OPTICAL BRIGHTNESS VARIATIONS IN A SAMPLE OF
NINETEEN RADIO-QUIET QUASI-STELLAR OBJECTS
PATRICIA LOUISE EDWARDS
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
The author is grateful to her advisor, Dr. Alex G. Smith, without
whom the quasar monitoring program would not have existed, and to the
other members of her graduate committee: Drs. Thomas D. Carr, H. L.
Cohen, Guy C. Omer, Jr., Billy S. Thomas and Robert E. Wilson. Robert
Leacock, Richard and Karen Hackney, Roger Scott, Ben McGimsey, Joe
Pollock, and Andy Pica have contributed to the observations of quasars
or related objects which provide a basis for comparisons. The quasar
monitoring program has been supported by grants from the National
Science Foundation. The current grant number is AST8000246.
The author benefited greatly from improvement in photographic
procedures resulting from research by A. G. Smith, R. L. Scott, and Hans
Schrader, and from the development of a microcomputer data reduction
system by Joe Pollock and Larry Twigg. Drs. S. T. Gottesmann, H. L.
Cohen, and H. K. Eichhorn provided most helpful discussions of least
squares curve fitting. The use of a computer program written by Dr. H.
L. Cohen is greatly appreciated.
During her enrollment at the University of Florida, the author has
received support from the Graduate School, the Department of Physics and
Astronomy, and the National Science Foundation. The financial and moral
support of her parents, Dr. Leslie Edwards and Dr. Carolyn Edwards, and
her sister, Dr. Lucy Edwards, is greatly appreciated. The author also
wishes to thank Mrs. Jeanne Kerrick for her advice and encouragement and
Mrs. Edna Larrick for typing this manuscript.
TABLE OF CONTENTS
ACKNOWLEDGMENTS . . . . . . . . .. . . ii
LIST OF TABLES . . . . . . . . .. . . v
LIST OF FIGURES . . . . . . . . . . vii
KEY TO SYMBOLS AND ABBREVIATIONS . . . . . . . x
ABSTRACT . . . . . . . . . . . . .xiii
1 INTRODUCTION . . . . . . . . .. 1
Quasars . . . . . . . . . . 1
RQQSOs . . . . . . . . . . 2
Characteristics . . . . . . . . 4
Questions . . . . . . . . . 5
Samples Chosen for the Present Study . . . 6
2 DATA ACQUISITION . . . . . . ... 11
Florida Monitoring Program . . . . . .. 11
Photographic Techniques . . . . . .. .12
3 DATA REDUCTION AND ANALYSIS . . . . .. 13
Comparison . . . . . . . . . . 13
Iris Photometry . . . . . . . . 14
Programs . . . . . . . . . .. . 14
Errors . . . . . . . . . . 15
Curve Fitting . . . . . . . .. 17
4 FIELD AT 1h +6 . . . . . . . .. 32
SA 94 . . . . . . . . ... . . 32
PHL 938 . . . . . . . . ... . 35
PHL 3375 . . . . . . . . ... .. . 40
PHL 1027 . . . . . . . . ... .. . 45
PHL 3632 . . . . . . . . .. .. . 45
PHL 1186 . . . . . . . . .. .. 54
PHL 1194 . . . . . . . . .. .. . 59
PHL 1222 . . . . . . . . .. .. . 64
PHL 1226 . . . . . . . . .. .. 64
Summary . . . . . . . . . . . 73
TABLE OF CONTENTS (continued)
5 FIELD AT 13h +360
. . . . . . . 75
B 46 .
CONCLUSIONS . . . . . .
Florida RQQSO Results . . . .
Other Studies of RQQSOs . . .
Comparison with QSRS Variability .
Comparison with Quasar Models . .
Summary . . .
BIBLIOGRAPHY . . . . . . . .
. . 155
BIOGRAPHICAL SKETCH . . . . . . .
