4 The astrometric results

Figure 2 shows the distribution of the deviation from the mean value of the angular separation $\rho$ for each double star exposure. The standard deviation of this distribution is $\sigma_{\rho}$= 0.046 pixels, equivalent to $\sigma_{\rho}$ = 0.017 arcseconds

Figure 3 shows the distribution of the deviation from the mean value of the instrumental position angle $\theta$ for each double star exposure. The standard deviation of this distribution is $\sigma_{\theta}$= 0.2 degrees. As expected, with decreasing angular separation, the accuracy of the position angle determination diminishes as well.

  

Figure 2: Accuracy of angular separation determination

  

Figure 3: Accuracy of position angle determination

  

Figure 4: Two colour diagram

  

Table 1: Photometric results

We used the four IDS double stars with INDEX codes 01109S5610, 05495N0701, 06386S2221, 23579N0743 in order to determine the CCD scale. Their relative positions are very accurate and listed in Brosche & Sinachopoulos (1988, 1989). We found the scale to be:

$$01109{\rm S}5610: \hspace{.3cm} .37695 \pm .00085$$

$$05495{\rm N}0701: \hspace{.3cm} .37676 \pm .00073$$

$$06386{\rm S}2221: \hspace{.3cm} .37902 \pm .00093$$

$$23579{\rm N}0743: \hspace{.3cm} .37668 \pm .00090$$

arcseconds per pixel.

These values, that are statistically the same at the 95% level of significance, result to the unweighted mean value of the CCD scale of 0.37735 $\pm$ 0.00056 arcseconds/pixel.

The output catalogue of the HIPPARCOS satellite (ESA, 1997) has one entry for every component of the double stars IDS23579N0743 (HIC 247, 248) and IDS06386S2221 (HIC 32144, 32145) from the above list. We tried to use the HIPPARCOS results of these two pairs in order to to determine the CCD scale more accurately. We first confirmed that both pairs are common proper motion ones and we remarked that their components have the same parallax on a very high statistical level of significance. So, they seem to have components of common origin. As concerns the scale, we obtain 0.37701 $\pm$ 0.00003 arcseconds/pixel for IDS23579N0743 and 0.38855 $\pm$ 0.00053 arcseconds/pixel for IDS06386S2221 . This last value is unexpectedly high. It might be caused by an overestimation of the accuracy of the HIPPARCOS results for this particular pair of stars (HIC 32144, 32145). This seems to be true for the secondary (HIC 32145) at least, because its entry in the Hipparcos catalogue gives an accuracy that is one order of magnitude lower than expected from the Hipparcos mission. We decided to use the scale calculated from the Brosche and Sinachopoulos catalogues.

  

Table 2: Astrometric results (epoch: 1991.91)

Finally, from several traces of equatorial stars in two different nights we found that the instrumental position angle calculated on the chip was 1.3 $\pm$ 0.1 degrees smaller than the correct one.
We present our astrometric measurements in Table 2. Average epoch of the observations is 1991.91. The first two columns of this table contain again equatorial coordinates (J2000) of the primaries. Column three the primary's Durchmusterung code. Column four the angular separation and its accuracy, column five the position angle and its accuracy. Column six contains the number of exposures made in V filter.