A Galactic Globular Cluster list (GGCl)


This list of 159 Galactic globular clusters includes revised V-band total magnitudes, as well as the implied half-light magnitudes, diameters, and surface brightnesses for 155 of the clusters. Integrated isophotal V-band magnitudes and diameters at the 22.0, 23.0, 24.0, and 25.0 magnitudes arcsecond-2 isophotes are given whenever the data are available (magnitudes for about 110 clusters, and diameters for about 115). While the list is designed for visual observers, with most of the data included to help guide visual and V-band imaging observations, other data are given as well to assist with understanding some of the clusters' astrophysical properties.

I. Introduction

This list of Galactic globular clusters is based on those by Harris (1996, 2010 edition; hereinafter "Harris") and Baumgardt et al. (2020a,b). It is primarily a list of clusters visible at optical wavelengths, though a few of the objects have been observed only in the infrared. Almost all of the data are based on optical observations, and are given here in the V-band where they were predominately collected. Very few clusters discovered since 2015 are included because they either (a) have not yet been confirmed as globulars, or (b) their V-band data are all but non-existent.

Background information on Galactic globular clusters and their relationship to the Milky Way, is given by e.g. Djorgovski and Meylan (1993, 1994), Archinal and Hynes (2003), van den Bergh (2011), and Beasley (2020), and many references therein. As well as Harris's list, previous catalogues of the Galactic globulars include (this is far from a complete list) those by Shapley (1930), Hogg (1959), Arp (1965), Alter et al. (1970), Alcaino (1973, 1977), Kukarkin (1974), Monella (1985), Djorkovski and Meylan (1993), Skiff (1999), and Archinal and Hynes (2003). Holger Baumgardt and his colleagues (Baumgardt et al., 2020a,b) are taking steps toward a new comprehensive catalogue; as of this writing (March 2021), a version with structural parameters primarily derived from N-body fits is available online (Baumgardt 2020b). This also has images for each cluster as well as radial velocities, proper motions, and orbital parameters.

As the present catalogue is aimed at amateur observers, I have focused on diameters, magnitudes, and surface brightnesses for the clusters, the immediately observable characteristics. Other data are of course given, especially those that influence what a visual observer will see through an eyepiece, or image with a camera operating in or near the visual part of the spectrum. Realizing, too, that the reason many of us observe these clusters is not just to appreciate their beauty, but to try to understand their present structure and composition, as well as their origin and evolution, I have also included data that will be of astrophysical interest, helping to tell the stories of the clusters.

This paper is organized in five sections beyond this introduction. Section II discusses the isophotal diameters at the 22, 23, 24, and 25 magnitudes arcsecond-2 isophotes; Section III presents the integrated magnitudes within those same isophotes; Section IV covers the determination of total V magnitudes; the fifth Section lays out the "other" data I've chosen to include; and the final Section briefly mentions newly-discovered candidate globulars that are not included in this catalogue.

II. Isophotal Diameters

Here, I determine diameters for Galactic globular clusters at the 22, 23, 24, and 25 mag arcsec-2 isophotes. The Trager et al. (1995) database of V-band surface brightness (SB) profiles (the run of SB versus radius "r") for the clusters provided almost all the data I used. There have been very few surface brightness profiles for globulars published in the 25 years since Trager et al. appeared, so -- with the few exceptions noted just below -- that formed my main data source. Brian Skiff (1999) used the same database to find the diameters at the 22 and 25 mag arcsec-2 isophotes; see his discussion for more background information.

Seven clusters (N4372, N4833, N6558, N6838, GCl107, Eridanus, and Pal15) not included in Trager et al. have aperture photometry that allowed me to derive SB profiles for them. Once I had those profiles in hand, I treated them identically to the SB profiles in the Trager et al. list.

When plotting each of the cluster SB profiles, I used r1/4 rather than log r or radius itself [see the Footnote]. (Figure 1 has two example profiles: NGC 6205 with adequate photometry, NGC 6496 without.) I based this variation on a comment by Harlow Shapley in his 1930 book Star Clusters (page 70): "The frequency [from star counts averaged by radius] is (roughly) inversely proportional to the fourth power of the distance from the center."

Even Shapley knew that r1/4 was a gross simplification. However, a simple linear relationship of the form

r1/4 = aV) + b,

fit across the appropriate surface brightness range, worked well for almost all the clusters. I used a plotting program ("DataGraph"; Adalsteinsson 2020) to make plots of r1/4 versus surface brightness, and used its "fit" function to find the coefficients a and b of the above equation.