LIST OF TABLES
1 Sample of Radio-Quiet Quasi-Stellar Objects . . . 8
2 Functional Forms Tested for Curve Fitting Procedure .27
3 Standard Stars in SA 94 . . . . . . . . 33
4 Comparison Stars for PHL 938 . . . . . ... .36
5 B Magnitudes of PHL 938 . . . . . . . .. 38
6 Comparison Stars for PHL 3375 . . . . . ... .41
7 B Magnitudes of PHL 3375 . . . . . . .. 43
8 Comparison Stars for PHL 1027 . . . . . ... 46
9 B Magnitudes of PHL 1027 . . . . . . ... .48
10 Comparison Stars for PHL 3632 . . . . . ... 50
11 B Magnitudes of PHL 3632 . . . . . . ... 52
12 Comparison Stars for PHL 1186 . . . . . ... 55
13 B Magnitudes of PHL 1186 . . . . . . ... .57
14 Comparison Stars for PHL 1194 . . . . . . 60
15 B Magnitudes of PHL 1194 . . . . . . ... .62
16 Comparison Stars for PHL 1222 . . . . . ... 65
17 B Magnitudes of PHL 1222 . . . . . . ... 67
18 Comparison Stars for PHL 1226 . . . . . ... .69
19 B Magnitudes of PHL 1226 . . . . . . ... .71
20 Standard Stars in M 3 . . . . . . . . 76
21 Comparison Stars for BSO 1 . . . . . ... 79
22 B Magnitudes of BSO 1 . . . ... .. . . . 81
23 Comparison Stars for B 46 ...... . . . . 84
LIST OF TABLES (continued)
44 Linear Correlation Coefficients . . . . .. 137
B Magnitudes of B 46 . . . . . . . ... 86
Comparison Stars for BSO 2 . . . . . ... .89
B Magnitudes for BSO 2 . . . . . . ... 91
Comparison Stars for B 114 . . . . . ... .94
B Magnitudes of B 114 . . . . . . ... 96
Comparison Stars for B 154 . . . . . .. 99
B Magnitudes of B 154 . . . . . . ... .101
Comparison Stars for B 194 . . . . . ... .103
B Magnitudes of B 194 . . . . . . . .. 105
Comparison Stars for B 201 . . . . . . .. 108
B Magnitudes of B 201 . . . . . . .110
Comparison Stars for BSO 6 . . . . . ... .113
B Magnitudes of BSO 6 . . . . . . ... .115
Comparison Stars for B 234 . . . . . ... .118
B Magnitudes of B 234 . . . . . . ... .120
Comparison Stars for B 312 . . . . . ... .123
B Magnitudes of B 312 . . . . . . ... .125
Comparison Stars for BSO 11 . . . . . ... .127
B Magnitudes of BSO 11 . . . . . . ... .129
Variability of the Sample of Radio-Quiet
LIST OF FIGURES
Color-Color Diagram for PHL, BSO, B Objects . . .
Average Rms of the Comparison Stars versus Magnitude.
. . 10
. . 18
3 Second Order Polynomial Fit in Magnitude
. . . . . 21
4 Second Order Polynomial Fit in I
5 Third Order Polynomial Fit in Ir
6 Fourth Order Polynomial Fit in I
7 Fifth Order Polynomial Fit in Ir
8 Line Plus Hyperbola . . .
9 K and o Values for Curves . .
ris Reading . .
'is Reading .
ris Reading .
is Reading . .
. . . . . .
SA 94 Field . . . . . . . . . .
PHL 938 Field . . . . . . . . . .
Variation with Time of PHL 938 . . . . .
PHL 3375 Field . . . . . . . . .
Variation with Time of PHL 3375 . . . . .
PHL 1027 Field . . . . . . . . .
Variation with Time of PHL 1027 . . . . .
PHL 3632 Field . . . . . . . . .
Variation with Time of PHL 3632 . . . . .
PHL 1186 Field . . . . . . . . .
Variation with Time of PHL 1186 . . . . .
PHL 1194 Field . . . . . . . . .
Variation with Time of PHL 1194 . . . . .
PHL 1222 Field . . . . . . . . .
. . . 22
. . . 23
S. . 24
S. . 25
. . . 26
. . 29
. . 34
. . 37
. . 39
. . 42
. . 44
. . 47
. . 49
. . 51
. . 53
. . 56
. . 58
. . 61
.. . 63
. . 66
LIST OF FIGURES (continued)
24 Variation with Time of PHL 12
25 PHL 1226 Field . . . .
26 Variation with Time of PHL 12
27 M 3 Field . . . . .
28 BSO 1 Field . . . . .
29 Variation with Time of BSO 1
30 B 46 Field . . .
31 Variation with Time of B 46 .
32 BSO 2 Field . . . . .
33 Variation with Time of BSO 2
34 B 114 Field . . . . .
35 Variation with Time of B 114
36 B 154 Field . . . . .
37 Variation with Time of B 154
38 B 194 Field . . . . .
39 Variation with Time of B 194
40 B 201 Field . . . . .
41 Variation with Time of B 201
42 BSO 6 Field . . . . .
43 Variation with Time of BSO 6
44 B 234 Field . . . . .
45 Variation with Time of B 234
22 . . . . . .
. . . . . . .
26 . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
LIST OF FIGURES (continued)
. . . 124
. . . 126
. . . . 128
. . . 130
. . . . 139
. . . 140
. . . . 141
. . . 142
. . . . 143
RQQSO Sample 144
ic Brightness 146
Ss and RQQSOs 150
B 312 Field . . . . . . . .
Variation with Time of B 312 . . .
BSO 11 Field . . . . . . .
Variation with Time of BSO 11 . . .
Variability Index versus U B . . .
Variability Index versus B V . . .
Variability Index versus U V . . .
Variability Index versus Iex . . .
Variability Index versus v i . . .
