The Southern Galaxy Catalogue (SGC)
by Harold G. Corwin, Jr., Antoinette de Vaucouleurs, and Gerard de Vaucouleurs

I. Introduction

The "Southern Galaxy Catalogue" (SGC) is based on IIIa-J plates (and film 
copies) taken with the 1.2-m Siding Spring (Australia) Schmidt Telescope.  The
original SGC includes classifications, diameters, and remarks estimated by eye
for about 5500 galaxies (5513 in this release) generally larger than D_25 
~ 2.0 arcminutes, including all those in RC2 and NGC at declinations less 
than about -17 degrees (the northern limit of the -20 degree zone of the
southern sky survey plates).

This version of SGC (2.0) features improved positions (at equinox J2000.0), 
diameters and axis ratios reduced to the RC3 systems at B = 25.0 magnitudes
arcsec^-2^, position angles, V magnitudes, B-V color indices, and redshifts 
(given as V = cz), taken from various sources, primarily 6dF (positions and 
redshifts) and ESO-LV (position angles and photometry).  It has been assembled 
primarily for use by amateur astronomers, but may find other uses as well.

SGC 2.0 consists of two main files:  1) "sgc20.dat" with diameters and axis 
ratios given as base 10 logarithms and 2) "sgc20am.dat" with major and minor 
diameters converted to arcminutes.  Auxillary files include a brief 
introductory file, a file showing the format of the data files, and this file 
of documentation.

The Notes are not yet cross-referenced into the J2000 SGC lists (the 
cross-referenced Notes will be available later), but the Notes for the 
previous version (1.4.3) remain available as does the previous version itself.

The data files, most recently assembled on a Macintosh computer, use only 
7-bit ASCII characters, and have UNIX line endings.

II. Catalogue Format

The data share a format similar to that of the SEGC (which see), but some
differences occur.  There is one line per galaxy as follows for "sgc20.dat":

<  RA  J2000   Dec >  <ESO num>< ><N/I/A/D     >  < Full Type        ><Lum >f n lD  fl        lR fl       n   PA fl       n  Vmag   B-V    S      cz     S
<> <> <--> <-> <> <>  <-------><-><------------>  <------------------><---->| | <--><> <->   <--><> <->   |   <-><>   <>  |  <--->| <--->| <-->  <---->| <-->

Here are a few examples:

00 00 22.1 -80 47 34  012- G012                   (R')SAB(rs)bc pec   I-II    1 1.28         0.10         1    22            12.96    .95  ELV     7900  6dF
00 00 26.6 -18 50 31  538- G015 = N7807           SB(rs)ab?                   2 0.97         0.25         2    75            14.59    .94  ELV     7655  6dF
00 00 29.5 -40 29 04  293- G027                   SB(r)bc:            III     2 1.22         0.44         2   139?           13.62    .74  ELV     3180  6dF
00 00 53.3 -47 21 25  193- G009                   SB(r)0+? sp                 2 1.08         0.62         2   177            14.13    .99  ELV     5903  6dF
00 00 55.5 -40 42 45  293- G029                   SB(rs)c:            I-II  : 1 1.04         0.33         1     7:           14.17    .72  ELV    14677  6dF
00 01 02.9 -43 19 50  241- G010                   E+2                         2 1.24:        0.14:        2   145?           12.86   1.19  ELV    11627  6dF
00 09 56.4 -24 57 49  472- G016 = N0024           Sbc:                II-III: 1 1.73:        0.59:        1    45            11.61    .58  RC3      545  6dF
00 09 59.3 -57 01 15  149- G019 = N0025           SB0-? pec                   2 1.14         0.25         2    86            12.89   1.07  ELV     9457  6dF
00 13 21.0 -32 54 08  350                         E pec                       1 0.41:        0.18:        1    90:                                           
00 13 23.3 -32 54 12  350                         E pec                       1 0.71:        0.30:        1    96            16.10   1.27  6dF    30911  6dF 

The columns are:

1) Position:  RA in columns 1,2, 4,5, 7-10; and Dec in columns 12-14, 16,17,
   and 19,20.

