|
VISUAL, RED AND INFRARED PHOTOGRAPHIC SURFACE
PHOTOMETRY OF NGC 55 AND NGC 253
By
GREGORY L. FITZGIBBONS
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1990
UNIVEtRSITY OF FLORIDA LIBRARIES
This work is dedicated to my mother and father, Virginia
R. Fitzgibbons and William E. Fitzgibbons, who believed for so
long.
ACKNOWLEDGMENTS
This work could not have been done without the help of
the many individuals who contributed time, resources and
support. I extend heartfelt thanks to my committee members,
Drs. Stephen T. Gottesman, John P. Oliver, Pierre Ramond, Alex
G. Smith and Haywood C. Smith, for their work. In particular,
my gratitude goes to Steve for his interest and suggestions,
especially in the early stages of the project, and to Alex for
providing me support as a research assistant.
Drs. Daniel Klinglesmith III and Christopher Harvel
taught me the rudiments of microdensitometry and answered many
of my questions in the beginning of my work. I am thankful
for their time and contributions.
The generous financial and material assistance of Kitt
Peak National Observatory and Cerro Tololo Inter-American
Observatory, and their former directors, Drs. Geoffrey
Burbidge and Patrick Osmer, is much appreciated. The support
of the National Science Foundation through grant 80LA/R-5 is
gratefully acknowledged, as is support from the Department of
Sponsored Research of the University of Florida.
Many people at Cerro El Roble gave much of their time to
make this project a success, Drs. Jorge May, Carlos Torres,
Jos6 Maza, and especially Edgardo Costa, who took many of the
iii
plates. His help was indispensable in obtaining the necessary
photographic images, in particular the troublesome IV Ns.
Sefor Gonzalez also is to be thanked for his telescope work.
Appreciation is extended to the U. S. Naval Observatory
Time Service Alternate Station at Miami for use of their
microdensitometer, and to the former directors, Don Monger and
Jim Martin. The program SCFTZ which operated the microdensi-
tometer was written by Jim and his help was always kindly
given.
I am grateful to William Pence for his comments regarding
the "spur" of NGC 253.
Much of the data reduction was performed at the North-
east Regional Data Center (NERDC) and their assistance is
appreciated.
My sincere thanks goes to my fellow graduate students and
friends for the good times we enjoyed, and especially to Joe
Pollock, Glenn Schneider and Roger Ball, who shared so many
times, good and bad, with me.
Finally, I wish to thank the three people who gave so
much of themselves toward helping me along: my mother,
Virginia Fitzgibbons, my father, William Fitzgibbons, and my
wife, Berna Lowenstein. Their patience and encouragement are
deeply appreciated, and I know they are as happy as I am to
see this labor come to a successful conclusion.
TABLE OF CONTENTS
PAGE
ACKNOWLEDGMENTS .
LIST OF TABLES .
LIST OF FIGURES. .
ABSTRACT . .
CHAPTERS
I. INTRODUCTION .
Objective. .
Previous Studies .
NGC 55 .
NGC 253 .
Adopted Distances.
II. PHOTOGRAPHIC MATERIAL AND REDUCTION
METHODS . .
Observations . .
The Telescope . .
Plates and Filters .
Photographic Techniques .
Calibration . .
Plate Digitization . .
Equipment . .
Technique . .
Data Reduction . .
Numerical Mapping Technique and
Reduction of the Outer Field
Reduction of the Inner Field. .
Density to Intensity Conversion .
Star and Blemish Removal. .
Photoelectric Calibration .
Plate Addition . .
Image Smoothing . .
III. PHOTOMETRIC DATA FOR NGC 55. .
General Description. . .
. . iii
. . vii
xiii
. . 1
. . 1
. . 4
. . 4
. . 8
. . 16
16
18
18
18
18
22
24
25
25
26
29
29
36
37
38
41
49
50
52
52
PAGE
Isophotal Contour Maps . .. 54
Luminosity Profiles. . ... 63
Major Axis . .. 63
Minor Axis . .. 69
Size of NGC 55 . .. 71
Asymmetry Profiles . .. 72
Major Axis . .. 72
Minor Axis . .. 76
Color Profiles . 76
Major Axis . .. 80
Minor Axis . .. 84
Integrated Parameters and Colors ... 88
IV. PHOTOMETRIC DATA FOR NGC 253 .. 106
General Description. . .. 106
Isophotal Contour Maps . .. 108
Luminosity Profiles . ... 118
Major Axis . .. 118
Minor Axis. . .. 123
Size of NGC 253 . .. 123
Mean Axis Profiles . ... 124
Ellipticity . ... 132
Color Profiles . ... 136
Major Axis. . .. 136
Minor Axis . .. 141
Integrated Parameters and Colors ... 145
Decomposition of the Observed
Profiles. . .. 160
V. SUMMARY AND CONCLUSIONS. .. 172
REFERENCES . . 180
BIOGRAPHICAL SKETCH. . .. 185
LIST OF TABLES
TABLE
1. DISTANCE TO NGC 55 .
2. DISTANCE TO NGC 253. .
3. PLATE MATERIAL NGC 55. .
4. PLATE MATERIAL NGC 253 .
5. APERTURE PHOTOMETRY OF NGC 55. .
6. NGC 55 PLATE CALIBRATION .
7. APERTURE PHOTOMETRY OF NGC 253
8. NGC 253 PLATE CALIBRATION .
9. ELEMENTS OF NGC 55 .
10. NGC 55 MAJOR AXIS LUMINOSITY
GRADIENTS . .
11. NGC 55 MAJOR AXIS SCALE LENGTHS
IN KPC . .
12. NGC 55 MINOR AXIS LUMINOSITY
GRADIENTS . .
13. NGC 55 MINOR AXIS SCALE LENGTHS
IN KPC . .
14. SIZE OF NGC 55 . .
15. MEAN V LUMINOSITY DISTRIBUTION
IN NGC 55 . .
16. MEAN R LUMINOSITY DISTRIBUTION
IN NGC 55 . .
17. MEAN I LUMINOSITY DISTRIBUTION
IN NGC 55 . .
18. PHOTOMETRIC PARAMETERS OF NGC 55 V
vii
PAGE
. 16
. 17
. 20
. 21
. 42
. 44
. 45
. 47
. 54
. 68
. 69
. 70
. 70
. 71
. 93
. 94
. 95
103
TABLE
19. PHOTOMETRIC PARAMETERS OF NGC 55 R
20. PHOTOMETRIC PARAMETERS OF NGC 55 I .
21. ELEMENTS OF NGC 253 .
22. SIZE OF NGC 253 . .
23. MEAN V LUMINOSITY DISTRIBUTION
IN NGC 253 . .
24. MEAN R LUMINOSITY DISTRIBUTION
IN NGC 253 . .
25. MEAN I LUMINOSITY DISTRIBUTION
IN NGC 253 . .
26. PHOTOMETRIC PARAMETERS OF NGC 253 V.
27. PHOTOMETRIC PARAMETERS OF NGC 253 R.
28. PHOTOMETRIC PARAMETERS OF NGC 253 I.
PAGE
. 104
. 105
. 108
. 124
. 150
. 151
. 152
. 161
. 162
. 163
viii
LIST OF FIGURES
FIGURE
1.
PAGE
. 35
Density contour map of plate 9309.
2. Photograph of NGC 55 . .. 53
3. Short exposure V isophote map
of NGC 55 . 55
4. Short exposure R isophote map
of NGC 55 . 56
5. Short exposure I isophote map
of NGC 55 . 57
6. Visual isophote map made from
added long exposure plates .... 59
7. Red isophote map made from
added long exposure plates ... 60
8. Infrared isophote map made from
added long exposure plates .... 61
9. Visual major and minor axis
profiles of NGC 55 ... 64
10. Red major and minor axis profiles
of NGC 55 . 65
11. Infrared major and minor axis
profiles of NGC 55 ... 66
12. Asymmetry profile for NGC 55
V data: major axis ... 73
13. Asymmetry profile for NGC 55
R data: major axis ... 74
14. Asymmetry profile for NGC 55
I data: major axis. .. 75
15. Asymmetry profile for NGC 55
V data: minor axis ... 77
ix
FIGURE
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
Asymmetry profile for NGC 55
R data: minor axis. .
Asymmetry profile for NGC 55
I data: minor axis. .
Major axis (V R) color profile
for NGC 55 .
Major axis (R I) color profile
for NGC 55 .
Major axis (V I) color profile
for NGC 55 .
Minor axis (V R) color profile
for NGC 55 .
Minor axis (R I) color profile
for NGC 55 .
Minor axis (V I) color profile
for NGC 55 .
Equivalent mean V luminosity
profile for NGC 55. .
Equivalent mean R luminosity
profile for NGC 55. .
Equivalent mean I luminosity
profile for NGC 55. .
Relative V integrated luminosity
curves for NGC 55 .
Relative R integrated luminosity
curves for NGC 55 .
Relative I integrated luminosity
curves for NGC 55 .
Photograph of NGC 253. .
Short exposure V isophote map
of NGC 253 .
Short exposure R isophote map
of NGC 253 .
x
PAGE
. 78
. 79
. 81
. 82
. 83
. 85
. 86
. 87
. 89
. 90
. 91
. 97
. 98
. 99
. 107
. 109
. 110
PAGE
FIGURE
33.
34.
35.
36.
37.
Red major and minor axis
profiles of NGC 253
Infrared major and minor
profiles of NGC 253
Mean V semi-major axis
profile for NGC 253
Mean R semi-major axis
profile for NGC 253
Mean I semi-major axis
profii3 for NGC 253
Mean V semi-minor axis
profile for NGC 253
Mean R semi-minor axis
profile for NGC 253
Mean I semi-minor axis
profile for NGC 253
Ellipticity curve for NGC
Ellipticity curve for NGC
Short exposure I isophote map
of NGC 253 .
Visual isophote map made from
added long exposure plates.
Red isophote map made from the
long exposure plate 5394.
Infrared isophote map made from
added long exposure plates.
Visual major and minor axis
profiles of NGC 253 .
axis
. 126
. 127
. 128
. 129
. 130
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
Ellipticity curve for NGC 253 I data
Major axis (V R) color profile
for NGC 253 . .
. 131
. 133
. 134
. 135
. 137
111
114
115
116
119
120
121
253 V data
253 R data
FIGURE
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
Major axis (R I) color profile
for NGC 253 . .
Major axis (V I) color profile
for NGC 253 . .
Minor axis (V R) color profile
for NGC 253 . .
Minor axis (R I) color profile
for NGC 253 . .
Minor axis (V I) color profile
for NGC 253 . .
Equivalent mean V luminosity
profile for NGC 253 .
Equivalent mean R luminosity
profile for NGC 253 .
Equivalent mean I luminosity
profile for NGC 253 .
Relative V integrated luminosity
curves for NGC 253. .
Relative R integrated luminosity
curves for NGr 253. .
Relative I integrated luminosity
curves for NGC 253. .
Decomposition of the V mean minor
axis . .
Decomposition of the V mean major
axis. . .
Decomposition of the I mean minor
axis . .
Decomposition of the I mean major
axis . .
xii
PAGE
. 138
. 139
. 142
. 143
. 144
. 146
. 147
. 148
. 154
. 155
. 156
. 167
. 168
. 169
. 170
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
VISUAL, RED AND INFRARED PHOTOGRAPHIC SURFACE
PHOTOMETRY OF NGC 55 AND NGC 253
By
Gregory L. Fitzgibbons
August 1990
Chairman: Stephen T. Gottesman
Major Department: Astronomy
Detailed photographic surface photometry is presented for
the Sculptor Group galaxies NGC 55 and NGC 253. Observations
in visual, red and infrared light from long and short exposure
plates are used to produce isophote maps as well as major and
minor axis profiles. The standard photometric parameters are
derived and V-R, R-I and V-I colors are presented for the
major and minor axes.
The integrated magnitude of NGC 55 is: V, = 8.17, R, =
7.94 and IT = 7.30. A thickening of the disk is detected that
increases with increasing wavelength, indicating the presence
of a faint, red halo. At the detection limits of V = 27.6
magnitude per square arc second, R = 27.8 magnitude per square
arc second, and I = 27.0 magnitude per square arc second, the
size of NGC 55 is 35'.9 x 9'.5 (18.6 kpc x 4.9 kpc), 42'.6 x
10'. 2 (22.1 kpc x 5.3 kpc) and 35'.8 x 12'.8 (18.5 kpc x 6.6
xiii
kpc) at the assumed distance of 1.78 Mpc. The asymmetry of
the galaxy is discussed.