Variability Index versus Redshift for the
Variability Index versus Relative Intrins
Variability Index versus Redshift for QSR:
KEY TO SYMBOLS AND ABBREVIATIONS
AB # Objects in the list of Bracessi et al.
B Magnitude in the blue wavelength range
B # Object in the list of Bracessi et al. (1968)
BSO # Object in the list of Sandage and Veron (1965)
B V Color difference, blue magnitude minus visual
C.L. Confidence level of variability P(X )
DEC Declination, angular position in the sky
measured north or south of the celestial
df Degrees of freedom in the X test
fu Flux unit = jan = Watt/m2/cycle per second
f/4 Focal ratio of the Newtonian focus of the
76cm telescope at the Rosemary Hill
Observatory of the University of Florida
GHz Gigahertz = 109 cycles per second
Hz Hertz = cycle per second
lex Infrared excess as used by Braccesi et al. (1968)
IRIS Iris reading on the Cuffey iris astrophotometer
K lo. /(n N)
Ly Lyman alpha, spectral line resulting from loss
of electron from the first shell of the hydrogen atom
KEY TO SYMBOLS AND ABBREVIATIONS (continued)
M 3 Third object 'n the Messier list, a globular
cluster at 13 40m +28.60
mfu Milliflux unit = 103fu
MHz Megahertz = 106 cycles per second
mm Millimeter = 10-3m = 10-1cm
MWP-2 Developer (Difley, 1968)
n Number of standard stars
N Number of unknowns in least squares curve fitting
OVV Optically Violent Variable, a subset of QSOs which show
variations in brightness greater than one magnitude on a
time scale of days
PET Microcomputer manufactured by Commodore
PHL # Object in the catalog of faint blue objects with small
Quasar Acronym for "Quasi-Stellar Radio Source" now used for
quasi-stellar objects with or without radio emission
QSRS Quasi-Stellar Radio Source
QSO Quasi-Stellar Object, an object whose optical image is
"star-like" and whose redshift is extragalactic
R.I.B. Relative Intrinsic Brightness = B 5 log z
RA Right Ascension, angular position measured in hours,
minutes and seconds of time, measured eastward from the
rms Root mean square, average deviation of the comparison
stars from the curve
RQQSO Radio-Quiet Quasi-Stellar Objects
SA 94 Mt. Wilson Selected Area number 94, at 2h53.3m +020'
KEY TO SYMBOLS AND ABBREVIATIONS (continued)
U Magnitude in the ultraviolet wavelength range
U B Color difference, ultraviolet magnitude minus blue
U V Color difference, ultraviolet magnitude minus visual
V Magnitude in the visual wavelength range
V.I. Variability index = X 2/df normalized to 30 df
z Redshift = AX/ Ao
3C Third Cambridge catalog of radio sources
A Difference in wavelength
X0 Wavelength measured in rest frame
0 Average deviation of the standard stars from the curve
X Chi Square
Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
OPTICAL BRIGHTNESS VARIATIONS IN A SAMPLE OF
NINETEEN RADIO-QUIET QUASI-STELLAR OBJECTS
Patricia Louise Edwards
Chairman: Dr. Alex G. Smith
Major Department: Astronomy
A photographic program, monitoring changes in the optical
brightnesses of nineteen radio-quiet quasi-stellar objects (RQQSOs), was
carried out at the Rosemary Hill Observatory of the University of
Florida. This study was done in conjunction with continuing variability
studies of quasi-stellar radio sources (QRSRs) and related objects. The
RQQSO observations cover the years 1974-1978.
The RQQSOs in this sample are located in two 60 fields, one
centered at lh 36m +60 and the other at 13h +360. They had optical
spectra and U B and B V colors similar to those of the QSRSs, but
had no radio flux above the detection limits of the early radio
surveys. One object, B 194, has subsequently been detected in the
radio. Another, B 234, was suggested as a possible detection.
Monitoring of the optical brightnesses was done to study the
characteristics of the variability of the RQQSOs for comparison with the
much larger sample of the QSRSs. A variability index (V.I.) was
computed to facilitate numerical comparisons and to check for
correlation between the extent of variability and other properties of
the quasi-stellar objects (QSOs).
Observations of each QSO and a sequence of nearby comparison stars
were taken on hypersensitized Kodak 103-a-0 photographic plates in
sealed cassettes with B filters at the f/4 Newtonian focus of the 76-cm
Tinsley reflector of the Rosemary Hill Observatory.
Of the nineteen RQQSOs studied, eight did not show evidence of
variability at a confidence level of at least 95 percent. Eight objects
(PHL 3632, PHL 1186, PHL 1226, BSO 1, B 154, B 234, B 312, and BSO 11)
were variable at a confidence level greater than 95 percent. An
additional three objects showed variability at a confidence level
greater than 99.9 percent. These strongly varying objects were
PHL 1194, B 46, and B 114. The proportion of RQQSOs which show
variability is similar to that of the QSRSs.
