetry of the OH curves is less than that of the optical curve. This is due to a very small (if any) delay between OH and optical minima in comparison with a distinct delay between the OH and optical maxima. This can be clearly seen Figs. 18f and 18i in comparing with the corresponding part of the optical curve, because of the very good sampling at the OH minima and maxima.

The variations of integrated flux amplitude from one cycle to another are here again quite similar for both main lines. Large variations in the values of the integrated flux maxima in the blue peak at 1665 as well as at 1667 MHz across all 3 epochs of observations are seen. The ratio of greatest value reached by the OH maximum to the smallest in the blue peak is 4 and 2-3 respectively at 1665 and 1667 MHz while a weaker ratio was measured in the red peak, less than 1.6 and 1.2 respectively at 1665 and 1667 MHz. On the other hand, the value of the integrated flux maxima at 1612 MHz did not change by more than a factor of 1.3 for both standard peaks.


 

 
Table 11: Central velocity and full width at half-maximum (FWHM) of the fitted components which have shown a longevity greater than 1 stellar period of S CrB at 1612 MHz for each interval of observations
a) Red peak
  LHC FWHM RHC FWHM
    km s-1   km s-1
interval 11 V1=4.31 0.861 V1=4.36 0.777
  V2=3.21 0.734 V2=3.20 0.757
interval 22 V1=4.29 0.830 (*)  
  V2=3.20 0.897    
interval 33 V1=4.24 0.879 V1=4.23 0.814
  V2=3.04 0.653 V2=3.05 0.723

b) Blue peak

  LHC FWHM RHC FWHM
    km s-1   km s-1
interval 11 V1=-2.21 0.638 V1=-2.24 0.674
  V2=-2.94 0.761 V2=-3.02 0.709
  V3=-3.89 0.956 V3=-3.92 0.966
interval 22 V1=-2.15 0.647    
  V2=-2.94 0.890 (*)  
  V3=-4.13 0.765    
interval 33 V1=-2.26 0.587 V1=-2.29 0.658
  V2=-3.02 0.738 V2=-3.10 0.669
  V3=-4.14 0.658 V3=-4.15 0.623

1: Interval of Julian days [45000 - 45400].
2: Interval of Julian days [46400 - 47000].
3: Interval of Julian days [49100 - 49800].
(*): No observations at this polarization for this period.



 

 
Table 12: Source: S CrB. The same as for the previous table but for the 1667 MHz
a) Red peak
  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
interval 11 V1=7.23 0.629 V1=7.12 0.711
  V2=5.93 0.830 V2=5.88 0.831
  V3=5.02 0.660 V3=5.16 0.554
      V4=4.67 0.803
  V5=4.22 0.802 V5=4.09 0.763
  V6=3.35 0.956 V6=3.34 0.928
interval 22 V1=5.82 0.817 V1=5.82 0.879
  V2=4.50 1.003 V2=4.48 1.015
  V3=3.39 0.927 V3=3.36 0.938
interval 33 V1=6.34 0.419 V1=6.27 0.345
  V2=5.76 0.603 V2=5.78 0.537
  V3=5.30 0.430 V3=5.23 0.451
  V4=4.68 0.619 V4=4.67 0.551
  V5=3.74 0.956 V5=3.89 0.843
  V6=2.92 0.782 V6=3.04 0.826

b) Blue peak

  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
interval 11 V1=-0.86 1.071 V1=-0.80 1.197
  V2=-2.44 1.333 V2=-2.42 1.435
interval 22 V1=-0.71 1.215 V1=-0.59 1.204
  V2=-2.41 1.590 V2=-2.39 1.512
  V3=-4.35 1.242 V3=-4.08 1.280
  V4=-6.03 1.299 V4=-6.01 1.314
interval 33 (*)   (*)  

1,2 and 3: Same intervals as given in the previous table.
(*): No fitting for this period due to the faintness of the signal.



 

 
Table 13: Source: S CrB. The same as for the previous table but for the 1665 MHz
a) Red peak
  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
interval 11 V1=4.41 0.480    
  V2=3.90 0.412 V2=4.04 0.696
  V3=3.31 0.788 V3=3.21 0.812
interval 22 (*)   V1=4.85 0.552
      V2=4.05 0.754
      V3=3.14 0.938
interval 33 V1=5.29 0.542 V1=5.30 0.454
  V2=4.61 0.484 V2=4.72 0.507
  V3=3.75 0.741 V3=3.93 0.782
  V4=2.94 0.844 V4=3.03 0.917

1, 2 and 3: Same intervals as given in the Table 11.
(*): No observations at this polarization for this period.
b) Blue peak.
No fitting were made for the third period for this peak due to the faintness of the signal.



 

 
Table 14: $\Delta I_{\rm min-max}$ for the integrated flux of S CrB
Frequency peak $\Delta I_{\rm min-max}$
(MHz)   interval interval interval
    1* 2* 3*
1612 blue 30% 37% 50%
  red 37% 14% 41%
1667 blue 60% 59% 81%
  red 54% 45% 58%
1665 blue 73% 69% 79%
  red 55% 37% 54%

*: Same intervals as given in the Table 11.


The values of $\Delta I_{\rm min-max}$ are given Table 14. Notice that the $\Delta I_{\rm min-max}$ of the integrated flux in both main lines were about the same and systematically greater than in the 1612 MHz satellite line. Further more, the greatest values of $\Delta I_{\rm min-max}$ in each line were all observed for the blue peak integrated flux while the faintest, observed in the second set of observations (i.e., interval of Julian days [46400 - 47000]), were all obtained for the red peak.

From the first and third interval of observations, for which both circular polarizations are available, the 1612 MHz emission was mainly weakly right-hand polarized. The 1667 MHz emission has shown a change in its behaviour, since it was first right-hand polarized with a degree of circular polarization reaching a value of [RHC-LHC] 1667= 0.2 in the first data set while it shows no degree of polarization for the second and third sets of data for both standard peaks. As for the 1667 MHz, the 1665 MHz line shows polarized emission for the first data set and no degree of polarization for third. But, contrarily to the 1667 MHz, the behaviour at 1665 MHz is different for the red and blue peaks. Indeed, while the red 1665 MHz peak emission shows a faint left-handed polarization (with the strongest value of about [RHC-LHC] 1665= -0.15) in the first set of observations, the blue 1665 MHz peak emission shows a right-handed polarization with a maximum value [RHC-LHC] 1665= 0.33 during the second maximum around the Julian day 45340.

4.3.3 Spectral components

Figures 20-23 display the temporal variations of the various fitted spectral components in the 3 OH maser lines, for the 3 epochs.
  \begin{figure} \includegraphics[angle=-90,width=8.8cm,clip]{ds1680f16.ps} \end{figure} Figure 20: Intensity variation curves of the Gaussian fitted components of S CrB at 1612 MHz in left- (LHC) and right-handed (RHC) polarizations for the two standard peaks for the 3 sets of observations (i.e., from January 1982 to March 1983, from January 1986 to July 1987 and from April 1993 to April 1995)


  \begin{figure} \par\includegraphics[angle=-90,width=8.8cm,clip]{ds1680f17.ps} \end{figure} Figure 21: Source: S CrB. The same as for the previous figure but for the 1665 MHz red peak. The blue peak was too faint for a reasonably good fitting


  \begin{figure} \par\includegraphics[angle=-90,width=8.8cm,clip]{ds1680f18.ps} \end{figure} Figure 22: Intensity variation curves of the Gaussian fitted components of S CrB at 1667 MHz in left- (LHC) and right-handed (RHC) polarizations for the red peak for the 3 set of observations (i.e., from January 1982 to March 1983, from January 1986 to July 1987 and from April 1993 to April 1995)


  \begin{figure} \par\includegraphics[angle=-90,width=8.8cm,clip]{ds1680f19.ps} \end{figure} Figure 23: Source: S CrB. The same as for the previous figure but for the blue peak

At 1612 MHz, all the components detected in the first data set could be observed in the next two epochs. Thus for this line, (i.e., Fig. 20) and for simplicity, only the mean value of the velocity of each component is given because of their great stability.