The integrated magnitude of NGC 253 is: V, = 7.35, R, =
6.41 and I, = 5.75. The disk appears to thicken with
increasing wavelength, again indicating the presence of a red
halo. At limits of V = 26.9 magnitude per square arc second,
R = 25.9 magnitude per square arc second, and I = 26.4
magnitude per square arc second, the size of NGC 253 is 36'.3
x 16'.4 (30.1 kpc x 13.6 kpc), 35'.0 x 12'.3 (29.0 x 10.2
kpc), and 42'.8 x 24'.2 (35.5 kpc x 20.1 kpc) at the assumed
distance of 2.85 Mpc. A large spur is detected to the south
that is most obvious in the infrared image. It increases the
size to 42'.8 x 31'.4.
A three component model of the visual and infrared
luminosity of NGC 253 is discussed. The relative contri-
butions of a r0.25 spheroid, an exponential disk, and a spiral
arm component are determined.
xiv
CHAPTER I
INTRODUCTION
Objective
The last ten years have witnessed an almost explosive
growth in surface photometry of galaxies. Photographic
plates, photoelectric photometers, and solid state charge-
coupled devices (CCDs) have all been used as detectors. And
yet, the luminosity, luminosity profiles, colors, and size are
unmeasured quantities for most bright galaxies. In addition,
an examination of published surface photometry (Davoust and
Pence 1982; Pence and Davoust 1985) shows that most of the
published observational material is in the B and V passbands,
although a trend toward longer wavelengths can be discerned.
It is desirable to supplement the blue and yellow data with
longer wavelength measurements for a variety of reasons: (1)
red and infrared passbands allow better penetration of
obscuring dust (Elmegreen 1981), (2) an underlying old stellar
component may be revealed (Schweizer 1976), (3) a halo or
extended spheroidal component may be detected (see, e.g.,
Spinrad et al. 1978), and (4) color indices such as (V-I) can
give information regarding the stellar constituents of a
galaxy and be of particular value when given as a function of
position.
2
Even for nearby galaxies much work remains to be done
before we can hope to understand their structure and
evolution. Their closeness causes them to have large angular
sizes, and this makes CCD or photoelectric photometry
difficult because these techniques are currently better suited
to making brightness measurements of objects of small angular
extent. On the other hand, photographic surface photometry
can be done for almost any galaxy accessible to a CCD, and
especially nearby objects because of the availability of wide
field cameras. Also, the use of hypersensitizing techniques
(Schoening 1978; Smith and Hoag 1979) allows the red and near
infrared spectral region (600 900 nm) to be photographed
with reasonably short exposure times. Thus, it is possible to
conduct a multicolor investigation of the structure of
galaxies with a particular emphasis on those longer
wavelengths which trace the cooler stellar component that
comprises the bulk of the luminous matter.
In this work we present detailed photographic surface
photometry of the nearby galaxies NGC 55 and NGC 253, in V, R,
and I passbands. The main objective is to provide a set of
photometric data that can be used to determine the properties
of the disk and halo components that comprise these galaxies,
and measure their contributions to the overall luminosity.
The result is a more complete picture of the structure of
these galaxies, which will in turn improve our understanding
3
of other galaxies. A photograph of each galaxy appears in
Chapter III and Chapter IV.
The observational material and reduction methods are
discussed in Chapter II. In Chapter III the standard photo-
metric parameters, as described by de Vaucouleurs (1962), are
determined for NGC 55. The size and shape of this galaxy are
discussed and compared to earlier work. Color indices and
integrated magnitudes are presented and corrections for
absorption are made. In Chapter IV, the standard photometric
parameters are determined for NGC 253. As for NGC 55, the
size, shape and colors are discussed and compared to previous
studies, and integrated magnitudes and colors are presented
and corrected for absorption. In addition, the luminosity
profiles are decomposed into a disk, spheroid and spiral arm
model. The results are summarized in Chapter V.
The choice of these two galaxies was based on several
criteria: the galaxies should be free from interacting
neighbors, they should be viewed at high inclinations so their
halos can be studied, previous B band photometry should exist
for comparison purposes, and kinematic data in the form of
rotation curves should be available. The first criterion was
the most difficult to satisfy. Both galaxies are members of
the Sculptor group, a loose collection of galaxies about 2.4
Mpc distant, with NGC 247, NGC 300 and NGC 7793 as additional
members (Puche and Carignan 1988). Neither NGC 55 nor NGC 253
has a close companion. Hummel, Dettmar and Wielebinski (1986)
4
place NGC 300 and an anonymous dwarf galaxy at projected
distances of 320 and 180 kpc, respectively, from NGC 55.
Puche and Carignan put the dwarf more than one Mpc away, but
have NGC 253 and NGC 247 about 250 kpc apart. Lewis (1969)
found displaced HI centroids in NGC 247 and NGC 253 and took
this to indicate that these galaxies form a "stable pair."
And Pritchet et al. (1987) in a study of the mean magnitudes
of carbon stars in NGC 55 derived a distance that supports it
forming a pair with NGC 300. Based on optical appearances,
there are no compelling reasons to believe recent interactions
have occurred. However, NGC 253 does show an optical "tail"
or "spur" (Beck, Hutschenreiter and Wielebinski 1982) which is
also detected in radio continuum observations (Beck et al.
1979), and it shows evidence of recent starburst activity.
Consequently, the possibility of past interactions cannot be
ruled out.
Previous Studies
NGC 55
Photographic surface photometry of NGC 55 was first done
by de Vaucouleurs (1961) in the photographic (pg) passband (a
Kodak 103a-0 plate and no filter). He found the maximum
dimensions to be about 10 x 0.20, but reliable measurements
stopped at 45' x 9', corresponding to about 26 magnitude per
square arc second, hereafter written as p,. = 26. The total
integrated magnitude is given as 7.9, which, for his assumed
distance of 2.3 Mpc, corresponds to an absolute magnitude of
5
M, = -19.1. Attention is called to the pronounced asymmetry
present in the galaxy's appearance. He found the luminosity
distribution to be approximately exponential in the outer
regions, but with the gradient increasing near the nucleus,
and the west side showing the largest change. Based on the
asymmetrical structure of NGC 55, de Vaucouleurs proposed that
it is similar to the Large Magellanic Cloud (LMC) and is seen
looking down the bar. Higher resolution surface photometry in
blue light is given by Sersic (1968) but it does not go to as
faint a level as that of de Vaucouleurs.
Spectroscopic observations of HII regions by de
Vaucouleurs yielded three velocity measurements which he used
to calculate a mass of 2 4 x 1010 solar masses (M) and a
M/L of 3.3 6.6 M/L. (L, = 1 solar luminosity). The
inclination was assumed to be 900 (edge-on) and the position
angle of the major axis was taken to be 1050.
The rotation velocities of the inner region of NGC 55
were measured by Seielstad and Whiteoak (1965) using the 21 cm
emission line of HI and an assumed inclination of 79.
Robinson and van Damme (1966), also observing HI, measured
velocities out to about 40' on the southeast side of the
galaxy, possibly indicating a turnover in the velocity curve.
The inclination was assumed to be 900. Both sets of
measurements show the center of rotational symmetry displaced
about 2.5' southeast of the optical nucleus. For an adopted
distance of 1.74 Mpc, Robinson and van Damme found the total
6
neutral hydrogen mass M., = 2 x 109 M., the total mass of NGC
55 to be 2.5 x 1010 M,, and a M/L = 7.2.
De Vaucouleurs (1981), with more extensive optical data
than previously used, mapped the velocity field and found
curious distortions of the velocity contours in the disk,
indicative of motion more complicated than purely circular
rotation.
Observing at 843 MHz, Harnett and Reynolds (1985) found
the peak emission offset 1.3' southeast of the nucleus. The
scale height of the radio continuum emission was 35" 7".
Using the surface photometry of de Vaucouleurs (1961),
and the rotation data of Seielstad and Whiteoak (1965) and
that of Robinson and van Damme (1966), Comte (1985) calculated
a simple two-component mass model of a constant M/L disk and
a spherical halo. For a disk M/L = 1, the derived total mass
is 3.9 x 1010 M. and the ratio of disk mass to halo mass is
0.03.
Hummel, Dettmar and Wielebinski (1986) observed NGC 55 at
6 cm and 21 cm. For an assumed distance of 2.0 Mpc, they
derived a total HI mass of 1.5 x 10' M, and a total galaxy mass
of 2 x 101 M. The HI extent is comparable to the optical
dimensions and is not centered on the bar. Also, the HI
distribution looks disturbed. The continuum emission is
mainly concentrated on the bar and dominated by a slightly
offset triple source, strongly suggesting ongoing star
formation. They, too, find an asymmetry between the mass and
7
light distributions. Almost 80% of the total mass resides in
the disk, the remainder in the off-centered bar component.
The centers of the two components are 3.5' apart. About 30%
of the blue light is in the bar region. The centroids of
total light and neutral hydrogen emission are 2.7' and 2.1'
southeast of the bar, respectively. The continuum emission is
also displaced about 1' southeast of the bar. They find an
inclination of about 800, a position angle of 1090 from their
HI data, and a M/L = 3.4. Additional continuum information is
available from Condon (1987) in the form of a map of 1.49 Ghz
observations.
From HI studies of the Sculptor group galaxies, Puche and
Carignan (1988) calculate the dynamical M/L for the group to
be 83 10. For NGC 55 they find M/L = 3.6.
Graham (1982) was able to resolve old red giants in NGC
55 and placed an upper limit of 1.9 Mpc on its distance.
Da Costa and Graham (1982) found three globular clusters,
including a young cluster similar to those found in the LMC,
and, assuming the distance given above, found an absolute
magnitude of -9.1. This is comparable to the young clusters
NGC 1818, NGC 1866 and NGC 2004 in the LMC. A distance
modulus of 26.57 (= 2.06 Mpc) is given by de Vaucouleurs
(1986) based on the brightest blue stars, the brightest
superassociation, and the brightest cluster. A much smaller
distance of 1.34 0.08 Mpc was derived by Pritchet et al.
8
(1987) using the mean magnitude of carbon stars found in NGC
55.
The apparent distribution of stars perpendicular to the
major axis was investigated by Kiszkurno-Koziej (1988). A
difference between the north and south sides led to the
conclusion that the inclination is between 80.80 and 81.70,
with the north side being closer to us. For a distance of 1.5
Mpc, Kiszkurno-Koziej determined the scale height of the young
disk to be between 0.10 and 0.14 kpc, and scale height of the
old disk to be between 0.18 and 0.21 kpc.
NGC 253
Unlike the case of NGC 55, an extensive body of
literature exists for NGC 253. Pence (1978) gives a listing
of publications through 1978. Here, rather than referencing
all published papers, only those which have some bearing to
the photometry or kinematics of NGC 253 will be mentioned.
Attention will be confined for the most part to results
published after 1978; the interested reader is referred to
Pence's listing for earlier material.
Sersic (1968) presented photographic photometry in the B
band of the bright inner region of NGC 253. Much deeper
photometry was done by Freeman, Carrick and Craft (1975) in
their photographic search for a corona to the galaxy. For an
assumed distance of 3.0 Mpc they found no clear evidence for
a corona at radii greater than 15 kpc along the minor axis,
down to g. = 29.2. Pence (1978) did detailed photographic
9
photometry of NGC 253 down to g. = 29 and photoelectric drift
scans in U, B and V. The maximum detected dimensions were 60'
x 26'. The standard photometric parameters were derived,
including a total magnitude BT = 8.05 and a total luminosity
LT = 2.65 x 1010 L, for an assumed distance of 2.5 Mpc. A
three-component luminosity model consisting of spheroid, disk
and spiral arm components was derived from the luminosity
profiles. Spinrad et al. (1978) looked for a halo in red (r)
and near infrared (i) light using photographic and photo-
electric photometry. Their profiles indicate an extensive,
faint spheroidal component that rapidly fades in brightness
with increasing radius, and has an almost constant r i =
0.7. The total spheroidal luminosity was found to be greater
than or comparable to the disk luminosity. In a qualitative
study of the optical halo around NGC 253, Beck, Hutschenreiter
and Wielebinski (1982) examined enhanced images of deep plates
taken in blue light, along with red and near infrared images.