Sharp drops in magnitude seemed to be more common in RQQSOs than
were sudden brightenings. Such drops occur less frequently in QSRSs.
No correlation between variability indices and U B, B V, U V
or Iex colors was found. A slight correlation between variability and
the v i given by Braccesi et al. (1970) for the B, BSO sample may
suggest that variability is enhanced for objects which are brighter in
There is a correlation between V.I. and magnitude and between V.I.
and redshift, in the sense that fainter and closer objects are more
variable. These two properties are coupled through a selection effect
due to the cut-off in apparent B magnitude, since the fainter RQQSOs can
only be seen at low redshift. Thus, it is difficult to determine
whether the increased variability is due to greater age or lower
In the 1950's, as radio telescopes became more sensitive, several
radio surveys were made of large areas of the sky. Positions of radio
sources were published in various lists, denoted as MSN, 3C, PKS, NRAO,
AO, Bl, CTA or CTD. Since radio telescopes operate at much longer
wavelengths than optical telescopes, their spatial resolution is much
less precise. Therefore, the radio position actually gives an area on
the sky, usually referred to as the "radio error box," in which the
source is located. Identifying the optical object corresponding to the
radio source can be quite difficult, since there may be many objects
within the radio error box. Many of the radio sources showed extended
distributions and corresponded to nebulae in our galaxy or to external
galaxies. However, some of the radio sources were "star-like,"
unresolved at radio wavelengths. These sources were called "quasi-
stellar radio sources," which was often shortened to "quasars."
In order to identify the optical image of the radio sources, much
more precise radio positions were needed. For some sources near the
ecliptic, this was accomplished by timing lunar occultations, which
gives very precise positions. In other cases, two radio telescopes were
used together as a radio interferometer. At some of the improved
positions the only optical object was a faint star-like object. One of
these radio sources, 3C48, was shown by Matthews and Sandage (1963) to
correspond to a 16th-magnitude object with a stellar appearance.
Spectra of this object showed that it did not have the spectrum of a
normal galactic star. At some wavelengths, there were broad emission
lines whose presence could not immediately be explained. A precise
position for 3C273 was obtained by Hazard et al. (1963) by means of
lunar occultation. At this position was a 13th-magnitude star-like
object, showing similar broad emission lines, some of which Schmidt
(1963) identified as the Balmer lines of hydrogen and a line due to
ionized magnesium, all redshifted by the factor z= AX/0o = 0.158. It
was then shown by Greenstein and Matthews (1963) that these same lines
appeared in 3C48 at a redshift of z = 0.367. Following this, many more
star-like objects were identified with radio sources. However, many
sources remained for which the radio error boxes were still quite large.
By studying the photometric properties of the quasars, Sandage and
Veron (1965) hoped to learn to make better guesses at the optical
identifications, which would still have to be confirmed by obtaining
spectra. They realized that the quasars were bluer than most stars, and
in particular that they occupied an area of the U B, B V color-color
diagram near the black body line, and separated from the main-sequence
stars (Fig. 1). Sandage and Veron (1965) used a double-exposure, two-
filter photographic method, producing ultraviolet and blue images
separated by a small displacement. Any object with a brighter-than-
normal ultraviolet image would be noticeable and would be a good
candidate for spectral confirmation as the quasar. When they applied
this technique to several fields in which the quasar had not yet been
identified optically, they were surprised to find extra objects,
"interlopers," with ultraviolet excesses but not near the radio
position. Realizing that these objects were probably related to the
faint blue objects at high galactic latitudes found in earlier surveys
by Iriarte and Chavira (1957), Haro and Luyten (1962), Humason and
Zwicky (1947), and Feige (1958), Sandage (1965) studied the space
density of these objects with respect to their apparent magnitudes.
These results led him to suggest that while the brighter of these blue
objects were galactic stars, most of the fainter ones were extragalactic
and could be expected to show large redshifts. When spectra were taken
of several of these objects, three showed extragalactic redshifts
(Sandage, 1965). The spectrum of BSO 1 was indistinguishable from those
of the quasi-stellar radio sources (QSRSs) and had a redshift of
z = 1.241. These extragalactic objects not associated with radio
sources are usually referred to as radio-quiet quasi-stellar objects or
In order to obtain a larger sample with which to study the space
density of these objects, Sandage and Luyten (1967) carried out
photometric studies on some of the blue objects found earlier in the
Palomar field at 1h36m +6 by Haro and Luyten (1962) using a U, V offset
method. These objects are referred to by their PHL numbers. Spectra by
Sandage and Luyten (1967), and by Burbidge (1968) confirmed many of
these as QSOs. Others proved to be galactic subdwarf stars. Braccesi
(1967) suggested that QSOs should also be brighter than the subdwarfs in
the infrared and that this could be used to weed the subdwarfs out of
the sample. The U, B offset plates taken of the field at 13h+360 by
Sandage and Veron (1965) were used to locate the objects which had
ultraviolet excess. This plate was then compared by Braccesi, Lynds and
Sandage (1968) with an infrared plate of the same field. Of those
objects with a strong ultraviolet image, those which also showed strong
infrared images were predominately QSOs, as confirmed by their
spectra. These objects were listed as BSO or B objects. The more
complete list, covering the same area, by Braccesi, Formiggini and
Gandolfi (1970) has later been referred to by AB numbers. These surveys
suggest that the RQQSOs are much more numerous than the QSRSs. Since
"radio quiet" simply means not yet detected in the radio, some of these
RQQSOs may be detected at radio frequencies as the sensitivity of radio
telescopes is increased.