I wrote the equation with r1/4 as the dependent variable to more easily calculate the radii at the four isophote levels of interest. In Table 1 (the cluster list), I've given most of the resulting diameters a nominal precision of 0.1 arcminutes -- a second digit beyond the decimal point almost always has no significance unless the diameter is very small. The statistical errors are on the order of ten percent, though are difficult to determine rigorously for the data at hand. The smaller clusters are best represented by diameters given to 0.1 arcminutes, the larger ones (~10 arcminutes across or more) to 1.0 arcminutes.

I've also tried to provide a subjective evaluation of the quality of the data for the clusters, scattering colons ":" and question marks "?" liberally around the table whenever I thought they were needed. Some clusters had SB profile data that had to be extrapolated either outwards or inwards to the desired isophote(s). I also noted these extrapolations by flagging the diameters with a colon or queston mark.

The primary source of uncertainty in the SB profiles is contamination by the star field in which a cluster is located. Subtracting the star field has, up to the Gaia and HST eras, been a statistical process. This problem is clearly present in the Trager data which has been assembled from radial star counts and aperture photometry; the photometry is also used for calibrating the star counts.

It is possible now to use the proper motions and velocities given by Gaia to determine cluster membership, at least down to about magnitude 20, roughly the reliable lower limit of the Gaia catalogue. Crowding affects the centers of most of the clusters, even in the Gaia data. However, deeper estimates, perhaps down to V = 25±, of cluster membership can also be made using color-magnitude diagrams from e.g. HST data to isolate the cluster stars using the main sequence inferred from the observed structure of the CM diagram. The selected stars can then be used for more accurate studies of the cluster, whether SB profiles, integrated magnitudes and colors, color-magnitude diagrams, space motions, or other investigations. Several research groups, including one led by Holger Baumgardt in Australia (see e.g. Baumgardt et al. 2018, 2020a,b), are working in this direction.

Professional astronomers studying globulars often quote cluster sizes in terms of "core radii" and "tidal radii". These sizes can be different depending on the model used [three popular models are the King (1966), Wilson (1975), and power-law (see e.g. Elson et al. 1987) models]. The interested reader should see McLaughlin and van der Marel (2005, and references therein) for more information on fitting the entire run of surface brightness to radius for the clusters.

Finally, I want to make clear that the r1/4 fits that I'm using across limited radii of the clusters provide little or no information of astrophysical interest. For that, standard representations of the cluster SB profiles with e.g. Wilson, King, or power-law models may be more useful. The r1/4 fits I adopted here allow me to determine cluster diameters at specific isophotes. I make no claims at all for them beyond that.

Integrated magnitudes within the different diameters are discussed in the next Section.

III. Isophotal Integrated Magnitudes

The integrated magnitudes within four isophote levels in the clusters come from the diameters determined using the luminosity profiles (see the previous Section) and the aperture-magnitude diagrams. Here, I use those isophotal diameters as input to the aperture-magnitude diagrams to find the integrated magnitudes within each of the isophote levels 22.0, 23.0, 24.0, and 25.0 V-magnitudes arcsecond-2.

Charles Peterson's valuable 1986 collection of aperture photometry of Galactic globulars is the primary source for these data. The CDS files of Peterson's paper are dated February 1996, so it is possible that additional data or corrections to his 1986 paper are contained in the files. As with the Trager et al. surface brightness profiles, there has not been much -- if any! -- aperture photometry of Galactic globulars published since Peterson's list appeared. In fact, the only additional aperture photometry that I am aware of is that from my own McDonald UBVRI photometry (Corwin 2019, Table 9) and from Robert Smyth who observed Palomar 12 from Siding Spring. Unfortunately, the light of three nearby bright field stars compromised Smyth's Pal 12 magnitudes enough that they had to be rejected; Smyth's colors are, in the mean, OK (see Corwin 2019 for more information). I sent the results for three clusters observered at McDonald to Peterson in the mid-80's (they are included in his list), but I observed another 15 clusters after that. I've also included their data in this present work.

As with the diameters, I fitted linear relationships to the largest aperture measurements for each cluster. I tried quadratic fits for a few clusters, but found that linear fits were easier to work with and were -- within the errors of the photometry -- no less accurate. So, most of the magnitudes come from the linear fits to photometry from the larger apertures. Figure 2 has three examples of the aperture-magnitude diagrams with the linear fits: 1) NGC 6205, 2) NGC 7492, and 3) GCl 107. NGC 7492 and GCl 107 have progressively fewer data points, and subsequently progressively larger errors in their fits.

In a few cases, I estimated the magnitudes from interpolation between the magnitude in the largest measured aperture and the total magnitude, even if isophotal diameters were impossible to determine due to missing data in the surface brightness profiles. I always marked the magnitudes and diameters as uncertain or questionable if there was any reason to suspect that they were less reliable than those of a cluster with clearly adequate data.