The range and mean value of the intensity $\Delta I_{\rm min-max}$ for those components exhibiting a longevity greater than one cycle for the three OH maser lines are given Table 15a. At 1612 MHz, the ranges of amplitudes of variations are systematically greater for the blue peak components than for those of the red peak. This can be interpreted as a greater degree of saturation for the 1612 MHz emission coming from the back part of the shell than from the front. This behaviour clearly demonstrates the inhomogeneity between the front and the back parts of the shell.

The |[RHC-LHC]$\vert\ge 0.10$ for the components of great longevity are given in Table 15b. We note that the behaviour of the components with regard to the polarization is quite different from one line to another, and surprisingly between both main lines. Even though the integrated flux at 1667 MHz for both standard peaks exhibits a degree of circular polarization greater than 10% in the first interval of observations, none of the components with a great longevity in the blue peak shows |[RHC-LHC]|> 0.10. On the contrary, while the red peak integrated flux at 1665 MHz shows [RHC-LHC] =0 in the third interval of observations, three spectral components with a great longevity appear to be strongly polarized (Table 15b).

   
4.4 R LMi

R LMi was observed during one stellar period between January 1982 and March 1983 and during almost 6 consecutive cycles from August 1989 to November 1995. This star is a type I emitter and, up to June 1994, no significant 1612 MHz emission could be detected. The eruptive 1612 MHz emission was studied by Etoka & Le Squeren (1997); thus, only the maser emission of the main lines is presented here.

4.4.1 Spectra

The spectra in both main lines and circular polarizations are displayed Fig. 5.
 

 
Table 15: Results for the components of great longevity of S CrB
a) Range and mean value of $\Delta I_{\rm min-max}$
Freq. Peak $\Delta I_{\rm min-max}$
(MHz)   Range Mean Value
    interval interval interval interval interval interval
    1* 2* 3* 1 2 3
1612 blue 19 to 36% 36 to 53% 33 to 55% 27% 44% 45%
  red 24 to 40% 20 to 27% 33 to 40% 31% 23% 37%
1667 blue 42 to 58% 36 to 58%   50% 49%  
  red 31 to 66% 14 to 50% 43 to 75% 49% 33% 53%
1665 blue            
  red 30 to 71% 39 to 40% 46 to 61% 47% 39% 53%

b) |[RHC-LHC]$\vert\ge $0.10

freq. peak V $_{\rm comp}$ interval [RHC-LHC]
MHz   (km s-1) *  
1612 blue -2.23 1 $\simeq 0.15$
  blue -4.05 1 $\simeq 0.15$
  red +3.14 1! 0.42
  red +4.29 1! 0.17
1667 red all comp. 1 >0.10
  red +5.90 1 0.26
  red +7.20 1 0.26
1665 red +3.00 3 0.06, 0.15**
  red +3.80 3 <-0.24, -0.15**
  red +4.65 3 <-0.24, -0.15**

*: Same intervals as given in the Table 11.
!: Around the Julian day 45260 framing an OH minimum.
**: Respectively for the first and second cycle of the third interval.


  \begin{figure} \par\includegraphics[angle=-90,width=8.8cm,clip]{ds1680f21.ps} \end{figure} Figure 24: Integrated flux variation curves of R LMi in both main lines and circular polarizations from January 1982 to March 1983 and from August 1989 to November 1995

It is clear that at the date of the displayed observations a rather strong right-handed polarization, for the blue peak, was observed. Also the spectral profile, the range of emission as well as the maximum detected intensity are roughly similar for both main lines. In particular, each of the two peaks shows two well defined main groups of components centered about the same velocity.

4.4.2 Integrated flux

The integrated flux variations for the two epochs of observations and for both main lines are displayed Fig. 24. The consecutive cycles of the second interval of observations are labelled from (0) to (6). The phase delay between the OH and the optical curve is 52 $\pm$ 15 days (cf. Table 2). From the 6 consecutive cycles, one can see a slow decline of the mean integrated flux value followed by a slow increase for the maser emission of the two peaks at 1665, as well as at 1667 MHz. Clearly, the variation of the mean value of the integrated flux from one cycle to another is not a random process but undergoes a slow variation over several cycles. Moreover, the 1665 MHz emission indicates that this long-term variation is roughly the same for the front and back part of the shell. Nevertheless, the minimum of the red peak long-term variation was delayed of 2 cycles compared to the blue peak one. Unfortunately, due to undersampling at 1667 MHz at the end of the OH cycle (3) and during the OH cycle (4), we can not certify that this trend was exactly the same in this line, even though the behaviour from cycles (0) to (3) and from cycles (5) and (6) are identical for a specific peak in both main lines.

The values of $\Delta I_{\rm min-max}$ are given in Table 18. For the second set of observations, which covers 6 consecutive cycles, $\Delta I_{\rm min-max}$ ranged from 36% to 93% with a mean value of 66% for the 1667 MHz line. They were slightly greater at 1665 MHz, ranging from 39% to 95% with a mean value of 72.5%. Moreover, there is an increase in the amplitude of variation while the mean integrated flux value decreases (i.e., from cycle (1) to (2) in the case of the blue peak and from cycle (1) to (4) for the red peak).

From Fig. 24, we note that the behaviour concerning the polarization is similar for the two main lines. Thus, while an almost zero polarization can be observed in both main lines for the red peak, a rather strong right-handed polarization is observed for the blue peak in both main lines. The degree of polarization for the blue peak reaches values higher than 0.45 in both lines in the very first set of data (cf. Table 19).


  \begin{figure} \par\includegraphics[angle=-90,width=8.8cm,clip]{ds1680f22.ps} \end{figure} Figure 25: Intensity variation curves of the Gaussian fitted components of R LMi at 1665 MHz in left- (LHC) and right-handed (RHC) polarizations for the red and blue peaks for the two epochs of observations (i.e., from January 1982 to March 1983 and from August 1989 to November 1995)


 

 
Table 16: Central velocity and full width at half-maximum (FWHM) of the fitted components which have shown a longevity greater than 1 stellar period of R LMi at 1667 MHz for each interval of observations
a) Red peak
  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
interval 11 V1=6.01 0.527 V1=5.95 0.568
  V2=4.93 0.723 V2=4.98 0.754
  V3=4.15 0.721 V3=4.07 0.629
  V4=3.39 0.687 V4=3.41 0.641
  V5=2.59 0.705 V5=2.59 0.625
interval 22 V1=5.79 0.529 V1=5.87 0.496
  V2=4.95 0.729 V2=5.03 0.731
  V3=4.36 0.541    
  V4=3.80 0.713 V4=3.99 0.698
  V5=2.90 0.647 V5=3.01 0.720

b) Blue peak

  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
interval 11 V1=-2.37 0.833 V1=-2.45 0.726
  V2=-3.33 0.633 V2=-3.21 0.568
  V3=-4.13 0.825 V3=-4.06 0.651
  V4=-4.96 0.521 V4=-4.86 0.685
interval 22 V1=-1.83 0.737 V1=-1.82 0.871
  V2=-2.65 0.775 V2=-2.48 0.609
  V3=-3.17 0.653 V3=-3.13 0.714
  V4=-3.96 0.807 V4=-3.97 0.746

1: Interval of Julian days [45000 - 45400].
2: Interval of Julian days [47700 - 50050].


  a) Red peak

 
Table 17: Source: R LMi. The same as for the previous table but for the 1665 MHz
  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
Interval 11 V1=5.00 0.714 V1=5.03 0.812
  V2=4.14 0.567 V2=4.23 0.561
  V3=3.34 0.628 V3=3.55 0.934
  V4=2.18 0.608 V4=2.18 0.681
Interval 22 V1=5.02 0.657 V1=5.04 0.682
  V2=4.38 0.525    
  V3=3.70 0.659 V3=3.81 0.683
  V4=3.01 0.696 V4=2.94 0.691
  V5=2.30 0.824 V5=2.38 0.589

b) Blue peak

  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
Interval 11 V1=-2.23 0.770 V1=-2.32 0.715
  V2=-3.40 0.731 V2=-3.23 0.575
  V3=-4.11 0.708 V3=-4.04 0.517
Interval 22 V1=-1.73 1.031 V1=-2.20 0.752
      V2=-3.18 0.590
  V3=-3.82 0.813 V3=-4.03 0.583

1 and 2: Same intervals as given in the previous table.