Positions for equinox J2000.0 from 6dF; if there is no 6dF entry, the 
positions are picked off DSS2 images (via HEASARC's SkyView), or taken from my 
NGC/IC files.  The positions are presented in a single sexagesimal format:

  hh mm ss.d -dd mm ss

These positions are usually accurate on the 1 arcsecond level, but I have 
found a few 6dF positions that are off the apparent nucleus or center of the 
galaxy.  I'm replacing these as I find them.

The very large galaxies with no obvious nucleus (e.g. Sculptor and Fornax, NGC
55) have overly-precise positions for ease of reading the electronic data 
file.  Their estimated positions would be better represented by a format 
"hh mm ss -dd mm.d".  The positions given for all galaxies are always accurate 
enough to unambiguously identify the objects to which they refer.

2) Name in columns 23-50 (note difference from SEGC)

The ESO number taken from SGC (corrected if needed) in columns 23-33; if the 
galaxy does not appear in ESO, then only a southern field number is given 
here.  If another common name applies to the galaxy, it follows in columns
34-50 with an equal sign separating it from the ESO number.  These are usually
NGC and IC numbers, or RC2 "A"[nonymous], DDO, or (occasionally) MCG or other
numbers/names.  MCG numbers in particular may be found using ESO-B or NED.

3) Full VRHS type, expanded from the RC2/3 coded type given in previous 
versions of SGC in columns 51-70.  

Note that colons and queries are tacked onto the end of the type, NOT onto the 
questioned feature, unless I had the original SGC workbook in front of me.  I 
will fix this in a future release.  Luminosity class and uncertainty flag for 
spirals with Hubble stages "Sab" to "Im" appear in columns 71-77.  The LC is 
expanded from the numerical class given in previous versions.  The uncertainty 
flag is from the SGC weight:  w = 0.25 is "?", w = 0.26 to 0.75 is ":", 
w > 0.76 or blank is no flag.

For SGC, I used the modification to the RC2 classification scheme with the 
"cD" galaxies inserted at stage T = -4.  This was a mistake, confusing the 
morphology of the galaxy images with the astrophysics of their dominance of 
clusters of galaxies.  I have started replacing these "types" with the 
appropriate VRHS morphological types.  A future release will have all of these 
corrected.

Here are two paragraphs from the SEGC introduction that are also appropriate
here:

Morphological types are on the de Vaucouleurs revised Hubble-Sandage (VRHS)
system as presented in "The Reference Catalogue(s) of Bright Galaxies" (RC1, 
G. and A.  de Vaucouleurs, 1964; RC2, G. and A. de Vaucouleurs and H. Corwin, 
1976; and RC3, G. and A. de Vaucouleurs, H. Corwin, R. Buta, G. Paturel, and 
P. Fouque, 1991; and in G. de Vaucouleurs, ApJS 8, 31, 1963).  Also see "The 
de Vaucouleurs Atlas of Galaxies"; R. Buta, H. Corwin, and S. Odewahn, 
Cambridge University Press, 2007, for an updated, extended, and illustrated 
introduction to the VRHS types.

Luminosity classes in van den Bergh's system have, where possible, been
assigned for spirals and irregulars later than Hubble stage "ab" (T = +2). 
I have extended the system to luminosity classes V-VI and VI (based on surface 
brightness estimates).  I believe it is unlikely that these additional 
luminosity classes are correlated with fainter absolute magnitudes, but this 
has yet, to my knowledge, to be rigorously tested.

In general, the classifications I present here are compatible with those in 
SEGC and in the Reference Catalogues.  Note again, however, that the 
classifications for the "cD" galaxies that appear in RC2/3 at T = -4 are not
morphological classifications.  Most of these large galaxies would receive 
morphological classifications in one or another of the S0 bins, though some
of course, will in fact be classified as true transitional types, E/S0, as de 
Vaucouleurs did for RC1.

The number of type estimates follows the LC and its uncertainty symbol (if
any).

4) Diameter in columns 81-84 with uncertainty flags in columns 85 and 86.