In most cases, the unreliable derived data were simply a reflection of missing observations. In some cases, though, I suspect that the inclusion or not of cluster and field stars limits both the accuracy and the precision of the aperture photometry. A slightly different centering of even a "large" aperture could make at least a small difference in the observed magnitude, as could the random inclusion or not of stars in the nearby "blank" sky comparison fields.

Table 1, the cluster list, includes all of these isophotal integrated magnitudes.

IV. Total Magnitudes

After some preliminary work with several lists of total magnitudes of the Galactic globulars, I settled on intercomparing data from the following four:

  Harris (1996, 2010 revision, "H")
  Baumgardt et al. (2020a, "B")
  Dalessandro et al. (2012, "D")
  McLaughlin and van der Marel (2005, "M")

A simple comparison of the lists, one minus another, leads to the standard deviations and associated weights, w = (0.1/σ)2, in Table 2:

Table 2:

  Source       Standard Deviation   Weight
  Harris              0.18            3.1
  Baumgardt           0.29            1.2
  Dalessandro         0.34            0.9
  MandM               0.18            3.1

These are after one or two cycles of 2-σ rejections of discordant data, and after running the standard deviations through a triangular comparison to find "true" standard deviations. There is more on this step below.

Baumgardt et al. (2020a) make the point that fainter clusters have larger statistical errors in their photometry than the bright ones. The practical result is that there may be pretty good agreement among the four sources for bright clusters (VT < 10), but not so good for the fainter (VT ≥ 10) objects. So, I've redone the comparisons for bright and faint samples. Table 3 has those comparisons after one cycle of 2-σ rejection, again to remove outlying data:

Table 3a: Mean differences for VT < 10.0:

            H min B  H min D  H min M  B min D  B min M  D min M
 mean diff  -0.086   -0.123   -0.114   -0.025   -0.052    0.019
  sigma_1    0.236    0.407    0.247    0.502    0.292    0.263
  sigma_n    0.025    0.070    0.031    0.086    0.036    0.048
        n     91       34       63       34       65       30

Table 3b: Mean differences for VT ≥ 10.0:

            H min B  H min D  H min M  B min D  B min M  D min M
 mean diff   0.232    0.009    0.024   -0.276    0.005   -0.403
  sigma_1    0.675    0.463    0.632    0.621    0.888    0.527
  sigma_n    0.095    0.164    0.153    0.220    0.209    0.199
        n     51        8       17        8       18       7

We can use these standard deviations from the differences, through triangular comparisons (see the Appendix), to find the standard deviations of each source. Table 4 has the results:

Table 4a: VT < 10.0:

  Source       Standard Deviation   Weight
  Harris              0.25            1.6
  Baumgardt           0.20            1.8
  Dalessandro         0.36            0.8
  MandM               0.19            2.2

Table 4b: VT ≥ 10.0:

  Source       Standard Deviation   Weight
  Harris              0.48            0.4
  Baumgardt           0.27            1.3
  Dalessandro         0.45            0.5
  MandM               0.56            0.3

I used the standard deviations and weights in Table 4 to calculate mean VT's and standard deviations for all the included clusters with VT from any of the four sources.

Then, I compared these mean VT's with the log(aperture)-magnitude (just "aperture-magnitude" hereinafter) diagrams for each cluster; examples are in Figure 3: NGC 6205 again, an example of a cluster with adequate and accordant photometry; and Palomar 15 with only one photometric data point. I plotted the total magnitude at twice the diameter of the largest aperture used for an actual observation. This usually put the total magnitude far enough out on the diagram that I could ignore it, or use it as an extrapolated point for estimating adjustments based on the observations.

As a first approximation, I used the half-light diameters from Harris as the "aperture" for the half-light magnitude (in our work on galaxies, we called these "effective" magnitudes, Ve = VT + 0.75; I'll adopt this notation for convenience). I examined each diagram and adjusted the half-light magnitude and half-light diameter until they matched the actual run of the observations. I adopted those as the final Ve and Ae for each cluster. These then gave an adjusted VTa which was of course just 0.75 magnitudes brighter than the adjusted Vea.

So, my final total magnitudes and half-light diameters are closely tied to the actual V-band aperture photometry. However, most of the recently-discovered clusters have no V-band aperture photometry, so depend on secondary calibrations based on star counts and summed stellar photometry. For these, I've simply adopted a weighted mean VT, and the half-light diameter from Harris's list, or from more recent determinations in the literature.

How well do these "adjusted" VTa's compare to the previously published data? Table 5 has the triangular comparisons answering that question, with the individual comparisons first, again after one cycle of 2-σ rejection. Rather than using simple zero-point offsets as I did above, I found the coefficients of linear equations of the form V1 = aV2 + b along with the associated standard deviations. The notation "impartial line" means that the solutions are essentially the means of the two ordinary least squares relationships between the magnitudes, the first as above, the second as V2 = a′V1 + b′ (see e.g. Isobe et al. 1990 and Stromberg 1940 for details and discussion).

