![]() ![]() |
![]() |
UFDC Home | Search all Groups | UF Institutional Repository | UF Institutional Repository | UF Theses & Dissertations | Vendor Digitized Files | | Help |
Material Information
Subjects
Notes
Record Information
|
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 (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 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 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 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 |