As more and more QSOs were identified and studied, it became easier
to describe them as a class. They are star-like objects often
coincident with a radio source. The spectra show broad emission lines
redshifted by large amounts. Absorption lines may also be present.
These two properties are enough to identify a candidate as a QSO, but
there appear to be other common attributes. QSRSs are strong sources in
the ultraviolet (Sandage, 1965) and in the infrared (Braccesi et al.,
1968). These characteristics were useful in choosing candidates for
spectra to confirm their extragalactic nature.
A fourth property possessed by many of the quasars is a variability
in optical and/or radio output. This possibility was recognized when
the magnitude of 3C48 was found to be different on a subsequent
observation (Matthews and Sandage, 1963).
There are two different directions in which QSO studies have been
aimed, based on the second and fourth properties. Because of the large
redshifts, QSOs were suspected to be the most distant objects observable
at the time. Therefore, they might be used to extend our baseline for
the investigation of cosmology--for example, determination of the Hubble
constant and the age of the universe (Sandage, 1972).
In cosmological studies the variability of QSOs proved to be a
nuisance because of the resulting uncertainty in absolute magnitude and
energy output. In other researches variability provided valuable
information because the radius of the active region is limited to the
distance that light can travel during the time scale of the variation.
Results of variability investigations imply that these sources are quite
small by galactic standards, which makes their tremendous energy output
even harder to explain. In addition, their relative energy output
across a wide spectrum, and relative changes in this energy
distribution, should contain information on the energy generation and
transformation mechanisms in the immediate vicinity. In particular, two
classes which have very different spectral energy distribution might
show different types or degrees of variability. Quasars typically have
their energy peak in the short radio range while RQQSOs have no
detectable emission in this range. If the source of the QSRSs'
variability were in the radio range, with much of the energy then being
transferred to optical wavelengths through various physical processes,
it might be expected that RQQSOs having little or no radio output would
be less likely to vary.
Samples Chosen for the Present Study
In order to study the optical behavior of RQQSOs, one or more
consistent samples with well-defined optical properties were needed. A
spectrum showing the object to be a quasar was the second requirement.
The existence of readily available photographic finding charts was
another major consideration. Two groups of objects, conveniently 12
hours apart and accessible during different times of the year, fit these
criteria. Eight objects in the PHL field Ih +60 studied by Sandage and
Luyten (1967) had spectra and finders given by Kinman (1966), Burbidge
(1968), and Burbidge et al. (1968). Redshifts were available for 11 of
the objects in the 13h +360 field whose finders were published by
Braccesi et al. (1968).
In the first of these fields the selection was based on ultra-
violet excesses only. In the second field infrared excess was also
considered in choosing which spectra should be taken. There was
probably also a preference toward the brighter objects. The limiting
magnitude is approximately 1 mag. fainter for the objects in the second
field. Positions and colors are given in Table 1 for the RQQSO sample,
which includes objects from both fields. These objects occupy a region
in the U-B, B-V color-color diagram shown in Fig. 1.
Sample of Radio-Quiet Quasi-Stellar Objects
Object Designations RA (1950) DEC (1950)
17, B 87
90, B 243
168, B 416
1 28 24.0
1 30 30.0
1 39 54.0
1 47 36.0
1 48 42.0
1 51 12.0
1 51 48.0
12 46 28.7
12 46 29.6
12 48 17.7
12 52 57.9
12 55 2.1
12 56 7.8
12 57 26.7
12 59 30.9
13 0 42.5
13 4 52.1
13 11 19.5
1 56 00
7 28 00
3 22 00
6 10 00
9 01 00
9 02 00
4 48 00
4 34 00
37 46 50
34 40 49
33 47 11
35 55 24
35 21 21
35 44 54
34 39 31
34 27 19
36 07 34
37 28 38
36 15 40
TABLE 1 extended
V B-V U-B Z Refs.
f, g, h
a Kinman, 1966.
b Burbidge, 1968.
c Sandage and Luyten, 1967.
d Burbidge, 1967.
e Braccesi et al., 1968.
f Sandage and Veron, 1965.
g Sandage, 1965.
h Braccesi et al., 1970, 1973.