 

 
Table 18: $\Delta I_{\rm min-max}$ for the integrated flux of R LMi
Freq. peak $\Delta I_{\rm min-max}$
(MHz)   interval interval
    1$^{\rm a}$ 2$^{\rm b}$
      (1) (2) (3) (4) (5) (6)
1667 blue 65% 74% 83% >63%   80% >59%
  red 38% 36% 48% 52%   73% >93%
1665 blue 57% 81% 87% 80% 63% 71% >63%
  red 41% 39% 49% 64% 93% 86% >95%

a: Interval of Julian days [45000 - 45400].
b: Interval of Julian days [47700 - 50050] covering 6 consecutive cycles labelled from (0) to (6) in Fig. 24.


 

 
Table 19: [ RHC-LHC] for the blue peak integrated flux of R LMi in both main lines
a) Interval of Julian days [45000 - 45400]
Frequency [RHC-LHC]
(MHz)  
1667 $+0.20\le$[RHC-LHC]$\le+0.46$
1665 $+0.50\le$[RHC-LHC]$\le+0.60$

b) Interval of Julian days [47700 - 50050] covering 6 consecutive cycles labelled from (0) to (6) in Fig. 24

Frequency [RHC-LHC]
(MHz)            
  (0) (1) (2) (3) (4) (5)
1667 +0.23 +0.29 +0.28     +0.10
1665 +0.44 +0.45 +0.45 +0.42 +0.45 +0.37



  \begin{figure} \par\includegraphics[angle=-90,width=8.8cm,clip]{ds1680f23.ps} \end{figure} Figure 26: Source: R LMi. The same as for the previous figure but at 1667 MHz

4.4.3 Spectral components

Figures 25 and 26 display the variations in intensity of the fitted red and blue peak components with the longest lifetime at 1665 and 1667 MHz respectively. From a comparison between the variability curves of the first (i.e., for the Julian days [ 45000 - 45400]) and the second (i.e., for the Julian days [ 47700 - 50050]) data sets it appears that the mean lifetime of the components can be greater than 5 cycles but less than 10 years. On average, in each peak, only 2 components have a longevity greater than 10 years at 1665 and at 1667 MHz. On the other hand, components with short lifetime can be rather numerous, more than 5 for a given peak can be observed during the very first cycle lying in the interval of Julian days [ 45000 - 45400].

Moreover, Figs. 25i and 26g,h and j clearly show that for this star, the degree of saturation differs from one component to another. Indeed, some of them exhibit large flux variations from one cycle to another while others undergo variations of the same order of magnitude.

Table 20 gives the range and mean values of $\Delta I_{\rm min-max}$ of the components of the blue and red peaks for both main lines observed over at least one cycle in the first set of observations.

 

 
Table 20: Range and mean value of the $\Delta I_{\rm min-max}$ for the components of great longevity of R LMi for the first set of observations (i.e., interval of Julian days [45000 - 45400])
Freq. peak $\Delta I_{\rm min-max}$
(MHz)   range mean
      value

1667

 blue 56 to 66% 59%
  red 37 to 69% 52%
1665 blue 38 to 70% 55%
  red 33 to 65% 45%


Table 21 gives the trend of $\Delta I_{\rm min-max}$ for those components with the greatest longevity over the 5 concecutive cycles of the second set of observations labelled from (1) to (5) in Fig. 24 for both the 1665 and 1667 MHz. Finally, in Table 22 are listed the components for which |[RHC-LHC]$\vert\ge 0.10$.


 

 
Table 21: Range and mean value of the $\Delta I_{\rm min-max}$ for the components of great longevity of R LMi for the five cycles labelled from (1) to (5) of the second set of observations (i.e., interval of Julian days [47700 - 50050])
Freq. peak V $_{\rm comp.}$ $\Delta I_{\rm min-max}$
(MHz)   (km s-1) cycle
      (1) (2) (3) (4) (5)
1667 blue all comp.1 71% 78% 77%   56%
  red all comp.2 43% 49% 48%   54%
1665 blue -2.20, -3.20, -4.05 80% $\searrow$ 55% 65%
  blue -1.75, -3.80 <60% $\searrow$ 45% 65%
  red +3.75 35% $\nearrow$ 73% $\searrow$ <44%
  red +5.00 33% 56% 50% 68% 58%

1: The real values are within a range of $\pm$ 10% from the given mean values for cycles (1) and (2) while within a range of $\pm$ 32% for cycle (3) and (5).
2: The real values are within a range of $\pm$ 15% from the given mean values.

   
4.5 RR Aql

This star was observed during two different epochs. The first set of observations was performed between January 1982 and March 1983. With its period of 394 days (cf. Tables 1 and 2), this interval covers one cycle. The second set of data was performed from July 1993 to April 1995 and covers a bit less than 2 cycles.

4.5.1 Spectra

For this star as for RS Vir, maser emission occurs between the red and blue peaks (cf. Fig. 6). According to the shape of the main line spectra, with rather blended components, only the small group centered at V=+29.00 km s-1 has been considered as an intermediate peak. In comparison with the 1612 MHz profile, the two adjacent groups of components in both main line profiles (i.e., at [+26.0; +28.0] km s-1and [+29.7; +31.0] km s-1) could be considered as belonging to the intermediate peak, but as they are usually blended with the blue and red peaks respectively, they have been treated as part of them.

4.5.2 Integrated flux

Figure 27 shows the integrated flux variations in the 3 OH maser lines in both circular polarizations for the two epochs of observations.
  \begin{figure} \par\includegraphics[angle=-90,width=8.8cm,clip]{ds1680f25.ps} \end{figure} Figure 27: Integrated flux variation curves of RR Aql in the 3 OH maser lines in both circular polarizations from January 1982 to March 1983 and from July 1993 to April 1995 for the blue, red and intermediate (Int.) peaks

The measured delay between the OH and optical curves was found to be $70 \pm 15$ days (cf. Table 2).


 

 
Table 22: |[RHC-LHC]$\vert\ge $ 0.10 for the components of great longevity of R LMi
Freq. peak $V_{\rm comp.}$ interval cycle [RHC-LHC]
(MHz)   (km s-1) * b  
1667 blue & red all comp. 1   from 0.24 to 0.42
  blue -3.95 2 from (1) to (5) 0.45 $\searrow$ 0.13
  blue -2.50 2 from (1) to (3) 0.34 $\searrow$ 0
  blue -2.50 2 (5) -0.28
  red +5.00 2 from (1) to (5) down to -0.14
  red +3.85 2 from (1) to (5) from 0.12 to 0.22
1665 blue -4.05 1   0.16
  blue -3.90 2 from (2) to (5) 0.66 $\searrow$ 0.54
  blue -3.30 1   0.59
  blue -2.30 1   0.70
  red +5.00 1   0.10
  red +3.45 1   0.19

*: Same intervals as given Table 18.
b: Labelled from (1) to (5) in the second interval (cf. Fig. 24).


 

 
Table 23: Central velocity and full width at half-maximum (FWHM) of the fitted components which have shown a longevity greater than 1 stellar period of RR Aql at 1612 MHz for the 2 intervals of observations. For comparison are given too in the last column the velocity of the components determined by Herman & Habing (1985, i.e., column labelled H&H)
a) Red peak
  LHC FWHM RHC FWHM H&H
  km s-1 km s-1 km s-1 km s-1 km s-1
Interval 11 V1=34.09 0.856 V1=34.10 0.806 34.14
  V2=32.64 0.974 V2=32.63 0.975 32.70
Interval 22 V1=34.21 0.738 V1=34.21 0.746  
  V2=32.73 0.961 V2=32.71 0.974  

b) Blue peak

  LHC FWHM RHC FWHM H&H
  km s-1 km s-1 km s-1 km s-1 km s-1
Interval 11 V1=23.92 0.739 V1=23.79 0.908  
  V2=22.81 1.066 V2=22.79 1.034 22.85
  V3=21.33 1.280 V3=21.34 1.155 21.18
Interval 22 V1=23.92 0.720/0.942(!) V1=24.03 0.524  
  V2=22.82 1.123 V2=22.86 1.113  
  V3=21.25 0.916 V3=21.30 1.039  

1: Interval of Julian days [45000 - 45400].
2: Interval of Julian days [49150 - 49800].
(!): The first mean value of the FWHM given here doesn't take into account the values of the FWHM guessed as unsatisfying ones (i.e., when they are clearly greater than the whole measured FWHM over the period of observations).