They found an extended halo that was more pronounced on the
reddest plates, and that roughly coincided with the radio spur
detected earlier by Beck et al. (1979). The minor axis could
be followed out 10' from the nucleus and they concluded, in
agreement with Spinrad et al. that any color gradient, if
present, was small.
Telesco and Harper (1980) examined the infrared emission
of NGC 253 between 1 and 300 pm over the inner 30" and
calculated an integrated luminosity of 2.8 x 1010 L,.
10
Photographic and photoelectric observations in the near
infrared were published by Uyama, Matsumoto and Thomas (1984).
The central regions of NGC 253 were examined at 0.7 0.9 pm,
1.25 pm and 2.2 pm. The color index (0.8 pm 2.2 pm) of the
disk ranged from 2.0 to 2.8, corresponding to colors of MO III
and M4 III stars. They found a color gradient increasing
toward the galaxy's center where the visual absorption also
increases. The inner disk was also observed in the near
infrared in the J (1.27 pm), H (1.65 pm) and K (2.2 pm) bands
and at the 2.6 mm line of CO emission by Scoville et al.
(1985). Their major axis profiles show two components: a
nuclear peak of diameter = 20" and an extended, inner disk of
diameter = 360". The 2.2 Km distribution is dominated by a
barlike feature at position angle 170 east of the major axis
and extending 120" on either side of the nucleus. The nucleus
itself is extremely red with respect to the disk. Both the
nucleus and the inner disk show excess 2.2 pm radiation,
probably caused by hot dust.
Young et al. (1989) used IRAS data to derive the global
properties of NGC 253 including infrared luminosities at 12,
25, 60, and 100 pm, and total masses of gas and dust. For an
assumed distance of 3.4 Mpc, the infrared luminosity (1 500
pm) is log L,, = 10.42 L,, the total dust mass is log Md,.t =
6.54 M., log M1. = 9.65 Me, and log M,2 = 9.37 MH. The
integrated fluxes at 12, 25, 60, and 100 pm are, respectively,
62.04, 147.34, 1157.2, and 1760.2 Jy.
11
The neutral hydrogen in NGC 253 was mapped by Huchtmeier
(1972, 1975) who found a total HI mass of 4.1 x 109 M, for an
assumed distance of 3 Mpc. Combes, Gottesman and Weliachew
(1977), unlike Huchtmeier, found a centrally peaked HI
distribution and a total HI mass of 4.9 x 109 M., using an
assumed distance of 3.4 Mpc. Whiteoak and Gardner (1977),
assuming the same distance, found a total HI mass of
3.8 x 109 M,.
Radio continuum observations made at 8.7 GHz by Beck et
al. (1979) detected emission 6' south and 5' north of the
plane of the galaxy. They found the emission to be nearly
constant between 2.5 and 4.5 kpc from the plane for an assumed
distance of 3.4 Mpc and detected the "spur" extending toward
the southwest, as mentioned earlier. Hummel, Smith and van
der Hulst (1984) made continuum observations at 4.9 GHz. They
divided the emission into three main components: a central
source (= 50% of the total emission), an inner disk containing
a bar (= 30%) and an outer disk (= 20%). The inner disk has
a half power diameter of = 5 kpc (for an assumed distance of
2.2 Mpc) and has a ridge of radio emission coinciding with the
optical bar. The outer disk has a total extent of = 15 kpc.
There is some suggestion of the presence of a radio spiral
arm. Other continuum observations of NGC 253 were made by
Harnett and Reynolds (1985), at 843 MHz, and Condon (1983,
1987) at 1.465 and 1.49 GHz.
12
Fabbiano (1988) found the inner disk of NGC 253 to be
quite bright in X-rays (0.2 10.2 kev), and noted extended X-
ray emission with no optical or radio counterpart in the
northern side. The major axis X-ray surface brightness
profile can be traced out to = 10' and is steeper than the
optical (blue) profile. The minor axis profile can be traced
out to = 5' toward the southeast and = 10' toward the
northwest. For an assumed distance of 3.4 Mpc the X-ray
luminosity L. = 1.1 x 1040 erg/sec and 2.5 x 1040 erg/sec
corresponding to the 0.2 3.5 key and 1.2 10.2 key ranges,
respectively.
The large infrared flux from the nucleus of NGC 253 has
led various investigators to classify this galaxy as a
starburst galaxy (Rieke and Low 1975; Rieke et al. 1980; Donas
and Deharveng 1984; Rieke, Lebofsky and Walker 1988).
Examination of the nuclear region at 1.413 and 4.885 GHz
(Condon et al. 1982) and at 1.490 and 4.860 GHz (Antonucci and
Ulvestad 1988) showed numerous bright, compact sources,
believed to be associated with the starburst activity. Turner
and Ho (1983), observing at 2 and 6 cm, confirmed the presence
of large numbers of HII regions and young stars. They
proposed the existence of 4.7 x 104 OB stars in the nucleus to
account for the observed flux. Hummel, Smith and van der
Hulst (1984) measured the power at 4.9 GHz to be about 1021
w/Hz in a beam half power size of approximately 0.2' x 0.1'.
From a compilation of measurements extending from 85.6 MHz to
13
10.7 GHz they calculated a spectral index of a = -0.43 0.07
for the nuclear region, which they attribute to a starburst.
Keel (1984), in a spectrophotometric study of the nuclei of
strong radio sources, found evidence of both nonthermal
activity and recent star formation in NGC 253. He proposed it
as a LINER (Low Ionization Narrow Emission-line Region), a low
ionization nucleus with a weak Seyfert core. Scoville et al.
(1985) calculated a minimum star formation rate of OBA stars
to be 2.3 M,/yr assuming the observed far-infrared luminosity
within a radius of 0.5 kpc is produced by young stars. And
McCarty, Heckman and van Breugel (1987) found spectroscopic
evidence for large-scale winds near the center of NGC 253,
probably from starburst activity, and extensive regions of
ionized gas along the minor axis. They also noted that the
high latitude ionized gas southeast of the nucleus appears to
coexist with the diffuse soft X-ray emission.
Numerous kinematic studies have been made of NGC 253.
Burbidge, Burbidge and Prendergast (1962) obtained a rotation
curve out to 290" from optical spectra and noted non-circular
motions in the nuclear region. Demoulin and Burbidge (1970)
also made spectroscopic observations of the central region but
with a higher dispersion and found material blue-shifted up to
120 km/sec. Turbulent velocities were noted southeast of the
nucleus, up to 150 km/sec. They proposed a violent event in
the center some 1 2 x 107 years ago, ejecting a cone or
column of gas. Huchtmeier (1975) published HI rotation
14
measurements out to 17.5'. The maximum velocity was 205
km/sec at a radius of 8' from which he derived a total mass of
1.4 x 1011 Me using a Brandt curve and a distance of 3 Mpc.
Gottesman et al. (1976) observed the central region of NGC 253
in the 21 cm absorption line and found an excess of blue-
shifted gas, indicating an outflow. Combes, Gottesman and
Weliachew (1977) also found evidence for noncircular motions
in the central region and noted faster rotation of HII than HI
about 4.5' northeast of the center. Using optical spectra,
Ulrich (1978) studied the gas flowing out of the nucleus and
put it at 0.01 Me/yr. She hypothesized a population of 103 06
stars could account for the flow. Pence (1978) used 25 Ha
interferograms of the entire bright disk to measure
velocities, extending from the nucleus out to a maximum
distance of 650" and derived a total mass of 1.42 x 101 M,, in
agreement with Huchtmeier. He also noted noncircular motions
in excess of 50 km/sec in several areas, and pointed out that
the streaming motions within 200" of the nucleus are similar
to that predicted by a simple kinematic model of gas flow near
a bar. Scoville et al. (1985), assuming a distance of 3.4
Mpc, estimated the total mass of the H2 inside a radius of 4
kpc to be approximately 2 x 109 Me, 7% of the dynamical mass
within the same radius. A plume extending northward about 1.5
kpc from the nucleus (for an assumed distance of 3.1 Mpc) and
moving at 168 km/sec was detected by Turner (1985) observing
the OH line at 1667 MHz, and may be related to the starburst
15
nature of this galaxy. Finally, Canzian, Mundy and Scoville
(1988) observed the molecular gas in the bar of NGC 253 in the
CO emission line. They found it to have an angular extent of
39" x 12" at position angle 640 and, at an assumed distance of
3.4 Mpc, a mass in molecular clouds of 4.8 x 108 M., 0.4 times
the dynamical mass of the region; hence, the gas is a
significant contributor to the potential. The gas appears to
rotate as a rigid body with a steep rotation curve, although
the gradient is not as steep as that found by Gottesman et al.
Canzian and co-workers concluded that models of barred spirals
may not apply to galaxies with so much gas.
Lewis and Robinson (1973) give the total mass of the
Sculptor group of galaxies as (4.4 1.5 p.e.) x 10" Mg,
almost all of which was determined from 21 cm rotation curves.
Their virial mass, excluding NGC 24 and NGC 45, is
6.1 x 1012 M.. From HI studies of the Sculptor group, Puche
and Carignan (1988) find the total mass to total blue
luminosity, M/L,, for NGC 253 to be 12.0, and for the
dynamical M/L, for the group, 83 10 M,/L,.
There have been two recent determinations of the distance
to NGC 253, both using different indicators. Blecha (1986),
in a search for globular clusters, detected 25 reliable
candidates within 13' of the center, with a r025 spatial
distribution. Matching the luminosity function to our
galaxy's globulars yielded a distance of 2.88 0.5 Mpc.
De Vaucouleurs (1986) gives a distance modulus of 27.42
16
(= 3.05 Mpc) based on the HI line width (B and H bands) Tully-
Fisher relation. Pence (1978) gives eight previous
determinations of the distance to NGC 253 or the Sculptor
group, the mean of which is 2.9 0.67 Mpc.
Adopted Distances
Recent distance determinations for NGC 55 are listed in
Table 1. Sandage and Tammann used blue and red supergiants;
Graham used old red giants; de Vaucouleurs used the brightest
blue stars, super associations and the brightest clusters; and
Pritchet et al. used carbon stars. The value from Puche and
Carignan is the mean of two values they list based on tertiary
indicators (de Vaucouleurs 1979a,b) which they derived using
the relations given by de Vaucouleurs and Davoust (1980) and
Bottinelli et al. (1983). Giving a weight of two to the value
of Puche and Carignan yields r mean distance of 1.78 0.34
Mpc which will be adopted for this work. At this distance,
one arc second is 8.6 pc; one arc minute is 518 pc. This
corresponds to 0.85 kpc/mm on the photographic plates used in
this study.
TABLE 1
DISTANCE TO NGC 55
Source Distance (Mpc)
Sandage and Tammann (1981) 2.3
Graham (1982) 1.9
de Vaucouleurs (1986) 2.06
Pritchet et al. (1987) 1.34
Puche and Carignan (1988) 1.56
17
Table 2 lists recent distance determinations for NGC 253.
TABLE 2
DISTANCE TO NGC 253
Source Distance (Mpc)
Sandage and Tammann (1981) 4.0
Blecha (1986) 2.88
de Vaucouleurs (1986) 3.05
Puche and Carignan (1988) 2.59
The Sandage and Tammann value is again based on blue and red
supergiants. The values of Blecha and de Vaucouleurs are
based on globular clusters and the Tully-Fisher relation and
have already been described. Four values listed by Puche and
Carignan, based on tertiary indicators, were combined and the
mean used. As for their NGC 55 values, these are based on the
work of de Vaucouleurs, de Vaucouleurs and Davoust, and
Bottinelli et al. A weight of four is given to this value of
the distance. The resulting mean value of all the distances,
2.85 0.54 Mpc will be adopted for this work. At this
distance, one arc second is 13.8 pc and one arc minute is 829
pc. This corresponds to 1.35 kpc/mm on the photographic
plates used in this study.