Table 5a: Coefficients and standard deviations of linear impartial lines, VT(Adj) = a(±ea)VT(B|M|H|D) + b(±eb), ±s.d., where "ea" and "eb" are the errors on the coefficients.

  Sources       a     e_a       b     e_b   s.d.
  Adj vs B    0.993  0.010   -0.002  0.086  0.20
  Adj vs M    1.014  0.020   -0.196  0.168  0.30
  Adj vs H    0.982  0.009   +0.133  0.078  0.18
  Adj vs D    0.950  0.036   +0.285  0.296  0.38

Table 5b: Coefficients and standard deviations of linear impartial lines for the published sources, VT(S1) = a(±ea)VT(S2) + b(±eb), ±s.d.

  S1  S2        a     e_a       b     e_b   s.d.
  B vs M      0.990  0.020   +0.131  0.171  0.34
  B vs H      1.016  0.013   -0.173  0.111  0.27
  B vs D      1.064  0.041   -0.425  0.357  0.50
  D vs H      0.993  0.029   +0.190  0.251  0.36
  D vs M      1.059  0.029   -0.439  0.256  0.34
  M vs H      0.970  0.017   +0.348  0.149  0.30

Triangular comparisons, with the standard deviations input from Table 5, lead to the "true" standard deviations in Table 6 (the "n" following two values means the square of the standard deviation is negative, usually indicating a very small "true" standard deviation compared to the others; the square roots shown in the table assume that the signs are positive -- which they are not!):

Table 6:

   B     M     Adj   |   H     D     Adj  |   H     M     Adj
  0.19  0.29  0.06   |  0.09  0.35  0.15  |  0.13  0.27  0.13

   H     B     Adj   |   M     D     Adj  |   B     D     Adj
  0.18  0.20  0.02n  |  0.17  0.29  0.24  |  0.42  0.27  0.18n

From the standard deviations in Table 6, the mean standard deviations and weights for each source are given in Table 7.

Table 7:

  Source       Standard Deviation   Weight
  Harris              0.13            5.9
  Baumgardt           0.27            1.4
  MandM               0.24            1.7
  Dalessandro         0.30            1.1
  Adjusted            0.15 (0.13)     4.4 (5.9)

These agree more or less with the numbers in Table 2 above for a triangular comparison of the four published sources after rejection of discordant points. (The Adjusted mean standard deviation excludes the two negative squares; assuming they are indicative of the positive squares, the mean becomes 0.13. I adopted 0.15 for adjusted magnitudes in the cluster list). The final total magnitudes that I adopted are either the adjusted values agreeing with the aperture photometry, or a weighted mean of the published magnitudes for those clusters with no (or unreliable) aperture photometry.

V. Other Data

Other data in Table 1 include names and alternate names, RA and Dec for J2000.0, the Shapley-Sawyer-Hogg concentration class, the axial ratio, the central surface brightness, V magnitudes for the tip of the red giant branch and the horizontal branch (RR Lyrae stars), the distance, heliocentric radial velocity, the cluster metallicity and age, and the total Galactic absorption in the direction of the cluster. Almost all of those data come from either Harris or from Brian Skiff's (1999) globular cluster list. I encourage you to check into those lists, if for nothing else than to chase down the many references to the publications carrying the earlier data. I have filled in missing data where they exist from more recent literature and I've given references to the papers from which I took the data.

Here is a bit of explanation about individual data columns in Table 1.

1) Names. The primary name I adopt is widely used throughout the literature, while the secondary name is usually less commonly encountered. There are some exceptions, e.g. 47 Tuc = NGC 0104, Omega Cen = NGC 5139, as well as all of the clusters in the Messier list.

2) Equatorial coordinates for J2000.0 are from a) my NGC/IC lists of "selected" positions that best represent the location of the cluster on the sky (see NGC/IC positions for the positions and their sources); b) Harris's catalog; or c) the original papers if neither of the primary lists carried the position. The coordinates are rounded off here to the nearest 0.1 seconds of time and 1 arcsecond. If needed, Galactic coordinates can be computed from these equatorial positions, or can be found in Harris's list.

3) The concentration class is from Sawyer Hogg (1959; based on those given by Shapley and Sawyer, 1927). This subjective assessment of the apparent "compactness" of each cluster is more or less correlated with surface brightness, either the central surface brightness, or that within the effective aperture. More recent estimates of the concentration (usually the ratio of the core and tidal radii; see e.g. McLaughlin and van der Marel, 2005) -- again loosely correlated with the Shapley-Sawyer-Hogg estimates -- are wrapped in theoretical models, so may be less appropriate for visual observers than the older estimates. But those older ones come from photographic plates taken primarily in the 1920s, so can be affected by the telescope size, exposure time, limiting magnitude, stellar background density, and so on.