 

 
Table 24: Central velocity and full width at half-maximum (FWHM) of the fitted components which have shown a longevity greater than 1 stellar period of RR Aql at 1667 MHz for the 2 intervals of observations
a) Red and intermediate (int.) peaks
  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
Interval 11 V1=33.85 0.671 V1=33.89 0.738
  V2=32.84 0.649 V2=32.87 0.695
  V3=32.08 0.735 V3=32.11 0.644
  V4=31.45 0.710 V4=31.52 0.667
  V5=30.54 0.776 V5=30.57 0.893
  $V_{\rm int.}=29.02$ 0.925 $V_{\rm int.}=29.00$ 0.864
Interval 22 V1=34.07 0.458 V1=34.08 0.476
  V2=33.57 0.498 V2=33.54 0.549
  V3=32.92 0.561 V3=32.91 0.548
  V4=32.31 0.653 V4=32.27 0.729
  V5=31.57 0.775 V5=31.58 0.710
  V6=30.59 0.737 V6=30.61 0.739
  $V_{\rm int.}=29.02$ 0.915 $V_{\rm int.}=28.93$ 0.754

b) Blue peak

  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
Interval 11 V1=27.87 0.512 V1=27.73 0.501
  V2=27.18 0.636 V2=27.00 0.692
  V3=26.25 0.764 V3=26.19 0.792
  V4=25.19 0.732 V4=24.98 0.881
  V5=24.21 0.764 V5=24.11 0.681
  V6=23.43 0.952 V6=23.45 0.785
  V7=21.89 0.810 V7=21.96 0.833
Interval 22 V1=27.86 0.408 V1=27.62 0.444
  V2=27.20 0.628 V2=27.05 0.494
  V3=26.47 0.553 V3=26.44 0.468
  V4=25.71 0.593 V4=25.78 0.622
  V5=25.07 0.542 V5=24.95 0.602
  V6=24.21 0.802 V6=24.14 0.764
  V7=23.42 0.515 V7=23.42 0.504
  V8=22.98 0.678 V8=22.93 0.623
  V9=21.94 0.731 V9=22.00 0.790

1 and 2: The same intervals as given in the previous table.



 

 
Table 25: Source: RR Aql. The same as for the previous table but for the 1665 MHz
a) Red and intermediate (int.) peaks
  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
Interval 11 V1=33.48 0.678 V1=33.42 0.698
  V2=32.60 0.716 V2=32.45 0.784
  V3=31.87 0.839 V3=31.65 0.789
  V4=30.64 0.906 V4=30.57 0.841
  $V_{\rm int.}=29.15$ 0.778 $V_{\rm int.}=29.09$ 0.651
Interval 22 V1=33.57 0.689 V1=33.57 0.691
  V2=32.67 0.757 V2=32.55 0.883
  V3=31.85 0.902 V3=31.75 0.774
  V4=30.58 0.657 V4=30.69 0.834
  $V_{\rm int.}=29.01$ 0.622 $V_{\rm int.}=28.67$ 0.656

b) Blue peak

  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
Interval 11 V1=27.93 0.535    
  V2=27.22 0.580 V2=27.16 0.779
  V3=26.38 0.773 V3=26.36 0.736
  V4=25.28 0.737 V4=25.26 0.854
  V5=24.22 0.785 V5=24.15 0.910
Interval 22 V1=27.63 0.378    
  V2=27.12 0.453 V2=27.13 0.830
  V3=26.51 0.561 V3=26.09 0.850
  V4=25.41 0.936 V4=25.04 0.886
  V5=24.33 0.722 V5=24.16 0.963
  V6=23.63 0.814    

1 and 2: The same intervals as given in Table 23.



 

 
Table 26: $\Delta I_{\rm min-max}$ for the integrated flux of RR Aql
Frequency peak $\Delta I_{\rm min-max}$
(MHz)   interval interval
    11 22
      (1) (2)
1612 blue 33% 43%
  red 32% 37%
1667 blue 49% 30% 42%
  red 53% 37% 53%
  intermediate 75% 65%
1665 blue 60% 33%
  red 53% 36%
  intermediate 70% 51% 66%

1: Interval of Julian days [45000 - 45400].
2: Interval of Julian days [49150 - 49800] covering 2 cycles labelled (1) and (2).


The values of $\Delta I_{\rm min-max}$ for the 1612 MHz integrated flux are comparable for the two standard peaks and show a difference of less than 10% between the first and second set of observations. Main lines show greater values of $\Delta I_{\rm min-max}$ than the 1612 MHz line but difference between red and blue peak values for a given cycle are very small (<10%). On the other hand, the differences between the $\Delta I_{\rm min-max}$of the standard peak and the inter-peak integrated flux reaches more than 30% (cf. Table 26).


 

 
Table 27: Greatest [RHC-LHC] observed in the standard peak integrated flux of RR Aql in the three OH lines.
Frequency peak [RHC-LHC]
(MHz)   interval interval
    1* 2*
1612   +0.21 <+0.08
1667   +0.33 $\simeq$+0.0
1665 blue   $\simeq$-0.08
  red +0.17 <+0.10

*: The same intervals as given in the previous table.


As concerns its polarization, this star is quite similar to RS Vir in the sense that the 1612 and 1667 MHz emission behave similarly. Generally, temporal variations in the degree of polarization can be observed in the 3 maser lines. Thus, a right-handed polarization observed at 1612 and 1667 MHz for the red and blue peaks as well as for the red peak at 1665 MHz in the first data set is no longer observed in the second set (cf. Table 27). The intermediate peak, observed in the main lines, shows a very small degree of polarization only at 1665 MHz for the first cycle of the second set of observations.

4.5.3 Spectral components

Figures 28-31 display the intensity variations of the fitted spectral components in the 3 maser lines for the two sets of observations which have a lifetime of a least one stellar cycle.
  \begin{figure} \par\includegraphics[angle=-90,width=8.8cm,clip]{ds1680f26.ps} \end{figure} Figure 28: Intensity variation curves of the Gaussian fitted components of RR Aql at 1612 MHz in left- (LHC) and right-handed (RHC) polarizations for the two standard peaks for the two sets of observations (i.e., from January 1982 to March 1983 and July 1993 to April 1995)


  \begin{figure} \par\includegraphics[angle=-90,width=8.8cm,clip]{ds1680f27.ps} \end{figure} Figure 29: Intensity variation curves of the Gaussian fitted components of RR Aql at 1665 MHz in left- (LHC) and right-handed (RHC) polarizations for the two standard peaks as well as for the intermediate one (components centered in the velocity interval [+28.50; +29.50] km s-1), for the two sets of observations (i.e., from January 1982 to March 1983 and July 1993 to April 1995)


  \begin{figure} \par\includegraphics[angle=-90,width=8.8cm,clip]{ds1680f28.ps} \end{figure} Figure 30: Intensity variation curves of the Gaussian fitted components of RR Aql at 1667 MHz in left- (LHC) and right-handed (RHC) polarizations for the blue peak for the 2 sets of observations (i.e., from January 1982 to March 1983: Cols. 1 and 2 and from July 1993 to April 1995: Cols. 3 and 4)


  \begin{figure} \par\includegraphics[angle=-90,width=8.8cm,clip]{ds1680f29.ps} \end{figure} Figure 31: Source: RR Aql. The same as for the previous figure but for the red and intermediate peaks, this latest being centered at V=29.00 km s-1

The number of such components is less at 1612 MHz for all peaks (only 3 in the blue peak and 2 in the red one). On the other hand, both main line spectra display a highly complex profile (cf. Fig. 6) where 4 components with lifetimes greater than one stellar period in each of the standard peaks at 1665 MHz while 6 and 9 of them can be observed in the 1667 MHz red and blue peaks respectively.

Here again, all the components at 1612 MHz are very stable, they can be observed in both sets of data over more than 10 years. At 1667 MHz, only 3 of the 7 components observed in the red and intermediate peaks have a lifetime greater than 10 years. These components are located at a velocity of V=31.50, 30.55 and V=29.00 km s-1. Likewise, only 5 of the 9 components observed in the 1667 MHz blue peak can be seen over more than 10 years. Those are centered at V=27.80, 27.10, 24.15, 23.43 and 21.95 km s-1. At 1665 MHz, changes in the spectral components are of the same order: two components are observed without ambiguity over the two sets of observations in the blue peak (at V=24.15 and 27.16 km s-1) against 4 in the red peak (i.e., the 4 components displayed in Fig. 29). Table 28 gives the range and mean values of $\Delta I_{\rm min-max}$ in the three OH lines for the two intervals of observations.