CHAPTER II
PHOTOGRAPHIC MATERIAL AND REDUCTION METHODS
Observations
The Telescope
Photographic surface photometry of NGC 55 and NGC 253,
which have angular sizes of more than one-half degree,
required the use of a telescope with a field of view of at
least several degrees. The 70/100 cm Maksutov Astrograph at
Cerro el Roble (Capilla de Caleu), operated by the University
of Chile, was used for all the observations. The telescope
had a 210 cm focal length and a plate scale of 98.2"/mm. The
field of view was 25 square degrees on 18 cm x 18 cm plates.
From geometrical considerations (Dawe 1984) the unvignetted
field had a radius of about 20 and therefore did not affect
the galaxy images. Located at latitude -32 58' 54", west
longitude 710 01' 10.5", and an elevation of 2220 meters, the
observatory was in the coastal range of mountains. The site
had good seeing (typically one arc second) and moderately dark
skies.
Plates and Filters
Observations were begun in late 1980 and continued
through late 1984. Of the many plates taken during this
period, those judged to be the best by visual examination were
19
selected for analysis. The plates used and other relevant
information are listed in Table 3 for NGC 55 and in Table 4
for NGC 253. The passbands used were defined as follows: the
visual (V) was a Kodak IIa-D emulsion behind a 2mm Schott GG
495 filter; the red (R) was a Kodak IIIa-F emulsion behind a
2mm Schott RG 610 filter; and the infrared (I) was a Kodak IV-
N emulsion behind a 2mm Schott RG 695 filter. The wavelength
intervals of each passband were approximately 500 650 nm,
610 700 nm, and 700 900 nm. The effective wavelength .,,
was calculated using
'S(A)IAdAL
Xeff S (1)
where S(X) is the product of the emulsion sensitivity and the
filter transmission at each X. For the V, R and I passbands
used here the effective wavelengths were 572.3 nm, 656.0 nm,
and 794.1 nm, the last value being that of Koo (1985). The RG
610 filter was broken in August 1982 and a filter known simply
as the "Russian" filter was substituted for it. This "new"
filter was 1.5 mm thick glass with a transmission curve that
closely matched that of a 2 mm Schott OG 590, although it did
not have as high a transmittance. It was used for the short
exposure plates and the matching of these observations with
those taken with the RG 610 filter is discussed later. The
resulting bandpass was about 590 700 nm and X.,, = 645.9 nm.
a) cr a a\ e a aaa
4)
o) U) CU)U4U)
NNNNN
raf HaHasa H aH
t m m x mUU
H H r-4 H' 4 r-I r-41r-4 W- 1 H
M$ OO lm 4O mtON M ma
4 0)
o Moooo oc a
*- V:0 2 2
M) oo.o0mio iFmmir
in c ??k h a
I 0 fai l l i o
H NHHH H->>>> )
a H H HHHHHH tH
H- H HH
< a "d +,I
S 0 44
Hd o
0 0
i Mc g o4
$ W 9 a
0 4
-4 -A 3 a o
4) rl *8) 4)
4 ,-4 H *v "
co 0l) CC (Id
I I rlr C- (n C4 I
,-Ir-4 Coo4 '41( ) N 0 4) 0.)
H i 0
04
S( ^ u
'**, u r-oooo tn cDP i~ -
*y aOQ aC Q)'< nMN s<'< Q) '*
'" WWW
*^ Um
(la6 m
S$ 21
4& 0000Nonoso00000
0 C"m
I I I I I I I I IQQQQQQQQQ
U -r4 -rl U U U U U U
00ooo C0000000
4"
U) r-l T ll
0 0~ 0
O~dH 333' 4H 0
0
o Z
s 4 ) .4 3 o
0 0000 0 N 0
4 0 04
0S000 o
04 2
w IV I"
22
The choice of plate and filter combinations was largely
determined by the purpose of the investigation. The outer
regions of galaxies are faint and the light received is only
a small fraction of the contribution from the background.
Hence, to enhance detection of these faint features, emulsions
with good contrast and relatively high signal-to-noise ratios
were used (Latham 1978; Furenlid 1978). In addition, noise
introduced during scanning with the microdensitometer, caused
by granularity, was reduced by the choice of fine-grained
emulsions. The filters, with the emulsions, approximate
standard photometric passbands and allow for photoelectric
calibration of the data. Early type stars and dust will be
most evident in the V exposures, HII regions in the R, and
underlying disk stars and old, cool stars will dominate the I
exposures.
Photographic Techniques
Many of the plates were hypersensitized. The IIa-D and
IIIa-F plates were baked in 4% H2 forming gas at 650 C and
stored in dry nitrogen until used. Each IV-N plate was bathed
for four minutes in a 0.001M AgNO3 solution while being
agitated on a Mount Wilson type rocker. This was followed by
hand agitation for two minutes in a bath of isopropanol. The
back of the plate was dried with a paper towel and the plate
was then placed emulsion up and rapidly dried under a stream
of cool air on a rotating turntable, following a procedure
similar to that described by Schoening (1978). Because IV-N
23
plates experience a rapid growth of fog level within hours of
silver nitrate bathing, each plate was used immediately after
being prepared. Also, the isopropanol dissolved the anti-
halation backing. This caused an increase in the sizes of
stellar images arising from light reflecting off the interface
between the emulsion and the glass substrate, but did not
appear to have any other adverse effects. Although hyper-
sensitization caused an increase in fog level and a decrease
in signal-to noise ratio (Hoag, Furenlid and Schoening 1978),
this loss was more than offset by the increase in speed of the
emulsion.
The V and R exposures were done with no moon in the sky.
At least two of the I exposures were taken with the moon above
the horizon (plates 6461 and 9309). This was not a problem
with the IV-N emulsion and RG 695 filter (Schoening 1978)
because the scattered moonlight is primarily smaller
wavelengths than the filter passes. In no instance was the
moon near the object being photographed.
The nucleus of each galaxy was considerably brighter than
the outermost regions; as a result, an exposure deep enough to
record the outer disk and halo overexposed the nucleus. This
necessitated the taking of both long and short exposures of
each galaxy with each emulsion. Each color had at least one
short and one long exposure.
In the beginning the IIIa-F and IV-N plates were
developed for nine minutes in MWP-2 (Diffley 1968) on a Mount
24
Wilson type rocker, but this practice was abandoned when less
than satisfactory development seemed to be the usual result.
Subsequently all plates were hung vertically in a large tank
of D-19 for five minutes with nitrogen burst agitation every
30 seconds. Plate 9281 was in the developer for about seven
minutes because of a timer problem, but suffered no adverse
effects. Following development, the plates were stopped,
fixed, washed and given a rinse in Photo-Flo solution. All
treatments were done at approximately 200 C.
Calibration
All of the plates were calibrated using a sixteen-hole
tube sensitometer described elsewhere (Fitzgibbons 1981). The
light source had a color temperature of about 3100 K. A set
of Schott filters matching those used in the telescope
plateholder was available for the sensitometer. The resulting
set of spots was used to construct the characteristic curve
for each plate.
Errors due to the photometric calibration are discussed
by de Vaucouleurs (1948). These can arise from the
calibration device itself and the exposure conditions, as well
as the history of the emulsion before exposure and during
development. To minimize the calibration errors, whenever
possible the plate in the telescope and the plate in the
sensitometer were cut from the same original 20 cm x 25 cm
plate. Both pieces were hypersensitized together, given
almost simultaneous exposures of the same duration, and
25
developed together immediately afterward. The sensitometer
was kept in a closet on the observing floor so the temperature
and humidity of it and the telescope were closely matched. In
some cases only smaller plates were available, either 18 cm x
18 cm or 4 in x 5 in. For these, plates from the same box
were hypersensitized together with one being used in the
telescope and another in the sensitometer. In all cases, both
the object plate and the calibration plate received identical
exposures and were developed together.
Plate Digitization
Equipment
The plates were digitized with the Boller and Chivens PDS
1010A microdensitometer at the U. S. Naval Observatory Time
Service Alternate Station at Miami, Florida. It was equipped
with a computer-controlled x-y stage that could accommodate
plates up to 25 cm x 25 cm. A program called SCFTZ controlled
the machine and performed a raster scan. The x and y limits
of the scans, scanning speed, distance between samples, and
the spacing in y between scans were entered as input data.
The output densities were recorded on magnetic tape after each
scan row.
At some time the machine caused density measurements in
excess of about 2.5 to become compressed and therefore no
longer linear (J. Martin, private communication). When this
occurred is not known. However, this should not have had an
adverse effect on the measurements since the calibration
26
plates would also have been affected, and few, if any regions
of interest in this study had densities as high as 2.5.
Another, more serious problem was the appearance of bands in
the digitized images of four plates. These are believed to
have been caused by drops in voltage to the microdensitometer
at semi-regular intervals. Later, a regulated power supply
was used to run the microdensitometer and these bands were
either absent or much reduced on the output images.
Technique
Before a plate was digitized, a coordinate system was
chosen to define the orientation of the field. This
coordinate system served to ensure that as many plates as
possible were oriented the same way so identical areas were
scanned on each. This facilitated the adding or subtracting
of images by preventing any shifts or rotations that might
otherwise be introduced. The coordinate system was defined by
selecting two widely spaced stars near an edge of a plate that
were at approximately the same right ascension or declination.
The object plate was then positioned such that these stars
defined the x-axis of a scan. The calibration plate was then
placed on the stage near the object plate and positioned such
that the x scan direction passed through each row of spots.
Although the coordinate system defined in this way was used
for many of the plates, other systems using different pairs of
stars had to be used, too. These were required when the
plates didn't have the same two stars used previously, either
27
because they were smaller plates, or because the telescope was
not pointed at the same position.
Digitization was done in a three-step process: first, the
calibration plate was scanned and then two scans were made of
the object plate. The microdensitometer was zeroed at the
level of the fog on the calibration plate, and a 50 pm x 50 gm
aperture used to scan the calibration spots, with two scans
being made through each row of eight spots. The step size in
x was 50 pm as was the offset in y between scans.
The plate with the object was divided into two
overlapping regions; an inner, or object field containing the
galaxy, and an outer, or background field covering most of the
plate and containing the inner field. The outer field, which
extends far beyond the visible object, was measured with the
same aperture as was used for the calibration plate. The area
chosen was typically 15 cm x 15 cm, although the small plates
and short exposure plates had smaller areas. Because of the
large area of the outer field, widely spaced samples were
taken to keep the number of data small. Pairs of scans were
made in the x-direction with a sampling interval of 50 pm and
a shift in y of 50 pjm between each scan. Each pair of scans
was separated in the y-direction by either 5 or 10 mm.
The inner field contained both the galaxy and sky back-
ground. The area was chosen to extend about twice as far as
the galaxy could be seen; on deep 18 cm x 18 cm plates this
was 5 cm x 5 cm. On smaller plates a smaller area had to be
28
chosen but one which still enclosed the galaxy. A star near
the nucleus of the galaxy was used to center the area so each
plate would have matching inner fields. In all cases the
sides of the inner field are parallel to the sides of the
outer field. The inner field scans were done with an aperture
of 12.5 pm x 12.5 im or 25 jpm x 25 [pm, the choice being
determined by the diameter of the smallest stellar images.
Scans were done with a step size of 12 Km or 25 lm, and the
spacing between scans was the same as the step size.
As a check on the photometry, for most plates density
readings were made of the same position on the plate, before
and after the inner field scans. In general, those scans
showing drifts larger than a few percent in density were not
used in the analysis. Exceptions to this occurred for two of
the short exposure plates which had large drifts in the
density readings caused by changes in the alignment of the
diaphragms in the microdensitometer during scanning. These
occurred at or near the end of a scan, and the affected
regions were discarded. For the plates for which before and
after density readings could not be directly compared, the
appearance of the background of the inner region was visually
examined on the digitized images. In the one or two cases
where the background showed large variations that looked like
photometer drift, the images were discarded. The removal of
variations caused by probable photometer drift is discussed
later in this chapter.
Data Reduction
Numerical Mapping Technique and Reduction of the Outer Field
The data from the microdensitometer scans of the inner
field consist of measurements of the galaxy and the sur-
rounding sky. The galaxy image has a component due to the sky
brightness which is small in the central regions where a
bright nucleus dominates, but which in turn becomes dominant
as the outermost extent of the galaxy is reached. Because the
contribution to the total luminosity of a galaxy arising from
its faintest parts can be large, the sky background must be
carefully removed so that only light from the galaxy remains.