4) The axial ratio (= b/a = 1 - e) is from White and Shawl (1987) or Harris. Many clusters are noticeably elliptical and the axial ratio (or the ellipticity "e") gives an indication of how "out of round" the cluster actually is. Flattening suggests rotation, and some clusters have indeed been found to rotate (see e.g. Bianchini et al. 2013 and references therein).

5) The central surface brightness, not corrected for foreground absorption (from Harris and references therein), is found from fits of the run of surface brightness with radius (see Section II, above). Harris makes the point that these numbers are limited by the seeing and resolution of the ground-based telescopes used to collect the data. Many of the clusters studied with the Hubble Space Telescope have considerably brighter central surface brightnesses than the numbers quoted here. Some of the clusters have "collapsed" cores that can strongly influence the central surface brightness. These numbers may also be influenced by the presence or not of bright stars (particularly red giants) superposed on the centers of the clusters. Even single stars can affect the central surface brightness.

6) The V-magnitudes of the tip of the red giant branch, and of the horizontal branch, are taken primarily from Brian Skiff's 1999 list. For the horizontal branch magnitudes, I've filled in blanks from more recent papers, or have in a few cases taken Harris's number if it is from a more recent paper than the one Skiff referenced. The largest difference between Skiff's and Harris's number is half a magnitude, but discrepancies of that size are rare. I note in passing that these two magnitudes (giant branch tip and horizontal branch) can be used as distance indicators for the globulars, and that the horizontal branch is usually the same as, or is close to, the magnitude at which RR Lyrae stars are found in the clusters. Skiff has an enlightening discussion of the relation of these parameters to the apparent resolveability of the cluster. This will be of special interest to visual observers.

7) Distances from the sun to the globular cluster are taken from Francis and Anderson (2014) for the majority of the clusters, while the remainder are from Harris, Baumgardt et al. (2019), or more recent studies as noted in the cluster list. Harris (and several other cataloguers) also give Galactocentric distances; there are several interesting correlations between various parameters for the clusters and the distance from the center of the Milky Way. One of these is mentioned in the next paragraph.

8) and 9) Heliocentric radial velocities and the metallicity come primarily from Harris, Baumgardt et al. (2019), or other papers as noted. Velocities are usually the mean value for several stars in the cluster, while metallicities are expressed in the usual manner of base-10 logarithms of the ratios of "metals" to hydrogen, on a scale zeroed at the solar metallicity of 1.0, or [Fe/H] = [Fe/H]cluster - [Fe/H]solar= 0.0. Virtually all globulars are deficient in metals with respect to the sun, so their [Fe/H]'s are negative. Metallicity for the globulars is related to at least two parameters. First is the location of the cluster (see e.g. Harris and Harris, 1999) -- "metal-poor" clusters are typically found in the corona ("halo") of the Galaxy, while the "metal-rich" clusters are almost always found in the Galactic bulge. Second is the absolute magnitude of the horizontal branch of the cluster -- Harris gives MV(HB) = 0.16[Fe/H] + 0.84 as a general relationship. See his discussion for an introduction to the many details, as well as references to the appropriate literature.

10) The total V-band absorption in the direction of the cluster is taken from Harris, or from NED's absorption calculator based on Schlafly and Finkbeiner's (2011) reclaibration of the extinction model by Schlegel, Finkbeiner, and Davis (1998). While the absorption estimates from the model are appropriate for galaxies seen beyond the Galactic absorbing clouds, they must be an approximation to the true absorption for many clusters because the objects are superposed on Galactic dust and gas as well as being behind or immersed in it. Also, the calculated total absorptions are subject to large uncertainties at low Galactic latitudes. More accurate estimates of the total absorption are available from color-magnitude diagrams or from spectrophotometry, so Harris adopted those estimates whenever possible.

11) Ages. Most of the Galactic globular clusters are older than 10 gigayears, often in the range of 11-12.5 Gyr, and sometimes as old as 13 Gyr. This helps explain their fascination for many astronomers -- the universe itself is just 13.7 Gyr old in the ΛCDM cosmological models now generally accepted. Many of the globulars, following their elliptical orbits around the nucleus of the Milky Way, in spite of occasional dips into the bulge of the Galaxy, have essentially aged in place in the corona, unperturbed by the continued star formation in the Galaxy's arms. With gravitational fields from tens or hundreds of thousands of stars holding them together, the globulars are survivors from the Milky Way's earliest eras. This makes their study particularly valuable for understanding stellar birth early in the universe.

