3. How useful is Binggeli and Jerjen's dataset?
Binggeli & Jerjen presented Sérsic profile parameters for 128 dwarf elliptical and dwarf lenticular galaxies, which they derived from the surface photometry of Binggeli & Cameron (1991 & 1993). The distribution of their objects on the sky is shown in Fig. 1, from which it is evident that only to the south-west of the cluster core direction (defined collectively by M84, M86 and M87) is there detailed coverage.
Turning now to Binggeli & Jerjen's total magnitudes, we were surprised to discover that the values they quoted were in fact those of Binggeli & Cameron rather than the systemic magnitudes (obtained by integrating Sérsic's law to ) that we would have expected. Binggeli & Cameron's photometric zero points were based on the total-magnitude scales of de Vaucouleurs & Pence (1979); Börngen (1980 & 1984 2) and the Virgo Cluster Catalog, hereunder VCC, of Binggeli et al. (1985). Ichikawa et al.'s (1986) total-magnitude scale was probably also used for the calibration of two or three plates, but Binggeli & Cameron were ambiguous on this point.
Although, Binggeli & Cameron were quite modest about the limitations of their photometry, Binggeli & Jerjen allowed for a photometric error of only 0.2 mag. in their correlation analyses. There are several reasons why the real photometric errors must be very much larger than this. These reasons are outlined below.
From comparisons with our independently calibrated Virgo Photometry Catalogue (Young & Currie, 1998) 3, hereunder VPC, it is clear that Binggeli & Cameron's adopted magnitude-scale standards do not define a single mutually consistent magnitude scale. This point is evident from Figs. 2, 3 and 4, in which serious scale discrepancies between the different sources are also noticeable 4.
We are confident that our VPC magnitude scales are not to blame for these scale discrepancies, as amongst other reasons, our photometry was calibrated with many hundreds of photoelectric aperture-photometry as well as simulated aperture-photometry measurements derived from CCD images. Furthermore, the agreement between VPC magnitude measurements and those of Durrell (1997), which were based on deep CCD photometry of relatively faint Virgo dwarfs, is better than 0.04 mag. We are also confident that the zero point of the VPC's B -band total magnitude () scale is accurate to several percent (note that it is independent of the general transformation we adopted to calibrate our plates with B and V -band photoelectric photometry measurements). For detailed comparisons with existing photoelectric photometry for VPC galaxies, see Sect. 8 of Young & Currie (1998).
Another [albeit related] reason why Binggeli & Jerjen must have severely under-estimated their photometric errors is that Binggeli & Cameron did not calibrate their photometry directly. Instead, they first calibrated their extrapolated total-magnitude scale with existing total-magnitude scales. The other Sérsic profile parameters derived by Binggeli & Jerjen were presumably derived on the basis of these magnitude-scale calibrations. The obvious weakness in this approach is that even if the different sources of standard objects had been accurately calibrated, there would be systematic differences between them on account of the different extrapolation (or in the cases of Börngen's and the VCC datasets, visual total-magnitude estimation) procedures.
A further problem is likely to be the scarcity of calibrators in certain fields. Binggeli & Cameron's Plate 1 was for example only calibrated with one galaxy.
If one compares Figs. 5 with Figs. 2, 3 and 4, it is clear that Binggeli & Cameron's magnitude-scale does not, as a whole, bear much resemblance to any of the scales invoked for calibration purposes. Binggeli & Jerjen must therefore have severely under-estimated the errors in both their total magnitude values and their central surface brightness, , measurements.
Young (1994 & 1997) has already presented some preliminary findings on the sizes of and origins of systematic errors in the faint ends of existing magnitude scales for Virgo galaxies. A much more detailed paper on this subject, covering the whole magnitude scale and dealing with the ramifications of the zero-point and scale errors uncovered, will be presented by Young et al. (in preparation).
In spite of the calibration problems described above, the Sérsic scalelength, , and the shape parameter, n, should be independent of zero point, so we would expect Binggeli & Jerjen's measurements of these quantities to be useful. Binggeli & Jerjen noted `quite good' agreement with Durrell's n values, with a rms (1 ) scatter of 0.10. Note that one should not be alarmed by the much larger scatter between these authors' log values, because this quantity is a strong function of n, assuming that the R -n correlation is genuine (which even Binggeli & Jerjen don't question-though they believe that it has a large intrinsic scatter).
In summary then, Binggeli & Jerjen's galaxy sample is not a complete sample of galaxies down to a well-defined total magnitude limit. Also, its coverage of the Virgo Cluster direction is very patchy. However, it does contain a large number of dwarf galaxies and is therefore useful on the basis of its size. Unfortunately the photometric zero points adopted for different plates are not mutually consistent, thereby rendering the and values of little use. However, this should not affect the n or values, which are probably more useful than the YC95 values because they are based on higher-resolution photometry.
© European Southern Observatory (ESO) 1998
Online publication: April 28, 1998