Fluctuations in the sky level are the major source of error
and care must be taken to remove them. Various means have
been developed to do this sky subtraction although many are
over-simplified and do not adequately allow for variations in
sky brightness or plate response over the area considered.
The Numerical Mapping Technique described by Jones et al.
(1967) was used for the determination and elimination of the
sky background. As they point out, (1) it allows exhaustive
use of the information in a photograph of an extended object
rather than the narrow sampling along selected cross-sections
widely used previously; (2) it permits much more rigorous
allowance for the photometric "local errors" of the
photographic emulsion and sky background irregularities of
various origins which are the main limitations to accurate
photometry of faint objects, such as the outer regions of
30
galaxies; and (3) it is designed for data in digitized form
from a scanning microdensitometer. The method has been suc-
cessfully used in other photometric studies (Barbon, Benacchio
and Capaccioli 1976; Tsikoudi 1977; Pence 1978) and only a
brief description of it will be given here. Much of the
original computer code has been rewritten to enable it to
handle larger arrays and to make the density to intensity
conversion. The algorithm for fitting the polynomial to the
background densities was left unchanged. The reductions were
done on the IBM 3090/400 computer of the NERDC at the
University of Florida.
Over the inner field of the plate the measurements are
composed of contributions from both the galaxy and the sky,
IG+S. The outer field has only the sky, Is, contributing to
it. In the Numerical Mapping Technique orthogonal functions
are used to analyze the measurements in the outer field and to
determine the best approximation to the sky background over
the whole plate by least squares techniques. The result is a
polynomial in x and y expressing the density at each point.
The sky values in the region of the galaxy are calculated by
interpolation over the "hole" caused by the inner field. The
contribution due to the galaxy alone, relative to the sky is
then given by
IG s (2)
Is
31
Plots of the scans of the outer field were examined for
each plate before any analysis was performed. Any scans or
pixels that showed obvious irregularities were not included in
the background sample. These were caused by scans that were
too close to the edge of a plate or that included a defect in
the emulsion. The remaining data were then used to map the
background.
The background sky measurements of the outer field are
composed of two distinct components: (1) a slowly varying
continuous component caused by real differences in sky
brightness, emulsion sensitivity, etc., and (2) a discrete
component due to point sources of light and defects in the
emulsion. The two components are treated separately and by
different methods.
The discrete component was removed before fitting the
continuous component. This was accomplished by fitting a low
order polynomial D(x,y), where
D(x,y) = D(1) + D(2)x + D(3)y + D(4)x2 + D(5)xy + ...
+ D(10)y3 (3)
by least squares to the density readings d(x,y) of the outer
field. Readings which had residual R(x,y) = d(x,y) D(x,y)
greater in magnitude than a specific rejection level, Ba, were
discarded and the process was repeated on the reduced data set
until no more residuals were rejected or the distribution was
Gaussian, which is to be expected in the ideal case. The
32
value of B used was between 2 and 3, depending on the analysis
cycle, and a is the standard deviation of the residuals. For
an unbiased estimate of the standard deviation of residuals 0,
Jones et al. used
ek = Ek (4)
ek N2 k)
where k = 10 denotes the number of coefficients used to define
D(x,y), N2 is the number of data points and N2 k is the
number of degrees of freedom remaining after subtracting one
for each coefficient D(k) in equation (2) and
N2
Ek = (d(x,y) Dn(x,y))2. (5)
n-=l
With the discrete component of the background removed the
continuous component can be mapped. This mapping is compli-
cated by two factors: (1) the remaining outer field data are
affected by small random fluctuations (noise), and a certain
amount of smoothing is necessary to avoid a rough and physi-
cally unrealistic representation; and (2) the distribution of
the data points has a large hole in the central region corre-
sponding to the inner field, and this will produce instability
over the inner field if a polynomial of too high a degree is
used to fit the data.
33
If the wavelength (shortest distance between two relative
maxima or two relative minima) of the approximating polynomial
is smaller than the width of the hole, it may oscillate un-
realistically over the inner field where there are no data.
To guard against this, an upper limit is placed on the degree
of the polynomial to keep the wavelength sufficiently large to
control instability. The highest permissible degrees in x and
y are determined by normalizing the coordinates by a linear
transformation so that the limits of the outer field extend
from -1 to +1 in both x and y. Then the shortest permissible
wavelength in normalized coordinates will be the length of the
transformed side of the rectangular hole. With the value of
the shortest wavelength determined, a standard orthogonal
system, Chebychev polynomials, is applied to find the highest
permissible degree of the polynomial, using Table 4 on page 41
of Jones et al.
For most of the plates used here, the continuous com-
ponent of the sky background is represented with a polynomial
Y(x,y) = D(1) + D(2)x + D(3)y + D(4)x2 + D(5)xy +...
... + D(43)xy7 + D(44)xe + D(45)ye (6)
of degree eight in both x and y. The only exception was plate
5410 which used a polynomial of degree seven. The polynomial
may be truncated at some term to effect the desired smoothing
of the data mentioned above. The terms of the polynomial are
calculated such that they are mutually independent in the
34
sense that if any term is terminated, the remaining series is
still a least squares fit (see Jones et al. p. 42). The cri-
terion for defining the cutoff is to minimize the standard
deviation of the residuals as given by equation (4) with k
equal to 36 or 45, depending on the degree of the polynomial.
By examining how the standard deviation of residuals changes
as the number of terms included in the polynomial is changed,
and by including terms that make a significant contribution to
decreasing the residuals an upper limit on the number of terms
is determined.
With the cutoff of terms in the polynomial determined,
the remaining background data of the outer field are fitted
again by least squares with the adopted number of terms. The
polynomial is used to produce a density contour map of the sky
over both the outer and the inner fields as shown in Figure 1
(plate 9309) which also shows the relative sizes of the inner
and outer fields for a typical plate.
Before approximating the variations in density of the sky
background, the number of data in the outer field were reduced
by taking samples every 2.5 mm in x and at either 5 mm or 10
mm in y, depending on what was used for the scan. Because the
background scans are in pairs with each scan having width 50
|m, an effective aperture of 100 pm x 100 pm was formed by
taking a mean of four pixels for each sample. This has the
beneficial effect of reducing the noise caused by granulation.
Sl,./. For t-h---s pt k -' x-axi is1
e ie
S' i ;
-1
Figure 1. Density contour map of plate 9309. The contours
are every 0.01D with the circular area having the lowest
value,-0.41. For this plate, k 45. The x-axis is 142.5 mm
in length; the y-axis is 145 mm. The small square shows the
inner field for this plate.
36
The result was a set on the order of 500 to 1000 background
samples depending on the size of the plate.
The outer field was at least several times as large as
the inner field, so that after removing the latter, there was
still enough area to use for sky background fitting. Tests
were run on plate 9416 to determine the effect of first,
having the background sampled at intervals of 5 mm and 10 mm
spacing in y, and second, on changing the size of the inner
field from 37.5 mm x 35 mm to 25 mm x 30 mm. In the first
case, although the number of data samples was reduced by a
factor of two, there was little difference in the contour maps
of the background and the interpolated inner field for the two
spacings. The final contour map of the galaxy image was only
slightly affected, with the image being slightly noisier for
the larger spacing test. The difference was negligible. In
the second case, the smaller inner field did not seem to have
a significant affect on the results: the contours of the
galaxy images were almost identical to the original ones, and
only the faintest contour levels showed a small, occasional
change. Thus, it appeared that the fitting of the background
was not adversely affected by the choices selected for the
inner and outer field measurements.
Reduction of the Inner Field
The inner fields were chosen to satisfy two criteria: (1)
they had to be large enough to completely contain the faintest
detectable light level of the galaxy on each plate, and (2)
37
the inner field boundaries should overlap the background scans
whenever possible. The first criterion was satisfied by
selecting areas considerably larger than the size of the image
on the plates and that contained the isophotes measured in
previous studies. The second criterion was important because
the inner and outer fields were scanned with different size
apertures. The data were put on a common system by taking the
mean of the readings along the common boundary measured
through both apertures, inner and outer, M1 and Mo, and then
multiplying the outer field values by their ratio Mi/M,.
Because the inner fields were measured with small
apertures, either 12.5 pm or 25 gm on a side, averages were
formed of 6 x 6 or 3 x 3 sets of pixels, respectively, to
reduce the number of pixels and improve the signal to noise
ratio. This reduction in the number of pixels was neces-
sitated by the constraints on getting plots of the data. The
resulting resolution was 7.1" and 7.4", respectively. This
loss of resolution is not important, especially in the outer
regions of the galaxies where no fine structure is expected.
Density to Intensity Conversion
Before the background sky was subtracted from the inner
field data, photographic densities were converted to inten-
sities using calibration plate data. This was accomplished
using a polynomial of the form (de Vaucouleurs 1968; Wevers
1984)
n
log E= a (log C)) (7)
1-0
38
to represent the characteristic curve, where E was the
relative exposure (=It) and 0 the opacitance (=106-1) of each
calibration spot, AD was the density above fog, and n was
usually 3 but sometimes 4. The values of the coefficients,
a1, were then determined by a least squares fit to the cali-
bration plate data. The advantages of this formulation were
that it was a relatively simple mathematical representation of
the data and it avoided the erors that would have resulted
from using the usual characteristic curve near the threshold
where the curve asymptotically approaches the log E axis.
The values of the coefficients a1 were read into the
subroutine which converted the density measurements of the
inner field to intensities, Ic~+(x,y). The value of the sky
background at each position, Is(x,y) was also calculated using
the computed sky density Y(x,y) from equation (6) and the a1
above. Then the intensity of the galaxy alone, Ig, was cal-
culated from equation (2) giving
( ) = IG+s(x,y) I(x, y)
Is (X, y)
Star and Blemish Removal
The inner field area surrounding each galaxy contained on
the order of 1000 stars and background galaxies. In addition,
the hypersensitization procedure sometimes produced blemishes,
especially on the IV-N emulsion. These unwanted features were
removed before any further processing was done to the image.
Because of the variety of sizes, shapes, intensities, and
39
local background values, an automatic rejection procedure was
not found to be successful. Consequently, for objects judged
to be well away from the galaxy, the size and position of each
object was specified and the affected pixels replaced by a
mean value formed from the surrounding pixels. For objects
near the galaxy image, a different procedure had to be used,
especially for those objects that were superimposed on regions
of the galaxy with a large intensity gradient. These objects
were removed a pixel row at a time, interpolating over the
affected pixels using pixels on either side of the affected
ones. Contour maps were then made of each image and were used
to judge the smoothness of the procedure and to make cor-
rections where necessary. The only significant problem was in
the southeast area of NGC 253 where a large, low surface
brightness region had two bright stars superimposed on it.
Detailed structure in this region was lost as a result of
removing the stellar images.
One of the plates, 8544, had a satellite trail cutting
through the galaxy image running north-northwest and south-
southeast and passing north of the nucleus. The trail was
removed by interpolating over the affected pixels and the
resulting contours are quite smooth.
After the stars and blemishes were removed from an image,
the mean intensities of each row and column were plotted.
Ideally, out away from the galaxy, these should have an
average value of zero, but this was usually not the case.
40
This nonzero level was noted by other photometrists (Jones et
al. 1967; Tsikoudi 1977) and Pence (1978) who hypothesized its
origin in a small drift in the response of the microphotometer
during the scan of the plate. In all cases where the mean
intensities of the rows and columns were not zero, a cor-
rection was applied by fitting a plane to the mean values of
the outer rows and columns and subtracting it from the image.
Because the sky densities were so low, the short exposure
plates showed the largest effects of photometer drift, up to
8% of the sky intensity for plate 8545. However, the data
more than about one magnitude below sky were not used, and the
drift effect was subtracted out, so this was not a problem.
As mentioned earlier, four of the images had low-
intensity bands of noise in them: those of plates 5394, 6487,
6488, and 9417. These bands were parallel to the scan
direction and quasi-periodic in nature. After any linear
corrections were made to the background as described above, a
sine function with period and amplitude approximately matching
that of the noise was subtracted from each column of the
image. Although it was not possible to completely remove all
of the noise in some cases, the level that remained was less
than 1% of the sky background. And because most of these
plates were combined with other plates of similar passbands to
form composite images later, the overall level of the noise
was further reduced.