 

 
Table 28: Range and mean value of $\Delta I_{\rm min-max}$ for the components of great longevity of RR Aql
Freq. Peak $\Delta I_{\rm min-max}$
(MHz)   Range Mean Value
    interval interval interval interval
    1* 2* 1 2
1612 blue 16 to 37% 40 to 47% 23% 44%
  red 21 to 30% 35 to 40% 25% 38%
1665 blue 38 $\nearrow$ 621 28 $\nearrow$ 35%2 51% 32%
  red 35 to 64% 30 $\nearrow$ 40%2 47% 35%
  inter.     33% 39%
1667 blue 33 to 73% 24 to 58% 56% 40%
  red 44 to 77% 34 to 60% 60% 50%
  inter.     74% 48%

*: The same intervals as given Table 26.
1: Rougthly increasing with the decrease of ${\vert V_{\rm component}-V_{\rm star}\vert}$ (i.e., $\Delta I_{\rm min-max} \nearrow$
from the component centered at V=24.20 km s-1 to the one located at V=27.20 km s-1).
2: Increasing with the increase of ${\vert V_{\rm component}-V_{\rm star}\vert}$.



 

 
Table 29: $\vert[{\rm RHC-LHC}] \vert \ge 0.10$ for the components of great longevity of RR Aql
Freq. peak V $_{\rm comp.}$ interval [RHC-LHC]
(MHz)   (km s-1) *  
1612 red 21.34 1 0.14
1667 blue all comp. 1 0.18
  blue all comp. 2 $\le$0.12
  red all comp. 1 0.15 $\nearrow$ 0.37$^{\rm a}$
  red 33.55 2 -0.22
  red 30.60 2 0.14
  inter. 29.00 1 0.19
  inter. 29.00 2$^{\rm b}$ -0.33
1665 blue 25.27 1 0.15
  blue 24.20 2 0.22
  red 30.60 1 0.21
  red 30.60 2 0.24

*: The same intervals as given Table 26.
a: Increasing with the decrease of ${\vert V_{\rm star}-V_{\rm component}\vert}$ (i.e. from the component located at V=+33.85 km s-1 to the one located at V=+30.55 km s-1).
b: In the first cycle of this set of data which is no longer observed in the next cycle.

At 1667 MHz, the difference in amplitude of variations for a given component of the blue peak from the first data set to the second is about 15 - 20% except for the component belonging to the most external part of the peak (i.e., V= +21.95 km s-1) which shows a difference in the amplitude of variations of more than 30% between the first and the second data set.

|[RHC-LHC]$\vert\ge $ 0.10 for the components of great longevity of RR Aql are given in Table 29. At 1612 MHz, both components of the red peak contribute equally to the observed right-handed polarization in the first set of observations. Nevertheless, the long-term components of the blue peak never reached a degree of polarization greater than 0.15 in the first data set (cf. Table 29) leading to the conclusion that the right-handed polarization observed in the integrated flux of this peak is mainly due to transient components.

At 1665 MHz, the greater part of the components show a weak right-handed polarization of less than 10%, except those given in Table 29. The inter-peak component of great longevity lying in the velocity interval [+28.5;+29.5] km s-1 shows a degree of polarization fainter than 0.10. In the first data set at 1667 MHz, the inter-peak component shows a degree of right-handed polarization of the same order as observed for the greater part of the blue peak components. On the other hand, it shows a strong degree of left-handed polarization ([RHC-LHC] =-0.33) in the first cycle of the second data set which totally disappeared in the next cycle.

   
4.6 U Her

For this star, we have more than 7 consecutive, well sampled cycles in both main lines. On average, monthly observations were performed in both circular polarizations at 1667 MHz and in right-handed polarization at 1665 MHz between July 1984 and March 1993. Sparse Observations in the left-handed polarization at 1665 MHz were performed from July 1984 to August 1991 then monthly to March 1993. This type I star shows 1612 MHz emission too but this emission will not be presented here because of its peculiarity; especially, eruptive emission could be observed from April 1984 to September 1989 as already reported by Etoka & Le Squeren (1997).

4.6.1 Spectra

Figure 7 displays the spectra of this source in both main lines and circular polarizations. One can see that the profiles and maximum intensity values are about the same in the two main lines. From the figure one can easily see that the spectral profiles are composed of many components. In both main lines, the blue peak is the widest, exhibiting 3 well defined groups of components, while the red peak is more compact. Finally, one can notice that rather strong left-handed polarization is only observed for the blue peaks at the period of the displayed observations.

4.6.2 Integrated flux

The variations of the integrated flux of U Her for the red and blue peaks in both main lines and circular polarizations are displayed in Fig. 32.
  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f31.ps} \end{figure} Figure 32: Integrated flux variation curves of U Her in both main lines and circular polarization from July 1984 to March 1993

The phase delay measured from the two latest, most sampled 1667 MHz cycles is about 75 days (cf. Table 2).

On the average, and considering both main lines, the integrated flux of the red peak is half that of the blue. $\Delta I_{\rm min-max}$ for this star shows the smallest values observed in the main-lines ranging from 18 to 43% with a mean value of 32% for the 1667 MHz and 12 to 40% with a mean value of 25% for the 1665 MHz (cf. Table 32). Note that, the red peak integrated flux systematically exhibits fainter $\Delta I_{\rm min-max}$ than the blue. Surprisingly, we notice that for this star, for a given peak the 1665 MHz integrated flux shows systematically fainter amplitudes of variations than the 1667 MHz one.


 

 
Table 30: Central velocity and full width at half-maximum (FWHM) of the fitted components which have shown a longevity greater than 1 stellar period of U Her at 1667 MHz from July 1984 to April 1993 (i.e., corresponding to [45900 - 49100] in Julian days)
a) Red peak
LHC FWHM RHC FWHM
km s-1 km s-1 km s-1 km s-1
V1=-8.42 0.606 V1=-8.44 0.589
V2=-9.25 0.629 V2=-9.29 0.656
V3=-9.83 0.526 V3=-9.72 0.522
V4=-10.42 0.556 V4=-10.37 0.675
V5=-10.89 0.674 V5=-10.91 0.571
V6=-11.53 0.643 V6=-11.37 0.735
V7=-12.54 0.588 V7=-12.63 0.611

b) Blue peak

LHC FWHM RHC FWHM
km s-1 km s-1 km s-1 km s-1
V1=-15.15 0.578 V1=-15.20 0.566
V2=-16.08 0.600 V2=-16.02 0.583
    V3=-16.34 0.661
V4=-16.73 0.609 V4=-16.65 0.527
V5=-16.98 0.495 V5=-16.91 0.585
V6=-17.34 0.517 V6=-17.35 0.546
V7=-17.60 0.575 V7=-17.71 0.534
V8=-17.88 0.496 V8=-17.99 0.543
V9=-18.40 0.643 V9=-18.70 0.721
V10=-18.96 0.585 V10=-19.02 0.603
V11=-19.38 0.587 V11=-19.41 0.593
V12=-19.62 0.592 V12=-19.73 0.621
V13=-20.11 0.773 V13=-20.11 0.717
V14=-20.33 0.766 V14=-20.37 0.738



 

 
Table 31: Source: U Her. The same as for the previous table but for the 1665 MHz
a) Red peak
LHC FWHM RHC FWHM
km s-1 km s-1 km s-1 km s-1
V1=-8.45 0.523 V1=-8.48 0.490
V2=-9.23 0.560 V2=-9.26 0.552
V3=-9.78 0.470 V3=-9.73 0.429
V4=-10.38 0.670 V4=-10.30 0.660
V5=-10.97 0.780 V5=-10.99 0.530
V6=-11.72 0.637 V6=-11.56 0.612
    V7=-12.75 0.543

b) Blue peak

LHC FWHM RHC FWHM
km s-1 km s-1 km s-1 km s-1
    V1=-14.75 0.447
V2=-15.11 0.642 V2=-15.25 0.596
V3=-16.08 0.675 V3=-16.09 0.632
V4=-16.57 0.609 V4=-16.54 0.587
V5=-17.25 0.752 V5=-17.08 0.721
V6=-17.86 0.532 V6=-17.86 0.603
V7=-18.19 0.655 V7=-18.10 0.560
V8=-18.58 0.668 V8=-18.66 0.696
V9=-18.92 0.620 V9=-18.92 0.655
V10=-19.43 0.614 V10=-19.61 0.608
V11=-20.24 0.694 V11=-20.35 0.685


Considering each peak separately, it is clear that in trend the long-term variations of the 1667 MHz emission is very similar to that of 1665 MHz. Then, taking into consideration the general trend of the OH integrated flux variation curves, we note a rather strong increase of $\Delta I_{\rm min-max}$ from cycle (5) for the blue peak and only from cycle (6) for the red peak. Furthermore, we note an increase of the mean integrated flux value (due to an increase of the minima and maxima values) along the 7.5 cycles displayed. Thus, as observed for the Mira R LMi, U Her main line emission also exhibits a long-term variation of the mean integrated flux value. Note that this OH long-term variation is not correlated with the optical light curve variations.

