Visual, red and infrared photographic surface photometry of NGC 55 and NGC 253

MISSING IMAGE

Material Information

Title:
Visual, red and infrared photographic surface photometry of NGC 55 and NGC 253
Physical Description:
xiv, 185 leaves : ill. ; 29 cm.
Language:
English
Creator:
Fitzgibbons, Gregory L., 1950-
Publication Date:

Subjects

Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1990.
Bibliography:
Includes bibliographical references (leaves 180-184).
Statement of Responsibility:
by Gregory L. Fitzgibbons.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001613900
notis - AHN8318
oclc - 23536028
System ID:
AA00003745:00001

Full Text









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