Logarithm of the major diameter in units of 0.1 arcminutes, or major diameters 
in arcminutes as explained above.  If the diameter was measured on more than 
one plate, the number given is an unweighted mean.  Occasionally, diameter 
measurements made on DSS images are given for galaxies in close pairs or 
multiplets.  These assume symmetry of the galaxy's image, and are always 
followed by the uncertainty symbol "?".  

From the SEGC introduction:  Standard uncertainty symbols ":", "::", and "?" 
also often appear in columns 85,86 (and columns 98,99 for the axis ratios) 
when warranted by irregularities in the image.  Common problems leading to 
uncertain diameters are the presence of irregular arms, plumes, and bridges; 
interacting companion galaxies; background or foreground galaxies; images of 
Galactic stars; uneven sky background on the plates (often real -- due to 
faint, diffuse nebulosity -- but sometimes due to large-scale plate defects); 
and, particularly for the early-type galaxies, uncertainty as to the maximum 
visible extent of the galaxy.

For a few entries, the measurement applies to more than one galaxy, or it 
includes unusual extended features (e.g. "plumes", "jets", or "bridges").  In
these cases, the diameter, axis ratio, and position angle are enclosed in
square brackets "[" in column 80 and "]" in column 105.  As noted above, I 
have made estimates of the diameters of the galaxies in these multiple systems
where possible, but have retained the entry for the multiple where data apply
to it rather than to just the included galaxies.

To maintain the similarity to the latest version of SEGC, I have left columns
88-90 blank; eventually, standard deviations for the diameters may be inserted 
here.

5) The axis ratio (R = D/d) or minor diameter is in columns 94-97, and the 
   standard deviation of logR may appear later in columns 101-103:

The log of the axis ratio log R = log (D/d) is reduced to 25.0 magnitudes 
sec^-2^.  The uncertainty flag is from the SGC weight; see the luminosity 
class uncertainty flag as explained above.  

6) n(DR), column 107:

The number of diameter measures is usually the same as the number of type 
estimates (above), but not always.  A blank here always means "1".

7) Position angle in columns 111-113 with uncertainty flags in column 114; 
   columns 119 and 120 are reserved for the standard deviation.

The position angle, measured from north through east, is taken from ESO-LV,
ESO-B, or is estimated to the nearest five degrees by me.  I have added
uncertainty flags where needed.  The PAs from ESO-LV (from photographic 
surface photometry) can be incorrect; NGC 7807 is an example where PA_ESO-LV 
= 75.7 degrees is wrong.  The correct value (from SEGC) is 27 degrees.  I will 
fix these as I further debug SGC.

8) V magnitude in columns 126-130, and color band flag (if not V) in column 
   131.

The V-band magnitude from RC3; or transformed from ESO-LV B and R, or from 6dF
B_J and R_F, to the RC3 system; in that order of preference.  In practice, 
this meant that the photometry from ESO-LV was usually adopted as it is
available for all but a small percentage of the SGC galaxies, with RC3 being 
given if it was available.  Data from 6dF or other sources was adopted only if 
no RC3 or ESO-LV data existed.

The transformation equations are those adopted for SEGC; comments follow.

ESO-LV:

  Buta and Crocker's (1992, AJ 103, 1804) updated RC3 transformation:

  B(BC) = 0.98*B(ESO) + 0.245, +- 0.27
  R(BC) = 0.905*R(ESO) + 1.02, +- 0.21

  and from the canonical B-R = 1.5(B-V) + 0.10 (see e.g. ESO-LV, page 15):

  (B-V) = 0.67*[B(ESO) - R(ESO)] - 0.07, +- ~0.2

6dF:

  Transformations (impartial relations) newly derived by me with SGC and SEGC 
  data:

  mu <= 14.7:
    V(RC3)-13.5 = 0.81*[Bj(6dF)-13.5] - 0.18*[Rf(6dF)-13.5] - ...
                  ... - 0.09*[mu(6dF)-15.0] + 12.66, +- 0.35
  mu > 14.7:
    V(RC3)-13.5 = 0.67*[Bj(6dF)-13.5] - 0.30*[Rf(6dF)-13.5] - ...
                  ... - 0.82*[mu(6dF)-15.0] + 12.75, +- 0.50

  where

  mu = Bj(6dF) + 5.0*logD_25(SEGC) - 2.5*logR_25(SEGC) - 5.26

  The (B-V) color index transformation is

  (B-V)(RC3) = 1.06*[Bj(6dF) - Rf(6dF)] - 0.38, +- 0.20

The "canonical" relationship between B-V and ESO's B-R was derived from 
stellar photometry, not photometry of galaxies.  Not only have I used it here, 
but ESO-LV used it to convert B and V galaxy photometry to B and R photometry 
for calibration of their plates.  Because the spectral energy distributions 
for galaxies and stars are different, use of a transformation based on stars 
for galaxies is an approximation and will lead to systematic errors, 
especially for emission-line objects.  This will affect the ESO-LV B and R 
photometry transformed back to Johnson B and V.  This may be responsible for 
the somewhat redder color indices listed for the galaxies here where the 
source is "ELV".  For future versions of SGC, I shall rederive the 
transformations using only galaxy photometry.

Note, too, that galaxies with low surface brightnesses (mu >~ 14.5) have a 
strong dependence on surface brightness itself in their 6dF magnitudes.
There is, however, no significant dependence on color, and only a slight
dependence on galaxy type.  The latter is absorbed by the surface brightness
term in the above transformations as later types tend to have lower surface
brightnesses.

Formal errors in the transformed magnitudes are given above for 6dF.  For 
magnitudes transformed from ESO-LV, "real-life" errors are estimated to be
+-0.30 to +-0.35 magnitudes (see the RC3 introduction, p. 42, and Buta and
Crocker, 1992).

If the listed magnitude is not a V-band magnitude, the nominal band is listed
in column 131.  Currently, this is limited to just "B" for Johnson B, or "b"
for B(j) measured on IIIa-J plates.

9) B-V color index in columns 133-137, B-V uncertainty flag in column 138, and
   magnitude and color index source in columns 140-143:

B-V is taken from the same source as the V magnitude.  Transformations to the
RC3 system are via these equations:

6dF to RC3:

  (B-V)(RC3) = 1.06*[Bj(6dF) - Rf(6dF)] - 0.38

ESO-LV to RC3:

  (B-V)(RC3) = 0.67*[B(ESO) - R(ESO)] - 0.07

As noted above, the ESO-LV transformation was derived from stellar photometry.

Many color indices from these transformations are obviously discordant from 
the colors expected for the galaxy types.  I shall replace these as new data 
become available.  In the meantime, I have flagged the discordant colors in 
column 138, whatever their source, as follows:

      Color index      Flag
         B-V < -0.2     d
  -0.2 < B-V <= 0.0     ?
   0.0 < B-V <= 0.2     :
   0.2 < B-V <= 1.2     [None; but I nevertheless advise caution!]
   1.2 < B-V <= 1.3     :
   1.3 < B-V <= 1.4     ?
   1.4 < B-V            d

Note that the highly-reddened Circinus Galaxy (A1409-65) has an observed B-V 
of 1.46; it would be rejected by the above criterion.  Other low-latitude 
galaxies are similarly affected by interstellar extinction.

Sources of the magnitudes and colors are (in order of preference):

  RC3 = Third Reference Catalogue (deVaucouleurs et al, Springer (New York),
        1991).
  ELV = ESO/LV (A. Lauberts and E. Valentijn, ESO (Graching), 1989).
  6dF = 6dF DR3, Jones et al, MNRAS 399, 683, 2009 and references therein.
  HIPA = HIPASS, Doyle et al, MNRAS 361, 34 2005.
  Smbd = SIMBAD online database at http://simbad.u-strasbg.fr/simbad/ .

For galaxies without colors, approximate *unreddened* colors may be estimated
from the morphological type:

     Type  B-V    Type  B-V    Type  B-V    Type  B-V    Type  B-V
      E    1.0    S0/a  1.0     Sb   0.8     Sd   0.6     Im   0.6
      S0   1.0     Sa   0.9     Sc   0.7     Sm   0.6    Other 1.0

These are obviously very uncertain, but may give you an idea of the visual
magnitude if the galaxy has at least a B (or "b") magnitude listed.