0 I 0
/ \ Fu-
Florida Monitoring Program
A photographic program monitoring the optical brightness of quasars
and related objects has continued since 1968 at the University of
Florida's Rosemary Hill Observatory located near Bronson, Florida.
Four-inch by five-inch photographic plates are exposed in a camera
located at the f/4 Newtonian focus of the 76-cm Tinsley reflecting
telescope. Summaries of the results of this monitoring program have
been published by Folsom et al. (1971), McGimsey et al. (1975), Scott
et al. (1976), Pollock et al. (1979), and Pica et al. (1980).
Since the telescope is rather small for use on such faint objects,
special efforts are made to maintain the quality and improve the speed
of the photographic plates. The plates are exposed in sealed cassettes
filled with dry nitrogen gas to prevent differences in the plate
responses due to oxygen and excessive humidity in the Florida air. The
plates are hypersensitized to provide an increase in speed without an
unacceptable increase in background density. Scott and Smith (1976)
showed that baking Kodak 103a-0 plates in an atmosphere of dry nitrogen
increases their speed. Subsequent evacuation and soaking the plate in
hydrogen gives an additional gain in speed. The plates are then kept in
a dry nitrogen atmosphere during storage, use and exposure. The plates
are developed 9m in MWP-2 (Difley, 1968) with mechanical agitation to
insure evenness of development.
For the RQQSO monitoring program the Kodak 103a-0 plates were
exposed through a Schott GG 385 filter which restricted their
sensitivity to the blue (B) region of the spectrum. This helped to
minimize possible color effects due to atmospheric extinction. The
RQQSO and an approximately one-half-degree field of stars surrounding it
were recorded on the photographic plate.
DATA REDUCTION AND ANALYSIS
A preliminary estimate of the brightness of the RQQSO can be made
almost immediately by comparing the image with a pair of neighboring
stars of similar magnitude, one of which is usually slightly brighter
than the object and the other slightly dimmer. Pairs of plates
suspected of showing variation can be studied using a blink comparator.
For numerical comparison a group of twelve or more stars is chosen
near the RQQSO covering a range of 3-4 magnitudes around that of the
object. The magnitudes of these comparison stars are obtained by
photographic transfer (equal-length exposures of the two fields on
different areas of one plate) from a nearby area for which photographic
or photoelectric magnitude sequences have been published. Since the two
RQQSO fields are very far apart, two calibration fields were needed.
Photographic determinations for the stars in SA 94 (Purgathofer, 1969)
near the 1h +60 field and a photoelectric sequence of the stars in M 3
(Sandage, 1970) near the 13 +360 field were conveniently placed. The
transfer process involved at least one exposure of the calibration field
for each RQQSO in the nearby field. Since a number of exposures of
these fileds were made it was possible to smooth the published magnitude
sequences for each of the two calibration fields to correct for any
secular changes since publication for any field effects in the
The prime reduction tool is the Astro-Mechanics Cuffey Iris
Astrophotometer, which measures how much light is blocked out by the
photographic image. An iris reading reflects the size of the iris
opening required so that the light passing through the iris and then
through the image on the plate is equal to that in a reference beam.
When these readings are made with the iris centered first on the object
and then on the stars of the comparison sequence, a curve is obtained
that represents the relation between iris readings and the magnitudes of
the comparison sequence. An example of such a curve is shown in
Fig. 3. The crosses represent the iris reading-magnitude points of the
standard stars in M 3 on a plate taken on June 6, 1977. The circles
show the parabolic curve fitted to the points using a least squares
A least squares program of Pollock et al. (1979) fitting a
parabola to the iris reading-magnitude curve is run on a Commodore PET
microcomputer. This program also displays the curve and the data points
on a screen to allow a visual check of the fit. The rms scatter of the
comparison star magnitudes is used as a measure of the precision of the
RQQSO magnitude determined. An expanded version of this program, run on
the Amdahl 470/V6 computer of the North East Regional Data Center
located on the University of Florida campus, provides cumulative data
reduction and allows for iterative smoothing of the comparison sequence.
Further statistical treatment of the first RQQSO magnitudes is done
on the PET, using a program which combines observations made during a
short interval on one night into a single magnitude, and then computes
the weighted Chi Square (see statistics texts such as Aitken, 1952, or
Brownlee, 1960). Chi Square tables then give a confidence level
reflecting how poorly the data fit an assumption that the object has no
intrinsic variation but only reflects the same plate scatter as the
comparison stars. In order to facilitate numerical comparisons between
strongly varying objects, the program uses a mathematical approximation
to Chi Square (X2) function for a given number of degrees of freedom
(df) to compute a variability index (V.I.), which is X2/df normalized to
30 degrees of freedom. A confidence level of 95 percent corresponds to
a V.I. of 1.46; 98 percent to a V.I. of 1.60; 99 percent to a V.I. of
1.70 and 99.9 percent to a V.I. of 2.00. Thus, variability indices
greater than 1.5 indicate real variability. Variability indices greater
than 2.0 reflect not only greater confidence but also greater variation
of the object.
