Dr. James Schombert
Dept. of Physics
University of Oregon
Eugene OR 97403
Dear Dr. Schombert
Your paper "THE BUTCHER-OEMLER EFFECT IN ABELL 2317" was sent to
an anonymous but competent referee. The referee's report contains
a number of minor comments aimed at improving the scientific presentation
and clarity of the manuscript. In addition, there are two significant
scientific concerns which require addressing:
1. Membership bias. Redshift 0.2 is also the land for many of the
Faint Blue Galaxies (e.g. Lilly etal 1995, Ellis etal 1995) and hence
its possible that if there are interlopers in your "cluster" sample,
f_b will be higher. I don't believe this is a serious problem in your
data but at needs to be discussed, and the amplitude of any contamination
needs to be assessed.
2. The referee is unconvinced that your photometric system can
unambiguously isolate galaxies undergoing normal star formation.
I think this will be a problem for the general reader that relies mostly
on H-alpha EW measures to determine this. Hence, you need to provide
more background on your photometric system and have a more precise
discussion of how it maps into the convention UBVR colors we all
know and love.
I would also like to add a couple of more concerns of mine that were
raised after reading the manuscript. These points also appear in
the referee's report.
1. You really must do a better job with the fading arguments and quantify
the amount of fading. In general, I think the fading vector between
z = 0.2 and z = 0.0 is too large to allow for star burst objects to
disappear in today's clusters. Between z = 0.6 and z = 0.0 I can believe
this, but not between z = 0.2 and z = 0.0.
2. Your highly reddened starbursts need better justification as A_v
>2.0 magnitudes is severe.
Please send one copy of the revised manuscript directly to me at
the University of Oregon.
Sincerely,
Greg Bothun
Scientific Editor
Referee's Report:
This paper presents a new estimate of the fraction of blue
galaxies (i.e., the Butcher-Oemler effect) in Abell 2317 based on
rest-frame Stromgren photometry. Although similar studies for this
cluster have already been carried out, the novelty of this paper lies on
the use of a larger, deeper sample in order to study the nature of the
blue galaxy population. Notwithstanding a couple of major concerns
raised below, I am willing to recommend a revised version of this
paper for publication in the Astrophysical Journal.
In previous papers of this series, the authors have addressed
the reliability of the Stromgren photometric indices to estimate the
fraction of blue galaxies in distant clusters, and discriminate among
various galaxy types. However, I am concerned about a possible bias in
the determination of the cluster membership for Abell 2317 that might
affect the results, and that has not been discussed in this paper (see
comment 5). Also, I do not find convincing the author's conclusion
that, using these data alone, the Butcher-Oemler galaxies are shown to be
normal star-forming systems (see comment 26). Finally, I would ask the
authors to proofread the revised version much more carefully to avoid
the ridiculously large amount of typos present in the current
version. Most of the comments below refer to minor details, but I
would like comments 3, 5, 6, 10, and 26 to be addressed before the
paper is accepted.
0.- p2: The sentence: "...Butcher-Oemler galaxies are shown to be
galaxies with normal star-formation rates..." should
be toned down greatly or rephrased altogether. I do not
believe that the authors have shown this with these data
(see comment 26).
1.- p3: "Ellis 1996", not in references
"Butcher and Oemler 1985", not in references (is it 1984?)
"...than a simply calculation..." => "...simple..."
"...trend then an..." => "...than..."
2.- p4: "Rakos and Schombert (1995)" => RS95
3.- p5: How were those 200 objects selected? Is it a magnitude-
limited sample? Give complete description of the sample
selection.
4.- p5: "objectsobjects" => objects
"66%" => "56%"
"mz indice" => index
5.- p5: How would the cluster velocity dispersion affect the
membership assignment derived using the mz index? What
would be its effect in such narrow-band colors? I am concerned
that the velocity dispersion of a Coma-like cluster, such
as Abell 2317, may bias the selection of cluster members
and their colors in low-redshift clusters (i.e., z<0.2). The
galaxy sample was selected from a projected area of radius 0.6
Mpc around the cluster centre. For a similar area, Coma
galaxies show a typical velocity dispersion of 1200 km/s
(Kent & Gunn 1982) which implies a shift of +/-100A (1sigma)
or +/-300A (3sigma) at 5000A. Describe the actual
criterion adopted to select cluster members, and assess any
possible bias introduced by the cluster velocity dispersion
in both the membership assignment and the derived colors.
6.- p5: What is the uncertainty in the measured uz, vz, bz, yz colors
and the mz index? Plot error bars in figures.
7.- p6: "...red vz-yz color..." => "...red vz-yz color (i.e.,
vz-yz > 0.65)..."
8.- p6: "...to produce a Seyfert classification." => was classified
as Seyfert.
"...uncertainly..." => uncertainty
"...boundarys..." => boundaries
"...the lower bound of starbursts..." => upper
"...subsequential..." => subsequent (check throughout the text)
9.- p7: What absolute magnitude corresponds to -20.7? (M_B?)
10.- p7: Section 3.1 should be shortened by at least one page, since
the first paragraph is an exact copy of the first paragraph
of section 3 in RS95, and the second paragraph is largely
a repetition of the ideas already described in section 2.
11.- p7: "...only sensitive solely..." => only sensitive
12.- p8: "...mergering..." => merging (check throughout the text)
"...reddening lines..." => line
13.- p8: If the majority of starburst galaxies are corrected for
2 to 3 magnitudes of extinction (as stated in the text),
then, according to the extinction vector shown in figure 2,
most starbursts would have extinction-corrected colors much
bluer than any other galaxy shown in this figure. Is the
size of the extinction vector correct? If so, the extinction
in the starburst sample may amount to only 1-1.5 mag.