Photoelectric Calibration
The zero-point calibration of each plate was done using
either photoelectric photometry measurements through one or
more apertures centered on the galaxy's nucleus (Jones et al.
1967) or photoelectric drift scans through the nucleus. Since
the plate measurements were normalized to the sky intensity,
the zero-point of each plate was just the sky magnitude.
For NGC 55, Longo and de Vaucouleurs (1983) listed eight
measurements in the Johnson V system, hereafter Vj, taken
through seven apertures. The measurements through the four or
five largest apertures were used for the calibration. These
apertures ranged from 22.2" to 131" radius for the long
exposure plates, and down to 10.4" for the short exposure
plate. Three measurements in the Cousins system were made
with an aperture of 131" radius and were provided by de
Vaucouleurs (private communication), two in R, and one in I,.
The pertinent data are listed in Table 5. Where more than one
measurement through a given aperture was available, the mean
of the measured magnitudes was used in the calibration.
Because many of the Vj measurements were made through small
apertures, annuli were formed by taking the difference between
the largest and smaller aperture values. This served two
purposes: it minimized the errors caused by problems in
centering a small aperture on the nucleus of a galaxy where
luminosity gradients are usually largest, and it reduced or
eliminated the effect of an overexposed nucleus.
TABLE 5
APERTURE PHOTOMETRY OF NGC 55
Aperture Radius Passband Magnitude Source"
10.4" V, 13.31 1
22.2 12.33 "
44.4 11.18 "
67.2 10.24
131 9.38 "
f" 9.32 "
SR, 8.99: 2
9.06 "
I, 8.60: "
Note: a (1) Longo and de Vaucouleurs (1983); (2) de
Vaucouleurs (private communication).
In the case of a single aperture measurement, the sky
magnitude is computed in a straightforward manner. If ma is
the measured magnitude through an aperture of radius a, and L,
is the integrated luminosity within the same radius derived
from the photographic plate, then
ma = k 2.51og La (9)
where
2n a
La = f fl(r,e)rdrdO. (10)
0 0
I(r,0) is measured in terms of the sky background, and the
constant term, k, in equation (9) is the magnitude of the sky.
The integrated luminosity is calculated by a program based on
43
one given by Jones et al. that sums up all the values of
I(r,0) in the digitized image in the area of interest.
For annuli, let F1, F2, mi, and m2 be the photoelectri-
cally measured fluxes and their corresponding magnitudes for
two apertures, and let F, = F2 F, be the flux corresponding
to the annulus. Then
F, = 10-0.42 10-0.4m (11)
and
ma = -2.51og(10-.'42 10-o4m") (12)
Using equation (10) to get L, = L, LI, the sky value is then
given by equation (9).
Two of the plates, 6473 and 6488, were quite deep, and
since only one aperture size was used for the photoelectric
measurements, annuli could not be formed to avoid any effects
of a heavily exposed nucleus. Instead, the magnitude scale,
and therefore the sky value, for each of these plates was
determined by a method similar to that used by Tsikoudi (1977)
to calibrate galaxy photometry. The luminosity profile along
the major axis was plotted on an arbitrary magnitude scale and
placed over the already calibrated profile for plate 6467.
The uncalibrated profile was then shifted along the magnitude
scale until the two profiles matched as best as could be
44
judged by eye. Table 6 lists the calculated sky magnitudes
for the NGC 55 plate material.
TABLE 6
NGC 55 PLATE CALIBRATION
Plate # Emulsion Exposure " Nb
6467 IIIa-F 115" 20.830.05 2
6473 IIIa-F 120" 20.640.05" -
6487 IIa-D 60" 21.030.05 4
6488 IIIa-F 150" 20.840.050 -
8544 IIIa-F 20" 20.860.05 2
9232 IIa-D 60" 21.050.07 4
9241 IIa-D 15" 20.760.04 5
9281 IV-N 60" 19.550.02d 1
9403 IV-N 60" 19.440.02d 1
9416 IV-N 70" 19.560.02d 1
9440 IV-N 20" 19.330.02d 1
Notes: The mean zero point in magnitudes per square arc
second derived from the aperture measurements and
corrected to a common color system.
b The number of measurements used to determine .
The sky value as determined from plate 6467 as
explained in the text. The error is that of plate
6467.
d Estimated uncertainty based on internal plate error.
See Chapter III.
For NGC 253, Longo and de Vaucouleurs gave fourteen meas-
urements taken through thirteen apertures in the V, passband.
Two of these were based on B magnitudes and were omitted from
consideration. Of the eleven measurements that remained, the
eight taken through apertures 10.4" to 80.7" radius were used
to form annuli as described above. One measurement, listed as
having V, = 12.65 and aperture size 16.9" was not used because
45
it was much fainter than the values listed for apertures of
10.4" and 14.7". The apertures and magnitudes that were used
for the calibration are listed in Table 7. The calculation of
the sky magnitude for each plate preceded as for NGC 55.
TABLE 7
APERTURE PHOTOMETRY OF NGC 253
Aperture Radius Passband
10.4" Vj
14.7 "
22.2
32.9
44.4
57.2
72.0
80.7
Source: Longo and de Vaucouleurs (1983).
Magnitude
12.32
12.06
11.32
10.76
10.20
9.72
9.40
9.19
As a check on the accuracy of the aperture photometry,
photoelectric photomet-y drift scans were used to compute sky
values. Using the photoelectrically measured magnitudes of
the galaxy through a given aperture at different positions,
the sky value was calculated at the same positions from the
integrated photographic data, just as for the aperture
photometry. Since the drift scans were measured through a
larger aperture than the digitized pixel size, the digitized
image was convolved with a Gaussian beam to degrade the
resolution to match the photoelectric data before preceding.
The photoelectric drift scans were in the Vj passband and were
kindly made available by Dr. W. D. Pence. The observations
were made using the 92 cm reflector at Mc Donald Observatory
46
with a 32.48" diameter aperture, each scan being in an east-
west direction and passing through the nucleus. Four or five
determinations were made for each of plates 5396, 5410, and
9243, giving sky values of p, = 19.880.14, 19.990.11 and
19.740.19, respectively. These compare well with the
aperture photometry values of ., = 19.940.07, 19.880.08 and
19.720.03.
No aperture photometry data were available for the red
and infrared passbands for NGC 253. However, Spinrad et al.
(1978) made east-west drift scan measurements of NGC 253 in r
and i passbands centered at wavelengths 604.0 and 746.0 nm,
respectively, using a 67 arc second square aperture. The red
and infrared digitized images were convolved with a Gaussian
beam to degrade them to their resolution and integrated
magnitudes were calculated at five or six locations corre-
sponding to the same positions in the drift scan data. Sky
magnitudes were determined as before using equations (9) and
(10).
The calibration plate 8545 was lightly exposed and its
characteristic curve did not extend to densities as high as
those in the galaxy image. It did, however, extend up to the
linear region where the relationship between density and
intensity is well behaved. As a consequence, the sky value
derived from drift scan measurements was not considered as
reliable. Instead, the magnitude scale for this plate was
determined by the method of matching the luminosity profile to
47
that of plate 5394 as discussed earlier for the NGC 55 plates
6467, 6473 and 6488. The uncertainty in the derived value of
k, the sky magnitude, was estimated from the fitting pro-
cedure. All of the sky magnitudes for the NGC 253 plates are
listed in Table 8.
TABLE 8
NGC 253 PLATE CALIBRATION
Plate # Emulsion Exposure " Nb
5394 IIIa-F 60" 20.060.08 5
5396 IIa-D 20" 19.860.07 6
5410 IIa-D 30" 19.800.07 8
8545 IIIa-F 20" 19.930.15" -
9243 IIa-D 10" 19.640.03 7
9301 IV-N 60" 19.160.18 6
9302 IV-N 15" 19.590.13 6
9309 IV-N 60" 19.030.22 6
9404 IV-N 60" 18.860.39 6
9417 IV-N 70" 19.130.19 6
Notes: a The mean zero point in magnitudes per square arc
second derived from the aperture me surements or
drift scans and corrected to a common color system.
b The number of measurements used to determine .
The sky value as determined from plate 5394 as ex-
plained in the text. The error is an estimate.
The magnitudes determined using equations (9) through
(12) were not all on a common system. The photoelectric
aperture measurements were in the Johnson V, and the Cousins
R, and I, passbands which had effective wavelengths of 550 nm
(Walker 1987), 640 nm and 790 nm (Bessel 1979). The drift
scans were in Vj and the Spinrad et al. (1978) r and i pass-
bands mentioned earlier. Only the Cousins I, and the I pass-
48
band used here were a close enough match to interchange. And,
as mentioned early in this chapter, two plates were taken
through a "Russian" red filter, hereafter R, which, with the
IIIa-F emulsion yielded an effective wavelength of 646 nm.
Each of these points will now be addressed.
To convert the NGC 55 data, the V,, R, and I, aperture
photometry taken through a 262" diameter diaphragm centered on
the nucleus was plotted versus effective wavelength. A smooth
curve was drawn through the points and the magnitudes corres-
ponding to the effective wavelengths of the system used here
were read off the graph. The conversions were as follows:
V = Vj 0.08; Ra = R, 0.02; R = R. 0.05; and I = I,.
NGC 253 has a strong color gradient near the nucleus so
mean magnitudes in r and i were taken from Spinrad et al. at
2.68' and 4.9' east and west of the nucleus. As for the NGC
55 data, a plot was made of magnitude versus effective wave-
length for each position. Since only two passbands were
available only a straight line could be fit to the data,
although this may not be a problem given that the plot of the
NGC 55 data for three colors had only a small amount of
curvature. The determination of the correction to be made to
I involved a small extrapolation while that for R, and R
required interpolations. The means of the results were:
V = Vj 0.08; Ra = r 0.20; R = r 0.24; and I = i 0.28.
The value for V was taken from the NGC 55 fit. The difference
between the two values used to get the mean I conversion
49
constant was 0.1 magnitude, the largest of all the
differences.
It is difficult to estimate an error in these conversions
but 0.1 magnitude is probably not unreasonable given the
simple nature of the process used. All of the magnitudes
given in Tables 6 and 8 were corrected using the above
conversions.
Plate Addition
The long-exposure intensity maps, now calibrated and
cleaned of stars and blemishes, were reduced in size by taking
a 2 x 2 average of all the pixels. This served two purposes:
it reduced the number of pixels in each image to the order of
256 x 256, resulting in a large saving in time and money in
the subsequent image processing, and it improved the signal to
noise ratio of the data. Angular resolution was reduced to
about 14" for the V and R plates, 15" for the NGC 253 I
plates, and 20" for the NGC 55 I plates. The resulting
intensity maps in each color were aligned and added together
for each galaxy. Those portions that completely overlapped
were used to form the final images. All of the individual
images were given equal weight except that of plate 6487 which
was given half weight because of residual noise in the
background.
As a final check on the photoelectric calibrations, the
added long exposure major axis profiles were compared to the
short exposure profiles. The two profiles in each color were
50
overlaid and any offset on the magnitude scales was noted.
All three colors of the NGC 55 data showed small discrepan-
cies which can probably be traced to the small size of the
apertures used for the photoelectric measurements and the
heavy exposure of the nuclear regions on the plates. The long
exposure profiles were judged to be most likely in error, and
they were shifted in magnitude to bring them into agreement
with the short exposure profiles. The adjustments amounted to
-0.14, 0.25, and -0.13 magnitudes for the V, R, and I
profiles, respectively. Only the V magnitude for NGC 253
showed a discrepancy between the long and short exposure
profiles, amounting to 0.32 magnitudes. The final values of
the sky magnitudes adopted for NGC 55 were: Cp = 20.910.04;
g, = 21.020.03; and R, = 19.390.01. The last two values are
more uncertain than their errors imply. For NGC 253 the sky
values were: A, = 20.150.05; JR = 20.060.08; and, finally,
Ri = 19.100.11.