Nevertheless, a slight difference exists between the 1665 and 1667 MHz emission, especially in the degree of polarization. Few observations in the left-handed polarization at 1665 MHz were collected near the OH maximum of the cycles (0), (1), (3), (5) as well as for the entire cycle (7) and the beginning of cycle (8). They allow us to detect the presence of a faint left-handed polarization for both 1665 MHz standard peaks. The ranges of [RHC-LHC] determined at 1665 MHz for these cycles are given Table 32.

At 1667 MHz, except for cycle (3) which shows no polarization, the signal is always left-handed polarized. Nevertheless, until cycle (3) for the blue peak and cycle (5) for the red only a very faint degree of polarization was measured (cf. Table 32) after which the degree of polarization is greater. All these large values of the degree of polarization in this line occurred simultaneously with the start of the increase of the amplitude of variations. Thus, a correlation might exist between the variations in these quantities. Another interesting fact is that the left-handed polarized signal started to increase one cycle in advance of the right-handed.

4.6.3 Spectral components

Figures 33-36 display the intensity variations of the fitted components with the greatest lifetime. Because of the small quantity of data in the left-handed polarization, at 1665 MHz only the variations of the components in right-handed polarization are displayed.


  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f32.ps} \end{figure} Figure 33: Intensity variation curves of the Gaussian fitted components of U Her at 1665 MHz in right-handed (RHC) polarization for the red peak from July 1984 to March 1993


  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f33.ps} \end{figure} Figure 34: Source: U Her. The same as for the previous figure but for the blue peak


  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f34.ps} \end{figure} Figure 35: Intensity variation curves of the Gaussian fitted components of U Her at 1667 MHz in left- (LHC) and right-handed (RHC) polarizations for the red peak from July 1984 to March 1993


  \begin{figure} \par\includegraphics[width=8.8cm,clip]{ds1680f35.ps} \end{figure} Figure 36: Source: U Her. The same as for the previous figure but for the blue peak

Spectral decomposition by Gaussian fitting reveals a great difference in behaviour between the various components, especially concerning the changes in the $\Delta I_{\rm min-max}$ from one cycle to another.


 

 
Table 32: $\Delta I_{\rm min-max}$ and [RHC-LHC] for the integrated flux of U Her over the 7 consecutive cycles labelled from (1) to (7) of the interval of Julian days [45900 - 49100]
Freq. peak $\Delta I_{\rm min-max}$
(MHz)   cycles
    (1) (2) (3) (4) (5) (6) (7)
1667 blue 29% 36% 36% 35% 36% 43% 42%
  red 18% 25% 26% 21% 23% 38% 40%
1665 blue 18% 25% 26% 21% 23% 38% 40%
  red 12% 20% 25% 16% 20% 30% 33%

Freq. peak [RHC-LHC]
(MHz)   cycles
    (0) (1) (2) (3)
1667 blue <-0.10 <-0.10 <-0.10 0
  red <-0.10 <-0.10 <-0.10 <<-0.10
1665 blue [-0.13; 0] [-0.13; 0]   [-0.13; 0]
  red [-0.15; -0.12] [-0.15; -0.12]   [-0.15; -0.12]

Freq. peak [RHC-LHC]
(MHz)   cycles
    (4) (5) (6) (7) (8)
1667 blue -0.21 $\simeq -0.10$ $\simeq -0.10$ $\simeq -0.10$  
  red <-0.10 <-0.10 -0.19 -0.15  
1665 blue   [-0.13; 0]      
  red   [-0.15; -0.12]   -0.10 -0.10


The range and mean value of the $\Delta I_{\rm min-max}$ for the components of U Her over the seven cycles are given in Table 33. At 1665 MHz, the greatest values of the amplitude of variations are attained by the most external group of components in both peaks (i.e., components located at V= -8.48 km s-1 and V=-9.26 km s-1 for the red peak and components located at V=-19.61 km s-1 and V=-20.35 km s-1 for the blue peak). The smallest amplitude values are reached by the component located at V=-10.30 km s-1 for the red peak and by the most internal component located at V=-14.75 km s-1 for the blue peak.

At 1667 MHz, the mean values of $\Delta I_{\rm min-max}$ are quite similar to the 1665 MHz. In the red peak, the greatest amplitudes of variations was observed for the component centered at V=-9.83 km s-1 which belongs to the most external group of components. In the blue peak, the greatest amplitudes of variations was observed for the group of components ranging in the velocity interval [-16; -18] km s-1.

It is clear that the increase of the mean value of the integrated flux value is due only to some specific components while other different components contribute to the changes of the amplitude of variations observed between cycles. The red peak components centered at V=-10.30 and -10.99 km s-1 at 1665 MHz are a good illustration of this fact (cf. Fig. 33). Even though they display large intensities they hardly contribute to the increase of $\Delta I_{\rm min-max}$ for the integrated flux observed in cycle (7) of Fig. 32. This increase is mainly due to the components located at V=-9.26 and -8.48 km s-1 and, with less importance, components located at V=-11.56 and -12.75 km s-1. Thus, the internal group of components of the peak (i.e., located at V= -10.30 and -10.99 km s-1) displays a quite different and dissociated behaviour from the two external groups of components of the peak (i.e., respectively located at V=-11.56 and -12.75 km s-1 and V=-9.26and -8.48 km s-1).

The 1665 MHz blue peak is even more complicated since the increase of $\Delta I_{\rm min-max}$ in the integrated flux from one cycle to another is due to an enhancement of intensity from different components in each cycle (cf. Fig. 34): the increase of $\Delta I_{\rm min-max}$ observed in cycle (5) and (6) of Fig. 32 is mainly due to the component located at V=-19.61 km s-1, while in cycle (7) it is due to the components located at V=-17.08 and -17.86 km s-1. On the other hand, the increase of the mean value of the integrated flux is only due to the two components located respectively at V=-18.66 and -20.35 km s-1. The components centered at a velocity closest to the stellar velocity (i.e., at V=-15.25 and -14.75 km s-1), show a very erratic behaviour. Moreover, a faint but continuous decrease of the mean value of the intensity of the component centered at V=-16.09 km s-1 is seen. We note that this last component is located at a velocity which is within 0.4 km s-1 of the 1612 MHz eruptive peak studied by Etoka & Le Squeren (1997).

At 1667 MHz, the increase of the integrated flux $\Delta I_{\rm min-max}$ in the red peak Fig. 32 is also due to the most internal group of components and located at the same velocity as in the 1665 MHz emission, i.e., at V=-8.44 and -9.29 km s-1. The component centered at V=-9.83 km s-1 as well as the components located at V=-12.54 and -11.53 km s-1 also contribute to this amplitude increase. Here again, but with less strength, the components centered at V=-10.37 and -10.91 km s-1 show a slow but continuous increase of the mean value of their integrated flux. Thus, the 1667 MHz red peak emission shows roughly the same behaviour as the one at 1665 MHz.

Most of the 1667 MHz blue peak components show quite smooth long-term variations. Thus, main contributors to $\Delta I_{\rm min-max}$ observed in the integrated flux in Fig. 32 for this specific peak are components located at V=-19.70, -18.40 and -16.91 km s-1 in a continuous process from cycle (5) to cycle (7).