14.- p9: "...photometrically classified galaxies of..." => galaxies
photometrically classified as
"...Ingoring..." => Ignoring
"...the this..." => this
15.- p9: Give reference for the present-day ratio of E/SO to red
Sp/Irr in clusters.
16.- p9: I do not entirely agree with the statement: "If the starburst
population fades, then this suggests that the blue population
does not evolve into the red population, but must be eliminated
from the cluster population in order to match the morphological
fractions of present-day clusters". For instance, an important
cluster population that has not been considered in the analysis
is the spheroidal galaxy class (or dwarf ellipticals). If
low-mass starbursts do indeed evolve into today's spheroidal
galaxies, as suggested by some studies (e.g., Koo et al. 1997),
then they are still part of the cluster population but will
not affect the present-day ratio of E/S0 to red Sp/Irr.
Vanishing of the blue population is not the only plausible
evolutionary scenario, and other scenarios should be also
stated in the discussion.
17.- p10: Give reference for the observed AGN fraction in 3C 295.
18.- p10: "...a single burst stellar populations..." => population
"...,whereas, the..." => whereas the
19.- p11: "...change dramatic..." => dramatically
"...increase up..." => increases
"...a elliptical..." => an
"...Bender 1996..." => Bender et al. 1996
"...are scatter..." => scattered
"...Schombert and Kreidl (1991)" not in references
"...Matsushima 1969..." not in references
20.- p12: "...a galaxies SED..." => galaxy
21.- p12: Justify in detail why the 4000-break falling in the F
passband translates into a low f_B in cluster 0016, as
derived from the observed J-F colors, (or drop this discussion
altogether, since it is irrelevant to the main thesis of the
paper).
22.- p13: "...that dominant..." => dominate
"...inspect..." => inspection
"...also shows..." => show
"...There some..." => There is
23.- p13: Although figure 3 suggests that non-E/S0 galaxies brighter
than L* tend to be bluer than their fainter counterparts,
neither figure 3 nor figure 4 show, in my opinion, that
"the brightest galaxies (regardless the morphological colors)
have bluest colors". Figure 4 does not show any significant
difference, and certainly there is no such trend for E/S0.
This sentence should be rephrased accordingly.
24.- p14: I do not understand the statement "...than (then?) fade below
detection limits by the present epoch." The bursting dwarfs
in the evolutionary scenario proposed by Koo et al. (1997)
would fade enough to become today's spheroidals, hundreds
of which have been observed in nearby clusters (e.g.,
Binggeli and Cameron 1991).
25.- p14: "...mergering..." => merging
"...of the either the..." => of either
26.- p14: One should be very cautious when summarizing the results
based on figures 1-4, i.e.:
"the main conclusion from Figures 3 and 4 is in agreement
with the morphological studies in that the Butcher-Oemler
population is primarily composed of galaxies with normal star
formation rates typical of present-day Sp/Irr". This statement
is based on the artificial boundaries defined in Figures 1 and
2 to characterize the various galaxy types. I have two major
concerns: a) as the authors discuss in section 2, the
boundaries between Sp/Irr and starburst, or Seyfert galaxies,
are very uncertain. Error bars should be indeed
plotted in these two figures. Considering these errors, is
the proposed Sp/Irr population significantly different from
that characterized as starbursts or Seyferts in the
neighbouring regions? Since there is no real separation
between the three galaxy types in the parameter space and
the errors may be quite sizeable, the likely contamination
of the Sp/Irr class by both starbursts and Seyferts will
diminish the value of the above conclusion; b) About one
third of the galaxies classified as Sp/Irr have bz-yz<0.15,
vz-yz<0.15, and mz>-0.15. There is only one true spiral that
lies in the same area of the parameter space in Rakos et al.
1996. Indeed, most of the nearby galaxies in that area are real
Seyferts. Hence, from the reference nearby sample studied by
Rakos et al. 1996, there is no clear observational evidence to
support the idea that at least a third of the galaxies
classified as Sp/Irr in Abell 2317 are indeed normal
star-forming galaxies. I suggest the authors should tone
down any results based on this conclusion, including the
sentences in the abstract, section 3.5, and summary.
27.- p15: "...is show..." => shown
"...that that..." => that
"...the the..." => the
"...subsequential..." => subsequent
"...is agreement..." => is in
28.- p16: Do Koo et al (1997) refer specifically to "gravitationally
induced starburst"? I could not find the actual reference.
29.- p16: "...transfered..." => transferred
30.- p17: "...They do not exist in present-day clusters and, therefore,
must either be destroyed or made undetectable...". See
Koo et al (1997) and comments 16 and 24 above, for another
plausible scenario in which some of the Butcher-Oemler galaxies
may be low-mass starbursts which will evolve into the present
population of cluster spheroidals.
31.- p17: "...present-day cluster..." => clusters
32.- p18: "...to also be..." => are also
"...dark matter dominant..." => dark matter dominated
33.- p19: "...Strogren..." => Stromgren
"...population know..." => known
"...primarily..." => primary
34.- p20: "...The results is..." => result
"...suspectable..." => susceptible
35.- References: Update all astro-ph references. Also, I could not
find Dressler et al (1994), Huchra and Burg, and Whitmore et
al (1993) as referenced in the text.