10) Redshift in columns 147-151, and redshift source in columns 153-156:

The recessional velocities from different sources show no clear systematic
differences on the +-50 km sec^-1^ level, so they are presented just as they 
have been published.  Some of the published redshifts, however, were obviously 
discordant.  All of these were checked against NED's lists for each galaxy, 
and the most probable redshift chosen.  The redshift source (6dF, RC3, ELV, 
HIPA, NED, etc. generally in that order of preference) is in columns 152-155.

Note that 6dF collected redshifts from 2dF, ZCAT (CfA), and SDSS as well as
from their own observations.  If the source is important to your work, see the
6dF documentation for more information.

Note on sources:

  Source "EL6d" is a mean of ELV and 6dF.  Rare, but it does occur.  Source 
    "Smbd" is Simbad, "NED" is NED, "HIPA" is HIPASS, "RC3" is RC3.
  The HIPASS redshifts can change slightly from one of their publications to 
    another.  I've usually taken the redshift from Doyle et al (MNRAS 361, 34,
    2005) as presented by VizieR.
  When NED is the source, I usually took NED's selected redshift, but not 
    always.  See NED at https://ned.ipac.caltech.edu/ for more information as
    well as the actual source of the adopted redshift.

------

A simple analysis of the catalogue data has been presented in the printed
version of SGC; but here is an "abstract".  logD-logN plots show that SGC is 
50% complete at about (original) logD = 1.5, but has a long "tail" to smaller 
diameters.  ORIGINAL SGC diameters and axis ratios may be reduced to the 
25.0 B-magnitude isophote (RC3's standard diameter system) as follows (see the 
RC3 Introduction, Volume 1, Section 3.4, Tables 6 and 9)

log D_25 = 1.06*logD - 0.380, s.d. = 0.083, for all types

and

log R_25 = 1.41*logR, s.d. = 0.08, for T >= 0
log R_25 = 1.42*logR, s.d. = 0.08, for T < 0

NOTE AGAIN that the diameters and axis ratios given in "sgc20.dat" and
"sgc20am.dat" have already been reduced to the B = 25.0 mag arcsec^-2^
systems of RC3.

------

Acknowledgments:

I'm pleased to include the acknowledgments here that first appeared in the
printed edition of SGC:

  Our most grateful thanks go to Professors Vincent C. Reddish and Malcolm S.
  Longair, successive Directors of the Royal Observatory, Edinburgh, and 
  Astronomers Royal for Scotland, for their hospitality during this project.  
  They extended every courtesy to us during our stays in Scotland, and made
  available to us the full resources of ROE.

  The Schmidt Telescope Unit at both ROE and at Siding Spring also gave us 
  help far beyond the requirements of their office.  In particular, we thank 
  Russell Cannon, UKSTU's Director, and its members Sue Tritton, Liz Sim, Andy 
  Longmore, Tim Hawarden, John Barrow, and Russell Eberst for access to the 
  full library of plates and films at ROE, and for assistance at crucial 
  times.  We are extremely grateful to Brian Hadley, Director of ROE's 
  Photographic Laboratories, for providing excellent copies of otherwise 
  unavailable Sky Survey fields.  Katheryn Gessas, Photographer in the 
  Astronomy Department at the University of Texas, provided valuable 
  assistance in preparing the diagrams for printing.

  Finally, we thank Andris Lauberts for sending a tape copy of his ESO/Uppsala
  Survey prior to publication, and for several valuable conversations 
  concerning our mutual work of exploring the southern sky.

  This work was supported in part by the National Science Foundation through
  grant AST-8111708 to the University of Texas.

In addition, I am indebted to the California Institude of Technology, Infrared 
Processing and Analysis Center, which provided access to the Southern Sky 
Survey film copies, computing, and other physical facilities during my work 
with NED from 1991 to 2011.  NED, CDS, LEDA, ADS, HESARC, and STScI all 
provided invaluable access to on-line catalogues, the DSS, and the 
astronomical literature.

Harold Corwin 
Latest revision 23 November 2019