Finally, considering all the cycles, the 1667 MHz red peak components of great longevity show a degree of polarization smaller than 10% expect for the component located at V=-10.40 km s-1, for which a left-handed polarization was observed ([RHC-LHC] =-0.14 during cycle (4)) and for the component located at V=-12.60 km s-1 for which a right-handed polarization was observed during cycle (7) with [RHC-LHC] =0.19. On the other hand, most of the blue peak components of great longevity show a left-handed polarization. [RHC-LHC] =-0.40 was reached for the component located at V=-16.05 km s-1 during cycle (6) and (7). Right-handed polarization was observed for the component located at V=-19.70 km s-1 (i.e., belonging to the most external group of components) from cycle (5) to (7) reaching a value as high as 30% during cycle (6), while the very most external component belonging to the same group showed a left-handed polarization of about 15% from cycle (4) to (7).

   
4.6.4 Comparison with existing maps in both main lines

Sivagnanam et al. (1990) obtained a VLA map of U Her in both main lines in left-handed polarization in March 1987. They detected compact components spread in 10 spots. Taking into account their channel separation of 0.47 km s-1, the correspondence between their detected components and our spectral composition is excellent.

For four of the components detected by Sivagnanam et al. (1990) at 1665 MHz we have long term variations. These components, centered at V=-10.3, -11.2, -14.9 and -16.3 km s-1 in their maps, correspond, within less than 0.21 km s-1, to our spectral components centered at V=-10.30 km s-1 and V=-10.99 km s-1 in the red peak and at V=-14.75 km s-1 and V=-16.09 km s-1 in the blue peak (cf. Figs. 33 and 34). These components, belonging to seven distinctive regions (labelled 1-6 and 8 by the authors) show a smooth variation along cycles. At 1667 MHz, we have long term variations for four of the detected components located in their maps at V=-16.3, -16.8, -19.6 and -20.1 km s-1 (corresponding, within less than 0.25 km s-1, to our fitted components respectively centered at V=-16.08, -16.73, -19.62 and -20.33 km s-1, cf. Figs. 36). These components, covering three regions in their maps (labelled 5, 9 and 10), also show smooth variations within cycles.


  \begin{figure} \par\includegraphics[angle=-90,width=8.8cm,clip]{ds1680f37.ps} \end{figure} Figure 37: Integrated flux variation curves of UX Cyg at 1612 a,b,c and 1667 MHz d,e,f in both circular polarizations for the group of components located at [-13.5; -9.0] km s-1 (group I), from January 1982 to March 1983, from July 1986 to January 1989 and from February 1991 to November 1995


  \begin{figure} \par\includegraphics[angle=-90,width=8.8cm,clip]{ds1680f38.ps} \end{figure} Figure 38: Integrated flux variation curves of UX Cyg at 1612 a,b,c, 1667 d,e,f and 1665 MHz g,h,i in both circular polarizations for the group of components located at [-9.0; -5.0] km s-1 (group II), for the 3 same intervals of time as for the previous figure


  \begin{figure} \par\includegraphics[angle=-90,width=8.8cm,clip]{ds1680f39.ps} \end{figure} Figure 39: Source: UX Cyg. The same as for the previous figure but for the group of components located at [-5.0; 0.0] km s-1 (group III)


  \begin{figure} \par\includegraphics[angle=-90,width=8.8cm,clip]{ds1680f40.ps} \end{figure} Figure 40: Source: UX Cyg. The same as for the previous figure but for the group of components located at [0.0; +7.5] km s-1 (group IV) and the one located at [+10.0; +17.0] km s-1 (group V), this latter at 1612 MHz only

Chapman et al. (1994) carried out interferometric observations of U Her in both main lines in July 1984 with MERLIN. For the same field of view their distribution matches the distribution from Sivagnanam et al. (1990), showing that the stability of spectral components is accompanied by a stable, temporal maser distribution in the shell.

All these facts together enable us to conclude that, for a major part of the observed regions emitting in the main lines, there exist some stable components lasting at least 3 to 10 years. The longevity of these components, added to their location stability, is consistent with a small velocity gradient ( $\epsilon \simeq 0.3$) as estimated by Chapman et al. (1994).

   
4.7 UX Cyg

This type II Mira has the largest period of the sample: about 560 days (cf. Table 1). This star was observed during 3 different epochs. The first epoch, from January 1982 to March 1983 covers 4/5 of a cycle and contains an OH maximum. The second set of data was taken with a coarser sampling between July 1986 to January 1989 but it allows us to follow the general trend of almost two cycles. The most recent set was taken from February 1991 to November 1995 over about 3 cycles in the main lines and over 2.5 cycles in the 1612 MHz satellite line. During the first and third epochs, all the observations were performed in both circular polarizations. For the second set of data, only the 1667 MHz observations were performed in both circular polarizations. The 1612 MHz observations were recorded in left-handed polarization and those at 1665 MHz in right-handed polarization.

4.7.1 Spectra

Figure 8 displays spectra of this source in the three OH maser lines and both circular polarizations. From this figure it is clear that the spectral profile of this source is very peculiar since we can observe 5 groups of components at 1612 and 1667 MHz which spread over more than 26 km s-1. This implies a large expansion velocity (i.e., >10 km s-1), unusual for this kind of star. For simplicity, let us define group I as the group of components in the velocity range [-13.5; -9.0] km s-1, group II in the velocity range of [-9.0; -5.0] km s-1, group III in [-5.0; 0.0] km s-1, group IV in [0.0; +7.5] km s-1 and group V in [+10.0; +17.00] km s-1. At 1665 MHz only groups II to IV can be observed.

4.7.2 Integrated flux

The integrated flux variations of each of these groups of components in the 3 OH maser lines are displayed in Figs. 37-40 for the 3 epochs of observations. At 1667 MHz, emission of group V was usually below the threshold of detection, thus, only the integrated flux variations of groups I to IV are displayed here.

On average, emission in group V is by far the lowest and has shown the smallest amplitude changes from the first set of observations to the third one. It hardly reaches 0.6 K km s-1 during the third epoch of observations and, in fact, was hardly above the noise level for most of the time. Emissions from groups II and III in the main lines and from groups III and IV in the 1612 MHz satellite line are also very low in the third data set. Thus, the resulting variation curves are rather chaotic. Nevertheless, it is clear that each of these 5 groups of components undergoes cyclic variations. Thus, from the variations of the group IV during the third set of observations, a slow decline of the integrated flux of this group along the three displayed cycles is clearly seen. This is reminiscent of the slow long-term variation observed in the main-line integrated fluxes of R LMi and U Her. The phase delay, measured in the last cycle of the third data set in the strong 1612 MHz group I integrated flux, is about $80 \pm 10$ days (cf. Table 2).


 

 
Table 33: Range and mean value of the $\Delta I_{\rm min-max}$ considering the components of great longevity of U Her as a whole and over the seven cylces of the interval of Julian days [45900 - 49100]
Freq. peak $\Delta I_{\rm min-max}$
(MHz)   Range mean
      value
1667 blue 12 to 71% 38%
  red 17 to 65% 35%
1665 blue 20 to 82% 37%
  red 17 to 69% 32%



 

 
Table 34: Central velocity and full width at half-maximum (FWHM) of the fitted components which have shown a longevity greater than 1 stellar period of UX Cyg at 1612 MHz from February 1991 to November 1995 (i.e., [48300 - 50000] in Julian days)
  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
group I$^{\rm I}$ V1=-10.07 0.685 V1=-9.94 0.811
  V2=-10.52 0.974 V2=-10.31 0.748
  V3=-10.98 0.634    
  V4=-11.67 0.806 V4=-11.63 0.779
  V5=-12.22 0.540 V5=-12.18 0.637
group II$^{\rm II}$ V1=-5.51 0.648    
  V2=-8.23 0.817    
group III $^{\rm III}$ V1=-2.52 0.742 V1=-2.58 0.893
group IV$^{\rm IV}$ (*)   (*)  
group V$^{\rm V}$ (*)   (*)  

I: Velocity range [-13.5;-9.0] km s-1.
II: Velocity range [-9.0,-5.0] km s-1.
III: Velocity range [-5.0;0.0] km s-1.
IV: Velocity range [0.0;7.5] km s-1.
V : Velocity range [10.0;17.0] km s-1.
(*): No component showing a clear cyclic variation over more than 1 stellar period.