Like any experimental or observational measurements, photographic
photometry is subject to error. There are three types, requiring
different treatments, involved in the data reduction described here.
The first type is a zero point shift occurring due to the
photographic transfer method of calibration. This error affects all of
the comparison stars in a field and so does not affect the variability
determination. It can be ignored except when comparing magnitudes with
those of other observers, measured photoelectrically or on a different
photographic system. No attempt is made at this time to convert the
photographic blue magnitudes of this photographic plate filter
combination to photoelectric magnitudes. A discussion of the relation
between these photographic blue magnitudes and the Johnson B
photoelectric magnitudes is given in Hackney (1973).
A second source of error, inherent in the photographic process, is
caused by granularity of the photographic emulsion and its response to
incident light. The size of these errors can be affected by observing
conditions (camera focus, atmospheric seeing and sky brightness) and by
the plate hypersensitization process. These random errors cannot be
removed, but it is very important to know the amplitude of the random
error since it is this with which the observed variation of the object
must be compared in order to evaluate whether the observed variation is
mainly inherent in the object or could have been entirely due to the
random errors. For this reason it is necessary to remove the third type
of error which, if allowed to remain, will increase the estimate of the
This third type of error acts like an error in magnitude of the
comparison stars. On the calibration plate, the effective magnitude of
the stars as seen by the plate may differ from the photoelectric
magnitude published, due to field effects in the telescope, adjacency
and sky background effects in the photographic emulsion and color terms
due to the difference in wavelength bandpass in the photoelectric and
photographic systems. Since each of the two calibration fields is used
for a number of sources, the constant terms can be separated from the
random terms and then smoothed out of the calibration sequence. An
iterative smoothing is also performed on the comparison stars in each
RQQSO field to prevent the random deviations in the calibration exposure
from contributing to each of the other observations. This also reduces
the effective weight of the calibration plate to its proper level.
The mean rms deviation of the comparison stars is used to
approximate the expected random error of the QSO in the X2 test. Since
the QSO may be somewhat brighter or fainter than the average comparison
star, this approximation might be expected to underestimate slightly or
to overestimate the random error of the QSO if the slope of the iris
reading-magnitude curve is not constant. The distribution of rms
deviation of the comparison stars in the combined RQQSO, quasar sample
was examined (see Fig. 2). The linear correlation coefficient is only
0.17, with a slope of 0.02 rms per one magnitude change in brightness,
which is equivalent to 0 09 over the range of magnitudes covered in this
sample; thus, this is not an important source of error in this
Since the form of the mathematical equation used in the least
squares curve fitted to the iris reading magnitude points may have an
effect on the magnitudes and associated error so determined, an
_ 0 0 0
investigation was made of the various forms and methods used for curve
fitting in iris astrophotometry programs.
In order to derive a magnitude from the iris reading of the object,
a curve is fitted to the iris reading magnitude points for the
comparison stars using a least squares procedure. A variety of systems
have been used. When the magnitudes of the comparison stars are not
known, a scale in terms of average iris reading is often used. Such a
procedure is described in Penston and Cannon (1970). It is difficult to
compare the amplitude of variability expressed in this way with results
of any other studies. In addition, variability thus determined is
suspect unless plate exposure, treatment, and sky conditions are
sufficiently uniform that the slope of the iris curve is essentially the
same on all plates.
Even when the magnitudes of the comparison stars are known or can
be determined, many different forms and methods have been used to fit
the iris curve, including a line fitted graphically by eye, a least
squares line, a smooth curve of unspecified shape fitted by eye, and
polynomials of varying degrees fitted by computer least squares
programs, some of which use iris reading as the independent variable and
some using magnitude. The RQQSO observations were reduced using the
same program as the Rosemary Hill Observatory quasar observations to
facilitate comparison between these two classes of object. The computer
program used in the data reduction to fit a parabola to the magnitudes
and iris readings of the comparison stars (for an example see Fig. 3)
assumes that all random errors occur in iris readings. The smoothing
technique assumes that constant terms are errors in magnitude and
smooths them out iteratively.
The analytical form of the iris reading-magnitude relation produced
by the Cuffey Iris Astrophotometer is not known. Empirically the curve
is essentially linear over most of the range of interest, but the
magnitude should increase asymptotically as the sky background makes an
increasing contribution to the photographic image.
In order to compare the different curve fitting methods and to
investigate the possible effects on variability determinations, a
calibration plate of the M 3 field taken on the night of June 6, 1977,
was reduced using a variety of analytical forms for the iris reading-
magnitude relation. A least squares curve fitting computer program
written by Dr. H. L. Cohen (1980) was used for polynomial forms and a
modified version to fit a line plus hyperbola. The resulting curves are
shown in Figs. 3 8. Results and comments are given in Table 2. In
these figures crosses represent the iris reading-magnitude point of the
standard stars in M 3. The circles show the least squares curve fitted
to the crosses in each case.