Image Smoothing
The added deep images of both galaxies were smoothed to
further reduce the effects of residual background noise and to
diminish small scale irregularities in the outer contours.
This was generally done in three steps similar to that de-
scribed in Jones et al. First, light smoothing was done by
convolving the images with a Gaussian beam whose full width at
the 1% level was five pixels. Moderate smoothing was done by
51
convolving a Gaussian beam whose full width was seven pixels.
Finally, heavy smoothing was done by first doing a 2 x 2 aver-
aging of all the pixels and then convolving a Gaussian beam
whose full width was five pixels. This final step was not
performed on the NGC 253 red data because of the low quality
of the data at the level where the smoothing would be done.
The light and moderate smoothings were used for the
intermediate areas of the galaxy and the heavy smoothing was
used for the outermost regions where the signal was weak and
the isophotes began to break up into islands. The result is
a decrease in the high frequency component in the structure
that diminishes the effects of small dust clouds, HII regions
and noise in the outer regions. The major structural features
are, however, preserved.
Smoothing was also done to the short exposure plates, not
to improve the signal to noise so much as to remove clutter
from the maps caused by resolved features such as individual
HII regions. Here, the images were convolved with a Gaussian
beam whose full width at the 1% level was five pixels. The
size of the beam used for smoothing is indicated on the
isophote maps.
CHAPTER III
PHOTOMETRIC DATA FOR NGC 55
General Description
NGC 55 is classified as a Magellanic type barred spiral,
SB(s)m sp (de Vaucouleurs, de Vaucouleurs and Corwin 1976,
hereafter referred to as RC2), although the classification is
marked as uncertain. Classification of this galaxy is
difficult because it is viewed nearly edge-on and is seen
looking along the bar from a "rear view" (de Vaucouleurs
1961). The appearance in blue light is shown in Figure 2,
reproduced from a print in the Atlas of Galaxies (Sandage and
Bedke 1988) where resolution into stars can just be discerned.
North is at the top and east is to the left in the photograph.
The scale of the print is shown by the bar in the northwest
corner.
One immediately notices the asymmetry in this system.
The optical nucleus is displaced from the center of the
galaxy, and, as mentioned in Chapter I, the radio nucleus,
mass center, and centroid of total light are all offset toward
the southeast (see pages 5,6 and 7 for the positions). The
western side appears to terminate rather abruptly and the
northern and southern halves are of different thickness near
the region of the bar. Numerous HII regions are readily seen
53
co
cmm
'U
ir l
I,)
44
0
4)
"-4
44
H
'II
CD
'U-
H
4)
.8-
r0
a)
p4
U)
$4
It)
C-)
V) U)
-7
54
as well as the presence of a considerable amount of dust, and
the former seem to be almost exclusively found on the eastern
side. Rotation is present with the east side of the galaxy
receding (de Vaucouleurs 1961). Table 9 presents some
observational parameters of NGC 55.
TABLE 9
ELEMENTS OF NGC 55
R. A. (1950)" 00h 12".40
Dec. (1950)" -390 28'.0
Galactic 1, b" 3320.90, -75.74
Supergalactic L, B' 2560.3, -2.4
Type" SB(s)m sp? (t=9)
Observed velocity Ve +131 km/sec
Corrected velocity Vo" +98 km/sec
Apparent distance modulus 26.48 0.4:
Corrected distance modulus 26.25 0.4:
Note: RC2, 1976.
Isophotal Contour Maps
The short exposure V,R and I isophote maps of the inner
region of NGC 55 are shown in Figures 3, 4 and 5. The contour
interval is g = 0.2 and the scale is shown by the bar. Also
shown is the full width at half maximum (FWHM) of the smooth-
ing Gaussian beam. The R map appears larger than the other
maps because it extends over a larger range of magnitudes.
Three bright HII regions 0'.9, 1'.7 and 4'.0 east of the
nucleus, and roughly on the major axis, are easily identified,
especially on the V and R maps. In addition, the dark dust
lane to the east of the nucleus and running northeast to
southwest between the inner HII regions and the nucleus can
N NGC 55 V
120"
Figure 3. Short exposure V isophote map of NGC 55. The
contour interval is 0.2 magnitude and the innermost isophote
is at L~ = 20.26 and the outermost at 22.26. North is up and
east is to the left. The bar shows the scale and the FWHM of
the smoothing beam is shown by the circle.
NGC 55 R
120"
Figure 4. Short exposure R isophote map of NGC 55. The
contour interval is 0.2 magnitude and the innermost isophote
is at N = 19.76 and the outermost at 22.36. North is up and
east is to the left. The bar shows the scale and the FWHM of
the smoothing beam is shown by the circle.
N NGC 551
I I
120"
0
Figure 5. Short exposure I isophote map of NGC 55. The
contour interval is 0.2 magnitude and the innermost isophote
is at l, = 19.13 and the outermost at 21.33. North is up and
east is to the left. The bar shows the scale and the FWHM of
the smoothing beam is shown by the circle.
58
also be seen. The nucleus itself appears almost circular or
slightly elongated along the major axis. Isophotes
surrounding the nucleus appear to be oriented at a smaller
position angle (93.08) than the outer isophotes, although this
shift in position angle is probably caused by the dust lanes
to the east and west of the nucleus, as can be seen in Figure
2. A blue (pg) light isophote map (de Vaucouleurs 1961) shows
much the same appearance.
Figures 6, 7 and 8 show the V, R and I isophote maps from
the added long exposure plates. Again, the bar shows the
scale of the map, and the circle the FWHM of the smoothing
beam. Here the contour interval is = 0.5. These maps do
not show the total extent of the galaxy but go to a level
where the contours start to break up into islands and the
residual background noise begins to show. A lobe of increased
brightness is visible on the east side about 10' from the
nucleus: this is interpreted as an inner, asymmetrical arm by
de Vaucouleurs. Also of interest is the tongue of material
about 8' from the nucleus in position angle = 2650 seen in the
I map. This feature closely matches a similar feature in HI
mapped by Hummel, Dettmar and Wielebinski (1986). The lobe at
the west end of the galaxy and the slight bulge along the
south side of the minor axis may be related to HI features,
too, although no optical emission is seen in the northeast
where they have HI extending in a tongue. They ascribe the HI
features to warping and wobbling of the HI plane.
NGC 55 V
10'
0
Figure 6. Visual isophote map made from added long exposure
plates. The contour interval is 0.5 magnitude. The innermost
isophote is at v 20.92 and the outermost at 25.92. North
is up and east is to the left. The bar shows the scale and
the FWHM of the smoothing beam is shown by the circle.
N NGC 55 R
10'
oO
O
Figure 7. Red isophote map made from added long exposure
plates. The contour interval is 0.5 magnitude. The innermost
isophote is at N 21.02 and the outermost at 26.02. North
is up and east is to the left. The bar shows the scale and
the FWHM of the smoothing beam is shown by the circle.
NGC 55 I
0 o
9
10'
Figure 8. Infrared isophote map made from added long exposure
plates. The contour interval is 0.5 magnitude. The innermost
isophote is at ,z = 19.39 and the outermost at 24.89. North
is up and east is to the left. The bar shows the scale and
the FWHM of the smoothing beam is shown by the circle.
A _
62
The position angle of the major axis is 1070.8 00.5 as
estimated from the long exposure isophote maps. This compares
well with determinations of 1050 (de Vaucouleurs 1961) and
1070 (Robinson and van Damme 1966) from optical data, and 1090
(Hummel, Dettmar and Wielebinski 1986) from 21 cm data. The
minor axis is taken as the line perpendicular to the major
axis and passing through the nucleus, even though this line
does not pass through the center of the image.
As mentioned in Chapter I, the published optical incli-
nation values for NGC 55 range from 790 to 900, and the HI
yields an inclination of = 800. From the appearance of the
galaxy in photographs, the lower value is probably ruled out
since no indication of far side structure or an elongated bar
is seen. The inclination, i, can be calculated using equation
(13)
Cos2i = ) (13)
(1 qo2)
where q is the ratio of the apparent minor axis to the major
axis, and q0 is the ratio of the true minor axis to the major
axis. Measurements of the axes at g = 23,24,25,and 26 in V,
R and I yield mean values of q of 0.21 0.02, 0.22 0.02,
and 0.25 0.04, respectively. The mean q from de Vaucouleurs
(1961) at Lp, = 23.7, 24.7 and 25.85 is 0.20 0.02, the same
as the value in the RC2. The mean value of the four flattest
(q 5 0.25) SBm galaxies in the RC2 (not including NGC 55) is
63
= 0.25 0.005. If late type galaxies that might be
mistaken for SBm, or vice versa, are examined, the means are:
= 0.19 0.05 (9 SBd); = 0.22 0.04 (2 SBdm); and
= 0.20 0.03 (6 Im). Inclinations for NGC 55 greater than
800 result from the possible real solutions of equation (13)
using the above estimates as qo. Only in the case of the
smallest value of q, 0.11, for the SBd galaxy NGC 7412 A, can
a value of i less than 800 be obtained. Hence, we conclude
that 800 < i < 90, and a value of i = 850 will be adopted
here.
Luminosity Profiles
Figures 9, 10 and 11 show slices along the major and
minor axes in each color. The bright, inner regions are from
the short exposure images, and the more heavily smoothed and
added long exposure images are used for the fainter magni-
tudes. Representative internal errors are shown by the bars
at the side of each figure for three brightness levels. Each
displays the effect of a 1 0 error, and is the standard
deviation of two to four sets of pixels from near the corners
of the images used to form the profiles.
Major Axis
The major axis profiles all show the marked asymmetry of
this galaxy. To the east of the nucleus bright HII regions
are apparent at 1'.5, 4'.0 and 9'.5, and obscuring dust is
evident at 0'.7 and 5'.0. Dust on the east side also causes
the depression in the luminosity profile between 5' and 9'.
o
N 41
\A
=,
$~4
(%J
VIr
0
wZ
L()
C-)r
0i
z1
I~e p. I.-*-
I I I I I ~ I -
-
t I I l m I m l
U)
q44
0
0)
44
-ri
0
5.4
0)
0r.
1:4)
04.
:I
04)
0)
r~X
14-)
$4.
rd 0
4.1
>
0
$4)
5444-
90'
-A 1
W (d
- I I I I I I I
OQ
Z0
Z-
SI I I I I I
O
CN
L.
0
u-
o H
VI
0
b
0
O
0
0
O
r)
U.
8
0
0)
0
-w
91
-4
0
05I
r4
$41
0
t*.
HO.
*sC
*
b
SI I I I
u
Ci)
7
I I I I I I I I
(-N
C4.
O LI
0
U)
8
44
0
M
4J
0
r*
0
14 4)
I
0
*rz
C Q
8-
*o <1
^S
0~
0 Mc
M *M
& b
67
On the west side, the drops in brightness at 2'.0 and 4'.5 are
seen to correspond to dust (see Figure 2). No HII regions are
cut by the major axis on this side. The large "bump" at 22'
on the east side of the R profile is not real but an artifact
caused by noise on one of the R plates.
All three colors show that the luminosity distribution is
approximately exponential in the outer regions, in agreement
with the pg (f,,, = 430 nm) study of de Vaucouleurs (1961).
This is especially noticeable for the west profile where the
decline is easily seen to extend over at least five
magnitudes. On the east side, the exponential decline can be
followed nearly as far but only below about pL = 23 and gC =
1 2 22.
The gradients of these exponential regions, in units of
magnitudes per arc minute, are given in Table 10. Column (1)
gives the color, or passband, of the observation. The
gradients of the east and west profiles of the major axis are
given in columns (2) and (3), with the magnitude range used
for the calculation in parentheses. The mean gradient of both
sides is given in column (4), and column (5) gives the
gradient of the entire west profile and the magnitude range.
The pg mean gradient is from de Vaucouleurs and is included
for comparison; the other pg values are the results of calcu-
lations made using his luminosity profile. In all cases the
values are only approximate because of the uncertainties in
the magnitudes at the fainter levels. It is the differences
68
between the gradients that are emphasized, not their actual
values.