 

 
Table 35: Source: UX Cyg. The same as for the previous table but for the 1667 MHz
  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
group I$^{\rm I}$ V1=-11.86 0.679 V1=-11.83 0.753
  V2=-11.56 0.684 V2=-11.44 0.831
      V3=-10.68 0.877
group II$^{\rm II}$ (*)   (*)  
group III $^{\rm III}$     V1=-4.03 1.028
group IV$^{\rm IV}$ V1=2.31 0.784 V1=2.55 0.748
  V2=3.14 0.802 V2=3.08 0.877
  V3=3.57 0.792 V3=3.49 0.735
  V4=3.73 0.872 V4=3.94 0.710
  V5=4.79 0.843 V5=4.61 0.593
      V6=5.28 0.726
group V$^{\rm V}$ (*)   (*)  

Explanation of (*) and the velocity range of the so-called groups I, II, III, IV and V are given in the previous table.



 

 
Table 36: Source: UX Cyg. The same as for the previous table but for the 1665 MHz
  LHC FWHM RHC FWHM
  km s-1 km s-1 km s-1 km s-1
group I$^{\rm I}$ (*)   (*)  
group II$^{\rm II}$ (*)   (*)  
group III $^{\rm III}$ V1=-4.44 0.731 V1=-4.07 0.959
  V2=-1.46 0.541    
group IV$^{\rm IV}$ V1=1.54 0.704    
  V2=2.35 0.674 V2=2.46 0.645
  V3=3.24 0.555 V3=3.08 0.626
  V4=3.61 0.819 V4=3.50 0.686
  V5=4.04 0.572 V5=4.10 0.701
  V6=4.94 0.627    
  V7=5.36 0.828 V7=5.31 0.917
group V$^{\rm V}$ (*)   (*)  

Explanation of (*) and the velocity range of the so-called groups I, II, III, IV and V are given in Table 34.



 

 
Table 37: $\Delta I_{\rm min-max}$ for the integrated flux of UX Cyg
a) The greatest value of $\Delta I_{\rm min-max}$
Frequency Group interval* $\Delta I_{\rm min-max}$
(MHz)      
1612 II 1 92%
  IV 1 90%
1667 III 3 95%
  II 2 92%
1665 II 1 96%

b) The smallest value of $\Delta I_{\rm min-max}$

Frequency Group interval* $\Delta I_{\rm min-max}$
(MHz)      
1612 I 2 28%
  II 2 34%
1667 IV 2 42%
  II 3 45%
1665 IV 3 31%

*: Interval of Julian days 1: [45000 - 45400].
Interval of Julian days 2: [46600 - 47550].
Interval of Julian days 3: [48300 - 50000].
c) The difference between the greatest and the smallest value of $\Delta I_{\rm min-max}$ for each group and frequency.

Frequency Group
(MHz) I II III IV V
1612 52% 58% 45% 46% 41%
1667 28% 47% 38% 32%  
1665   58% 36% 36%  


Table 37 gives the smallest and the greatest value of $\Delta I_{\rm min-max}$ observed in each line as well as the difference between the smallest and greatest value of $\Delta I_{\rm min-max}$ observed in each group and at each frequency. For this star, taking into consideration all groups, the mean values of $\Delta I_{\rm min-max}$ are similar for the satellite line and the main lines: 65%, 71% and 65% at 1612, 1667 and 1665 MHz respectively. The difference between the smallest and the greatest value is quite large for the three maser lines, exeeding 30% for all lines. Surprisingly, it is the greatest at 1612 MHz, leading to the conclusion that this emission is unsaturated.

Some changes can be seen in the degree of polarization in the three maser lines. Values of |[RHC-LHC] $\vert\ge 0.10$ are listed in Table 38.


 

 
Table 38: |[RHC-LHC]$\vert\ge $0.10 for the integrated flux of the 5 groups of UX Cyg
Freq. Group interval [RHC-LHC]
(MHz)   *  
1612 III 1 -0.36
  IV 1 0.35
1667 IV 1 0.20
  I 2 0.40
  II 2 >0.40
  III 2 0.73
  IV 2 0.16
  IV 3 -0.22, $\leq -0.19$, $-0.43^{\diamond}$
1665 II 1 0.41
  III 1 0.59
  IV 1 $-0.12 \leq$ [RHC-LHC]< 0
    3 -0.29, -0.39, $-0.49^{\diamond}$

*: The same as given in the previous table.
$\diamond$: For the three cycles of this interval.


The 1612 MHz satellite line is the one which shows the least change. A strong degree of left-handed and right-handed polarization is observed for groups III and IV respectively during the first interval of observations. Otherwise the signal shows very faint polarization, especially in the very last interval of observations.

In the first interval of observations, the 1667 MHz line shows significant polarization only for group IV. During the second set of observations all groups show right-handed polarization, for which the highest values are given Table 38. Finally, during the third interval of observations, groups I, II and III show faint polarization (with the exception of the 100% right-handed polarization observed in the second cycle of this interval for group III). Only group IV shows a rather strong left-handed polarization, observable during all three cycles of this interval. At 1665 MHz, in the first set of observations, group IV exhibits a faint left-handed polarization while groups II and III show a strong right-handed polarization. For the third interval of observations, groups II and III also show similar behaviour, different from group IV. Thus, groups II and III exhibit a degree of polarization close to zero, while group IV shows, like at 1667 MHz, a strong left-handed polarization which increases from the first cycle to the third one.

Thus, in considering all the data sets, the 1665 and 1667 MHz lines exhibit a similar general trend in polarization quite different from the 1612 MHz satellite line. It should especially be noted that the behaviour of group IV during the third data set, which is identical for both main lines, shows an increase of the degree of polarization from the first to the third cycle, i.e., while the mean integrated flux value is decreasing.

4.7.3 Spectral components

Figure 41 displays the intensity variations of fitted components showing the greatest lifetime in the 3 OH maser lines and both circular polarizations for the third data set.

  \begin{figure} \par\includegraphics[angle=-90,width=8.8cm,clip]{ds1680f41.ps} \end{figure} Figure 41: Intensity variation curves of some Gaussian fitted components of UX Cyg in the 3 OH maser lines in left- (LHC) and right-handed (RHC) polarizations for the third set of observations (i.e., from February 1991 to November 1995). First column: 1612 MHz (i.e., boxes a,b,c,d), second column: 1667 MHz (i.e., boxes e,f,g,h) and third column: 1665 MHz (i.e., boxes i,j,k,l,m)

Because of the faintness of groups I, II and III in the main lines, the fitting was quite difficult to achieve. It is certainly the least satisfactory of the study, since some components "were lost'' in the noise or in the edge of nearby stronger components. As a result, the spectral decomposition for these groups in the main lines was of poor quality. Only the results for group IV are displayed in these lines. For the same reason, only the results for groups I and III are displayed at 1612 MHz.

One can see that the variations of all the displayed components follow the optical cycle according to the usual delay expected between OH and optical curves. The range and mean value of the $\Delta I_{\rm min-max}$ considering all the components of great longevity of UX Cyg in the interval of Julian days [48300 - 50000] are given Table 39.

 

 
Table 39: Range and mean value of the $\Delta I_{\rm min-max}$ considering all the components of great longevity of UX Cyg in the interval of Julian days [48300 - 50000]
Freq. $\Delta I_{\rm min-max}$
(MHz) Range mean
    value
1612 30 to 70% 52%
1667 20 to 68% 48%
1665 24 to 59% 42%


The components at 1612 MHz exhibit the grestest values of $\Delta I_{\rm min-max}$ while the smallest range and mean value are observed at 1665 MHz. It is clear from Fig. 41 that the variations of some of the components are rather chaotic and also for many of faint intensity.

Finally, at 1612 MHz none of the displayed fitted components exhibits a degree of polarization greater than 10%. On the other hand, at 1667 MHz, components centered at V=3.80 km s-1 and V=2.40 km s-1 belonging to group IV show left-handed polarization as strong as [RHC-LHC] =-0.27 while at 1665 MHz the components in the same range of velocity: at V=3.55 km s-1 and V=2.40 km s-1, also show a strong left-handed polarization, as strong as [RHC-LHC] =-0.34 and -0.62 respectively. At this frequency, the component located at V=4.05 km s-1 also shows a strong left-handed polarization with [RHC-LHC] =-0.35.

next previous
Up: Long-term OH variability of

Copyright The European Southern Observatory (ESO)