12) Finally, the B-V color index, also not corrected for Galactic extinction, is -- with two exceptions -- from Harris. The exceptions:

a) For NGC 7492, I've adopted my unpublished McDonald photometry, B-V = +0.57, which agrees with the B-V = +0.59 transformed from g-r in Vanderbeke et al. 2014 (Harris's 2010 value, +0.42 from Johnson 1959, would make this the bluest known Galactic globular and is probably wrong).

b) For Palomar 12, I've adopted +0.86, a mean of the color from Racine's (1975) large aperture (+0.88), and R.J. Smyth's smallest (+0.83, unpublished Siding Spring photometry, but see Corwin 2019). Harris's (2010) value, +1.07, a mean of Racine's two apertures, is apparently contaminated by Racine's small aperture measure B-V = +1.22 (uncertain, or including several red stars). As mentioned above in Section III, superposed stars included in his two largest apertures strongly affect Smyth's photometry for Palomar 12.

The observed colors for the clusters are of course strongly dependent on Galactic extinction; after correcting for that, the mean intrinsic color is <(B-V)0> = 0.70 +- 0.14 (see Figure 4 and its caption).

VI. Candidate Galactic Globular Clusters

New globular cluster candidates are being found with some frequency at the moment (~2015 to the present) thanks to deep infrared sky surveys; see e.g. Ryu and Lee (2018), Palma et al. (2019), Garro et al. (2021) and references therein. While many of these newly-discovered objects may indeed turn out to be genuine Galactic globulars, I have chosen to not include them here because of their lack of visibility in the visual part of the spectrum, and their subsequent lack of reliable V-band information. Most are being found at very low Galactic latitudes which makes their detection at optical wavelengths very difficult with current technology -- the total absorption from dust in the visual range is large enough to render many of these clusters completely invisible in the V-band.

There are nevertheless a few clusters in the present list with total V-band absorptions of ten magnitudes or more, and two clusters -- GLIMPSE 02 and 2MASS GC01 -- with AV of more than twenty magnitudes.

Two clusters included here -- GLIMPSE 02 and Sgr 2 = Laevens 5 -- are not included in the Baumgardt et al. (2020b) list, and Sgr 2 was found after Harris's catalog appeared.

Finally, the Baumgardt et al. list has two clusters -- BH 140 and GLIMSPE 01 -- that I have not included here; these are listed in Table 8 along with four from Harris that currently also have questionable status. One object, TJ 17, included in Peterson's (1986) collection of aperture photometry, is a galaxy.

Table 8: "Globular clusters" in Baumgardt et al. (2020b), Harris, and Peterson (1986) not in the current list.

Name           J2000 Position      V_T    D_e  mu_e  Class    
Koposov 2   07 58 17.0  +26 15 19  17.6   0.4  16.6  Open cluster?
Koposov 1   11 59 18.4  +12 15 36  17.1   0.5  16.1  Open cluster?
BH 140      12 53 53.5  -67 10 38   9.1  12.7  15.1  Open cluster:
BH 176      15 39 07.5  -50 03 10  14.0   1.8  16.3  Open cluster = ESO 224-008
TJ17 = TB1  17 18 41.5  -27 50 14  16.5:  0.3? 13.6? Galaxy (see Terzan and 
                                                      Bernard, 1978)
VVV Cl001   17 54 42.5  -24 00 53  17.7   5.6  21.9  Possible globular cluster
GLIMPSE 01  18 48 49.7  -01 29 50  22.2   1.3  23.8  Open cluster?

VII. Conclusion

Relying primarily on published data, I've found revised total and half-light magnitudes, half-light diameters, and integrated isophotal magnitudes, all in the V-band, for 155 Galactic globular clusters. An additional four clusters are in the list as examples of recently-discovered, heavily-obscured objects found at low Galactic latitudes. I've also collected other data relevant to visual observations of the clusters, as well as data to help with understanding the clusters' astrophysical properties. All these data are presented as an updated list of Galactic globular clusters. References to other recent catalogues and studies are included, so that issues not addressed here may be explored at the reader's convenience.

VIII. Acknowledgements

This project came about thanks to a comment by Scott Harrington in the amastro discussion group about the magnitudes of NGC 4147 and NGC 5694 (total mags, Scott, are 10.25 and 9.94 respectively; Table 1 has a lot more!). Brian Skiff made suggestions that considerably improved this discussion, and his 1999 globular list has also provided data and inspiration. Thanks also go to Charles Peterson for his big collection of UBVRI aperture photometry for the globulars (including his own observations for many of them); and to Bill Harris for his remarkable "McMaster Catalog" of Galactic globulars -- he showed us how to get the job done. I have also made use of the NASA/IPAC Extragalactic Database (NED) which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration.


