3 Comparison with previous measurements

In this section, we compare the present VI CCD data with the available photographic and CCD data. The difference ($\Delta$) is always in the sense present minus others.

3.1 Photographic data

BV photographic observations were published by Hartwick (1975) up to $V\sim$ 18 mag with a zeropoint accuracy of $\sim$ 0.1 to 0.2 mag in V. Statistical results of the photographic V magnitude comparison are given in Table 3. The differences are a function of brightness as also pointed out earlier by Ortolani et al. (1990). However, they are within the uncertainty present in the photographic measurements.

3.2 CCD data

Ortolani et al. (1990) have carried out BVRI CCD photometry of the cluster region using a detector whose sensitivity varied across the chip. The accuracy limits of their photometric calibration range from 0.07 to 0.15 mag. A comparison of their V and (V-I) data with ours could be carried out as S. Ortolani kindly provided us with a computer-readable table of their photometric results. The transformation equations relating their ($X_{\rm Or}$, $Y_{\rm Or}$) coordinate system to ours ($X_{\rm pres}$, $Y_{\rm pres}$) are found to be,

$$X_{\rm Or} = -268.503 + 1.431 X_{\rm pres} - 0.007 Y_{\rm pres}$$

$$Y_{\rm Or} = -379.783 + 0.007 X_{\rm pres} + 1.431 Y_{\rm pres}.$$

There are 1150 stars in the Ortolani et al. (1990) data whose positions coincide within 1 pixel with the stars measured by us. The differences between the data sets are plotted in Fig. 2 and their statistical results are given in Table 3. A photometric comparison of the stars common between field and cluster regions is also shown in the figure indicating excellent agreement between the independent VI photometric data of the two regions.

  

Table 3: Statistical results of the difference $\Delta$ in the sense present minus Ortolani et al. (1990) CCD and Hartwick (1975) photographic values. V and (V-I) are from the present photometry. The mean and standard deviations ($\sigma$) are based on N stars. A few points discrepant by more than 3$\sigma$ have been excluded

  

Figure 2: Comparison of the present photometry with Ortolani et al. (1990) CCD data has been shown by filled circles. The differences are in the sense present minus their data, plotted against the present CCD photometry. Crosses denote the photometric comparison of the field region with the cluster region. The differences are in the sense cluster minus field data

  

Figure 3: The V, (V-I) and I, (V-I) diagram for all stars observed by us. The (V-I) scale is compressed to show the presence of very red stars. Only for the brighter stars $(V \le 18)$, the cluster population can be distinguished from the rich population of field stars

However, a comparison of present photometry with Ortolani et al. (1990) indicates the following. Except for a few outliers, which appear to be mostly those that were treated as single in our measurements and as blended doubles in theirs, the distribution of the photometric differences seems fairly random with a constant zeropoint offset of $\sim$ 0.3 mag in V and of $\sim$ 0.2 mag in (V-I). However, differences in colour (V-I) are smaller for bluer objects and increase slightly for redder ones (see Table 3). A least-squares linear regression between the data points yields

$$\Delta(V-I)=0.06(\pm 0.009)(V-I)_{\rm ccd}+0.08(\pm 0.02)$$

with a correlation coefficient of 0.2. Obviously, there is a zeropoint difference in both V and I between the two sets of ground based CCD data. This is due to the difficulties encountered by Ortolani et al. (1990) in calibrating the data (see Guarnieriet al. 1998). The zeropoint offsets between HST data and ground based CCD data by Ortolani et al. (1990) found by Guarnieri et al. (1998) are similar to ours. This and the discussions in the sections to follow indicate that our photometry agrees fairly well with the HST data.

Now we turn to the results derived from the present VI CCD photometry.