TABLE 10
NGC 55 MAJOR AXIS LUMINOSITY GRADIENTS
Color East" Westa Mean" West"
pg 0.31(23-27) 0.41(23-27) 0.35b 0.31:(20-27)
V 0.75(24-28) 0.55(24-28) 0.65 0.45(19-28)
R 0.70(23-25) 0.32(23-25) 0.51 0.40(19-28)
I 0.96(23-27) 0.41(23-27) 0.68 0.44(19-27)
Notes: a All values are in magnitudes per arc minute.
b Value from de Vaucouleurs (1961).
Examination of Table 10 shows that the east profile is
substantially different from the west profile when compared at
similar brightness levels. In general, the gradient increases
with wavelength on the east side of the major axis, while no
such trend is recognized for the corresponding sections of the
west side. This implies that the relative contribution of
blue objects to the luminosity at a given position in the east
profile increases with increasing radial distance. With the
exception of the pg data, the gradients on the east side are
all larger than those on the west. Hence, the eastern part of
the disk has a much different morphology than does the western
part, more than can be explained by simply adding a contri-
bution due to a small inner asymmetric arm. If one considers
the appearance of the LMC on the side with its asymmetric arm,
this is not to be unexpected.
69
The gradients can also be expressed in terms of a scale
length. For an exponential luminosity distribution the inten-
sity goes as I(r) = I0exp(-ar), or, in terms of a magnitude,
m(r) = m(0) + 1.086(ar). Table 11 lists the scale lengths 1/a
in kpc for the same data as in Table 10.
TABLE 11
NGC 55 MAJOR AXIS SCALE LENGTHS IN KPC
Color 1/a East 1/a West <1/a> 1/a West
pg 1.81 1.37 1.61 1.81:
V 0.75 1.02 0.87 1.25
R 0.80 1.76 1.10 1.41
I 0.59 1.37 0.83 1.28
Minor Axis
Unlike the major axis profile, the minor axis profile is
almost uniform in its appearance: the "bump" visible on the
north I profile at R. = 26 is due to background noise. All
three colors are well approximated by an exponential decline
in brightness, although the I profile may show a hint of the
beginning of a r0.25 shape. From this we conclude that NGC 55
shows little evidence of a spheroidal bulge component. This
is in keeping with its classification as a late-type galaxy,
which implies that it should exhibit little or no spheroidal
bulge.
Gradients, in magnitudes per arc minute, are listed in
Table 12. The numbers in parentheses are the magnitude inter-
vals used in the calculations of the gradients. The corres-
ponding scale lengths in kpc are given in Table 13.
TABLE 12
NGC 55 MINOR AXIS LUMINOSITY GRADIENTS
Color North" Southa
pg 1.05(23-27)b 0.825(23-27)b 0.9
V 1.81(19-28) 1.56(19-28) 1.
R 1.56(19-28) 1.47(19-28) 1.
I 1.72(19-26) 1.27(19-27) 1.
Notes: a All values are in magnitudes per arc mini
b Value from de Vaucouleurs (1961).
Color
pg
V
R
I
TABLE 13
NGC 55 MINOR AXIS SCALE LENGTHS IN KPC
1/a North 1/a South
0.54 0.68
0.31 0.36
0.36 0.38
0.33 0.44
Mean"
'38(23-27)
68(19-28)
52(19-28)
50(19-26)
ite.
<1/a>
0.61
0.33
0.37
0.37
All of the colors agree in that the gradient of the
southern half of the minor axis is less than that of the
northern half. This is most easily interpreted as the result
of an asymmetric distribution of stars perpendicular to the
plane of the galaxy. Direct evidence of this can be seen in
Figure 2 where faint material is visible either in front of or
to the south of the nucleus. As was noted for the east side
of the major axis, the longer wavelength surface brightness
decreases more rapidly than does the pg brightness. This
implied increase in the relative contribution of blue sources
with distance above and below the plane of the galaxy is
difficult to understand. It is, therefore, worth considering
that the pg photometry may be in error, especially since it
71
was done using techniques that have since been replaced by
more sophisticated methods.
Size of NGC 55
The size of NGC 55 is determined from the maximum extent
of the luminosity profiles. For each color the profiles are
extended to reach the faintest level attained by any one part
of a profile by extrapolating up to a few tenths of a magni-
tude. The dimensions are given in Table 14 where column (1)
lists the passband, column (2) gives the surface brightness at
the measured position, and columns (3) and (4) give the major
and minor axis sizes in arc minutes and kpc. A colon marks
uncertain values: these arise from those profiles with
"bumps" on their lowest portions. As a result, the value of
2a for the red image and 2b for the infrared image are over-
estimates. De Vaucouleurs pg data are again included for
comparison.
TABLE 14
SIZE OF NGC 55
Color L 2a 2b
pg 26: 45' 9'
23.3 kpc 4.7 kpc
V 27.6 35'.9 9'.5
18.6 kpc 4.9 kpc
R 27.8 42'.6: 10'.2
22.1: kpc 5.3 kpc
I 27.0 35'.8 12'.8:
18.5 kpc 6.6: kpc
Note: The pg data are from de Vaucouleurs (1961).
The large extent of NGC 55 in the pg is somewhat remark-
able. De Vaucouleurs reported it could be traced out to an
72
unreliable 10 x 00.2 (31.1 kpc x 6.2 kpc). Apparently, this
outermost part of NGC 55 must be very blue if it is real,
especially on the west side, because it does not appear on the
V, R and I profiles. This point will be considered further
when the colors and integrated magnitudes of NGC 55 are
discussed.
Measured at the 3 a level the V, R and I major axes are:
35'.4 (18.3 kpc), 41'.2 (21.3 kpc), and 37'.1 (19.2 kpc). The
corresponding minor axes are: 9'.1 (4.7 kpc), 9'.5 (4.9 kpc),
and 9'.2 (4.8 kpc).
Asymmetry Profiles
Major Axis
Asymmetry profiles, formed by taking the difference be-
tween the two sides of the major axis luminosity profiles,
Am(N S) = R(N) R(S), are shown in Figures 12, 13 and 14.
Each color shows the inner 7' of NGC 55 to be nearly symmetri-
cal, the two sides differing in brightness by a factor of
about 2 at most. However, beyond 7', the east profile starts
rising while the west profile continues its exponential
decline (see Figures 9, 10 and 11). It is here that the dis-
similarity begins to show and it continues to grow for as far
as the measurements extend. The visual asymmetry profile
shows an almost steady increase in the difference between the
east and west sides of the major axis beyond 7'. The red and
infrared profiles, however, show a small decrease in the
growth rate of the asymmetry starting around r > 11'. This
01
4C
0 14
) 4)
.4.*
rl
.I
0
> 40'.
4)
10
0)
I0
0u
0 "0 4)
0 N"
4)
(M-3)Lu V r v
4)
LC)
'U
4-)
0 4)
to
.44
0r4
8O
4-)
0
z
4)
$44)9
04
44
09
$44 0
0
.4
(Od
4-t)
4)
0
$4
S44
r44~
Ln
0
(M-3)w V
U)
x
$4
0
4)
4*
C0
4n *
0
*4 4J
C-
01
*
0
*-I
'4
*4 4(
0 $
0 0?
'^g ^
^ r
LO >S^1
(M-3)L v
o
-
m
76
coincides with the region where the east side of the lumi-
nosity profile begins its turnover toward an exponential
decline.
Minor Axis
Figures 15, 16 and 17 show the asymmetry profiles for the
minor axis data, Am(E W) = g(E) g~(W). The inner arc
minute in each color is quite uniform. A slight increase in
asymmetry seen at r 1' is attributable to the luminous area
south of the bar seen in Figure 2 and mentioned in the
discussion of the minor axis gradients. All three colors show
the asymmetry increases with increasing distance from the
plane of the galaxy. In each case the southern half of the
minor axis is brighter than the northern half. The large
increase in the I asymmetry profile for r 2 3' may not be real
but the result of photometric errors, and differences in the
slopes of the curves for r 3' may not be significant. The
sudden downturn in the I profile at r 2 4' is caused by the
"bump" on the north luminosity profile at .L 2 26 where the
photometric errors are large.
Color Profiles
Three color indices are formed for each axis: V-R, R-I
and V-I. However, it should be kept in mind that because of
the uncertainty in the photoelectric R and I calibrations, the
possibility of (large) systematic errors in the color indices
cannot be discounted.
(S-N)U V
L.
4-)
-)
.H
o I
rl
0)
0-
+,
S)'
' 0
-4
c t
4)
z44 g
Og
a) -
- o W
Sc0
$.4
4-
41 c
0 0
-4 .a
4da
a4
rXo
b *Oj
41
4
0 0
0
*E4
-4)
00
(%*4
C0
U,
C
4J
.4 m
0
14
0
- g-
04) t
40
(S-N)w V
V-
(S-N)LU V
1
4J
l.4
LO +u
03
*4
4 *4
E-4
lif
r1
4J 44
41
.-!
O:
o o
r-u
0
1.4
r,-I
0 V
4J
o
-m
>,
a)
")
r-.
944
34M
& H
*H*
h *
Major Axis
The major axis color profiles are shown in Figures 18, 19
and 20. As can be seen, the nucleus is not a feature distin-
guished by its color, which is about V-R = 0.33, R-I = 0.40,
and V-I = 0.73. The region within about 5' of the nucleus is
the bright inner disk and bar. It is somewhat bluer than the
outer disk, having a nearly constant V-R = 0.2, R-I = 0.4, and
V-I = 0.6. The prominent red peak to the east of the nucleus
occurs at the edge of the dusty region about 1' across and
clearly visible on the photograph, isophote maps and major
axis luminosity profiles. A valley approximately 3' east of
the nucleus in the V-R profile is an artifact produced as a
result of the removal of a satellite trail on the short
exposure red image. This feature shows as a peak at the same
location in the R-I profile.
Beyond about 4' to 5' east and west of the nucleus, the
outer disk of the galaxy is encountered. Errors in the photo-
metry cause the color indices to be especially untrustworthy
at distances in excess of about 15' on the east side and 10'
on the west, so all discussion will be restricted to this
range. Within these limits, the east profile is redder than
the west, V-I = 0.9 versus 0.8. This is not surprising given
the large amount of visible obscuration on the east side of
the disk.
A prominent feature is seen at 9'.5 east of the nucleus.
This red area is seen as a bright peak on the major axis lumi-
0
Oz
0
z U4)
I
0
$4-
04
$4-
0
co
$4
r4
U)
r4
Ne
0
I.-,
LU N
H
rzd
d-A I
LC
Ln
- C)
z
L
C
I-dl
C\
0
0
U0
U)
z
$4
0
O u
4M
44
0
.-I
0
-'.4
$4
0
*I-
X
$4
O
.e
*l
br
U m I
L-
,
---=~--~
LO
u
z
T-
I-A
0
0
O
84
nosity profiles. It is also visible on the isophote maps as
the west edge of the bright island at about 10', and on the
photograph where some details can be seen. An enhancement in
star density occurs here and de Vaucouleurs (1961) lists two
HII regions here. Color indices of V-R = 0.9, R-I = 0.5 and
V-I = 1.4 indicate this feature is probably one or more
heavily reddened young star clusters. The location would
place it on the near side or the inside of the asymmetric arm.
Minor Axis
Figures 21, 22 and 23 show the minor axis color profiles.
Photometric errors render color measurements unreliable at
distances in excess of 2 3'. In addition, a comparison of the
long and short exposure I north profile shows variations up to
a few tenths of a magnitude in the region of overlap. The two
profiles were joined between 1' and 2' to form the composite
profile and the transition is not as smooth as could be hoped
for. Coupled with the large gradient in the north infrared
profile, the R-I and V-I colors on this side are prone to be
less reliable than their counterparts on the south side.
Within about 1'.2 either side of the nucleus the color is
approximately constant: V-R = 0.2, R-I = 0.4, V-I = 0.6.
Again, the nucleus is not distinguished by its color. A
slightly redder area is seen less than 0'.5 due south, which
appears to be related to the dust in this region.
In general, the south side of the minor axis has nearly
constant or slowly increasing color indices with increasing
85
LO
0
VO
Z
z
^ -Q
0
I4-
0 0
a4
r-4
0
.r1
Z o
0
)- 0.
) N *ri
d --A
y an
-LO
-o
I
0
44
I I
0
o
rl
,-4
O
0
14
0
H
*0
M
4
O
*a
C
C(
- bi
,--14
I
|