Adalsteinsson, D.  2020, "DataGraph 4.6.1", Visual Data Tools, Chapel Hill, NC 
  (see https://community.visualdatatools.com/datagraph/ ).
Alcaino, G.  1973, "Atlas of Galactic Globular Clusters", Universidad Catolica 
  de Chile.
Alcaino, G.  1977, PASP 89, 491.
Alcaino, G.; Liller, W.; Alvarado, F.; and Wenderoth, E.  1991, AJ 102, 1371.
Alter, G.; Balazs, B.; Ruprecht, J.; and Vanysek, J.  1970, "Star Clusters", 
  Budapest: Akademiai Kiado, 2nd ed.
Archinal, B.A. and Hynes, S.J.  2003, "Star Clusters", Richmond, VA: 
  Willmann-Bell, Inc.
Arp, H.  1965, in "Galactic Structure" ed. A. Blaauw and M. Schmidt, Univ. 
  Chicago Press, Chapter 19 
  (see http://articles.adsabs.harvard.edu/pdf/1965gast.book..401A).
Barba, R.H., et al.  2019, ApJL 870, 24.
Baumgardt, H.; Hilker, M.; Sollima, A.; and Bellini, A.  2018, MNRAS 478, 1520.
Baumgardt, H.; Hilker, M.; Sollima, A.; and Bellini, A.  2019, MNRAS 482, 5138.
Baumgardt, H.; Sollima, A.; and Hilker, M.  2020a, PASA 37, 46.
Baumgardt, H. et al.  2020b, "Fundamental parameters of Galactic globular
  clusters" (see https://people.smp.uq.edu.au/HolgerBaumgardt/globular/ )  
Beasley, M.A.  2020, in "Reviews in Frontiers of Modern Astrophysics. From
  Space Debris to Cosmology", ed. Kabath, P.; Jones, D.; and Skarka, M., 
  p. 245, Springer Nature Switzerland, Cham.
Bellazzini, M.; Ferraro, F.R.; Ibata, R.  2002, AJ 124, 915.
van den Bergh, S.  2011, PASP 123, 1044.
Bianchini, P.; Varri, A.L.; Bertin, G.; and Zocchi, A.  2013, ApJ 772, 67.
Bica, E.; Ortolani, S.; and Barbuy, B.  1994, A&A 283, 67.
Bonatto, C. and Bica, E.  2008, A&A 479, 741.
Carballo-Bello, J.A.; et al.  2016, MNRAS 462, 501.
Carraro, G.; Zinn, R.; and Moni Bidin, C.  2007, A&A 466, 181.
Corwin, H.G. 2019.  See http://haroldcorwin.net/ubvri/galptmy.html .
Dalessandro, E. et al.  2012, AJ 144, 126.
Djorgovski, S.G. and Meylan, G.  1993, in "Structure and Dynamics of Globular 
  Clusters", ed. S.G. Djorgovski and G. Meylan; pp. 325, 347, and 373, 
  Astronomical Society of the Pacific, San Francisco.
Djorgovski, S.G. and Meylan, G.  1994, AJ 108, 1292.
Elson, R.A.W.; Fall, S.M.; and Freeman, K.C. 1987, ApJ 323, 54.
Fernandez-Trincado, J.G.; et al.  2020, A&A 643A, 145.
Francis, C. and Anderson, E.  2014, MNRAS 441, 1105.
de la Fuente Marcos, R.; et al.  2015, A&A 581A, 13.
Gaia Collaboration; Brown, A. and a cast of thousands.  2020, A&A, in press 
  (see https://doi.org/10.1051/0004-6361/202039657 ).
Garro, E.R.; et al.  2021, A&A, in press.
Harris, H.C. and Harris, W.E.  1999, Chapter 22 in "Allen's Astrophysical 
  Quantities", ed. A.N. Cox, AIP Press/Springer Verlag.
Harris, W.E.  1996, AJ 112, 1487.
Harris, W.E.  2010, see https://physwww.physics.mcmaster.ca/~harris/mwgc.dat .
Hogg, H. Sawyer.  1959, "Star Clusters", in Handbuch der Physik, Vol. LIII, 
  ed. S. Flugge, Springer-Verlag, Berlin.  
Isobe, T.; Feigelson, E.; Akritas, M.G.; and Babu, G.J.  1990, ApJ 364, 104.
King, I.R.  1966, AJ 71, 64.
Kukarkin, B.V.  1974, "The Catalogue of Globular Star Clusters of Our Galaxy", 
  Sternberg State Astronomical Institute, Moscow.
Kurtev, R.; Ivanov, V.D.; Borissova, J.; and Ortolani, S.  2008, A&A 489, 583.
Laevens, B.P.M.; et al.  2014, ApJL 786, 3.
Laevens, B.P.M.; et al.  2015, ApJ 813, 44.
Longeard, N.; et al.  2019, MNRAS 490, 1498.
Longmore, A.J.; et al.  2011, MNRAS 416, 465, 
McClure, R.D.; Hesser, J.E.; Stetson, P.B.; and Stryker, L.L.  1985, PASP 97, 
McLaughlin, D.E. and van der Marel, R.P.  2005, ApJS 161, 304.
Minniti, D.; et al.  2017, ApJL 838, 14.
Monella, R.  1985, Coelum LIII, 287 (see CDS catalogue VII/103).
Mutlu-Pakdil, B.; et al.  2018, ApJ 863, 25.
Ortolani, S.; Bica, E.; and Barbuy, B.  1995, A&A 296, 680.
Ortolani, S.; Bica, E.; and Barbuy, B.  2003, A&A 402, 565.
Palma, T.; et al.  2019, MNRAS 487, 3140.
Peterson, C.  1986, PASP 98, 192.
Racine, R.  1975, AJ 80, 1031.
Sarajedini, A. and Geisler, D.  1996, AJ 112, 2013.
Schlafly, E.F. and Finkbeiner, D.P.  2011, ApJ 737, 103.
Schlegel, D.J.; Finkbeiner, D.P.; and Davis, M.  1998, ApJ 500, 525.
Shapley, H.  1930, "Star Clusters," McGraw-Hill, New York.
Shapley, H. and Sawyer, H.  1927, Bull. Harvard College Obs. 849, 11.
Skiff, B. A.  1999, "Observational Data for Galactic Globular Clusters", 
  Lowell Observatory FTP site, Flagstaff (see 
  ftp://ftp.lowell.edu/pub/bas/deepsky/globulars.ref and ~/globulars.tab ).
Stromberg, G.  1940, ApJ 92, 156.
Terzan, A. and Bernard, A.  1978, The Messenger 15, 14.
Trager, S.C.; King, I.R.; and Djorgovski, S.  1995, AJ 109, 218.
Vanderbeke, J.; et al.  2014, MNRAS 437, 1725.
Webbink, R.F.  1985, in IAU Symp 113 "Dynamics of Star Clusters", ed. J. 
  Goedman and P. Hut, p. 541, IAU (see also CDS catalogue number VII/151).
White, R.E. and Shawl, S.J.  1987, ApJ 317, 246.
Wilson, C.P.  1975, AJ 80, 175.

Appendix: Triangular Comparisons

Triangular comparisons of the standard deviations can, in principle, give the true statistical errors of each source. For a comparison between sources "1" and "2", we write the usual sum of squares,

12)2 = (σ1)2 + (σ2)2.

There are identical equations for sources 1 and 3, and sources 2 and 3:

13)2 = (σ1)2 + (σ3)2 and

23)2 = (σ2)2 + (σ3)2.

Rearranging and combining terms, we find the individual standard deviations:

1)2 = (σ12)2 - (σ23)2 + (σ31)2,

2)2 = (σ12)2 - (σ31)2 + (σ23)2, and

3)2 = (σ23)2 - (σ12)2 + (σ31)2.

There is a possibility that one source may have much smaller errors than the other two. In that case, the right hand side of the equation may be negative. I've flagged these negative squares in the tables with an "n" following the square root of the number with its sign switched to "+". This is certainly not mathematically viable, but nevertheless seems to give an idea of the size of the standard deviation.

Figure Captions and Footnote

Figure 1: Surface brightness profiles with r1/4 fitted across the 22.0 to 25.0 magnitude arcsecond-2 isophotes.

Figure 2: Aperture-magnitude diagrams with linear fits to sections across 22-25 magnitudes arcsecond-2.

Figure 3: Observed aperture-magnitude diagrams. The aperture is expressed in base-10 logs of 0.1 arcminutes (to avoid negative logs).

Figure 4: Histogram of intrinsic B-V colors. The mean color is (B-V)0 = 0.70 ± 0.14. Only three clusters have (B-V)0 close to or larger than 1.0: NGC 6528 (0.99), NGC 6553 (1.10), and Palomar 6 (1.37). With large and uncertain extinctions, these probably have intrinsic B-V colors more in line with the remainder of the clusters.

Footnote: My work with Gerard de Vaucouleurs -- who used the r1/4 law to model the radial light distribution in elliptical galaxies -- of course had a bit to do with this choice, too! However, I did not come across Shapley's book until well after de Vaucouleurs's death in 1995. So, I was unable to ask him if his use of the r1/4 relation arose from his having read Shapley's book, or if he independently developed it as a graduate student in the 1940s. I know that in the years I worked with him, he regarded the r1/4 law as his own discovery. Nor would I be surprised to learn that he developed it on his own -- r1/4 is easy on a slide rule: two square roots! But I also know that he had a copy of Shapley's "Star Clusters" in his personal library.

Harold G. Corwin, Jr.
Email: hcorwin1153@sbcglobal.net

Latest update: 19 March 2021