In Fig. 3 it is assumed that all of the random errors are in iris
reading and that iris reading may be expressed as a second order
polynomial in magnitude. This is the form used in the RQQSO and quasar
studies done at the Rosemary Hill Observatory of the University of
Florida. Three unknown coefficients are determined in the least squares
procedure. In order to determine the magnitude of the QSO from a known
iris reading, this form must be inverted and the correct root chosen.
3G]nl N 9VtN 9
c~r ~u O~ m N
< + a)
%( 2 *o^^F
EOcn lh NOV~ N
o ( )
+ + n
+ n O 0
^ D^0^ -
3On IINVP Q
(( -- i
^ 5 ^
"~~~I j *..-
E Bc fINOVN'4 8
o o 0 N
+ I +01
o a N )
< U) -0 N
-i m m
11 a L "L
Lj 0 L o
-J + 0n
pq + -
a 01 o0
en In ,
Ll > J4:
0) C) 0
a3 m I
10 r U
H tdn E
U7 3 41
< m to
'1 m N m m E (N In 44
4:1- 0 -'
Higher order polynomials of this type become quite difficult to invert
to determine the magnitude of the QSO.
Other observers, Penston and Cannon (1970), Braccesi et al. (1970),
Burkhead and Seeds (1971) and Bonoli et al. (1979) use computer least
squares fits expressing magnitude as a polynomial in iris reading. This
form does not require inversion to determine the magnitude of the QSO,
but this form assumes that all of the random errors occur in
magnitude. This seems unrealistic since the magnitudes of the standard
stars are usually determined photoelectrically with quite high
precision while the iris reading is subject to the noise in the
photographic emulsion. Figures 3 6 show magnitude expressed as second
through fifth order polynomials in iris reading (3 6 unknowns).
Penston and Cannon (1970) use a second order polynomial in iris
reading and many observers have followed their method. Burkhead and
Seeds (1971) investigated computer fitted polynomials in iris reading
for a field with more than one hundred stars. They found the best
computer fit (minimum K = Zoi/(n-N)) with a fifth order polynomial, but
they usually used a third order polynomial for cases with a small number
of stars or fit a smooth curve by eye. Braccesi et al. (1970), also
with a very large number of standard stars, found a minimum at third
order although the third order terms were small. The M 3 field studied
here had minimum K at first order (a line). A linear form does not fit
near the sky background causing the magnitude to be underestimated.
With only 15 standard stars, high order terms are susceptible to
unreasonable curvature in the range where the curve should be linear.
E L* H
i i i
4 3 2
I I ...
2 34 5
Fig. 9. K and a Values for Curves.
In a search for a relatively simple mathematical form which would be
linear over most of the range but would increase asymptotically as the
sky background was reached, one such form was found: a line plus
hyperbola (Fig. 8, L + H in Fig. 9). It has three unknown coefficients
and an equation: M = A(1) + A(2)x(IR SKY) + A(3)/(IR -SKY). It fits
slightly better than the second order polynomial curves and has the
proper behavior at the sky background. It does not require inversion to
determine the object magnitude. It does have two disadvantages. The
iris reading of the sky background must be known. This quantity was not
measured on many of the old plates. Also, the line plus hyperbola has
magnitude as a function of iris reading. In a linear least squares
curve fitting program, this assumes that the random errors are in
magnitude. The line plus hyperbola may be promising for future data
reduction of the total RQQSO and quasar samples if combined with a curve
fitting program which will minimize residuals in both variables
Numerical comparison of the goodness of fit of the various
mathematical forms tested on the M 3 field is shown in Table 2 and
Fig. 9. Two quantities are calculated C, the average rms scatter of the
points about the curve (expressed in magnitude units), and K, which is
ZEC/(n-N), where N is the number of unknown coefficients and n is the
number of standard stars (in this case n = 15).
Where the minimum K value occurs seems to depend on the particular
field and on how many comparison stars there are. A linear form is too
simple and has the wrong behavior at the sky limit. A parabola in iris
reading also does not approach the sky background properly. Most quasar
monitoring fields have only 10-15 comparison stars, in which case high
order polynomials allow too much curvature in the middle range. The
line plus hyperbola would allow a better shape than a parabola in
magnitude, but requires a more complex least squares curve fitting
All of the RQQSO data described in the following chapters were
reduced using a parabola in iris reading, in order to be consistent with
the reduction of the larger sample of quasars observed at Rosemary Hill
Observatory and given in Pollock et al. (1979) and Pica et al. (1980).
Use of the line plus hyperbola form will require re-reading of many
older plates and development of a new curve-fitting program which will
minimize the residuals in both variables simultaneously.
FIELD AT 1h +60
The comparison stars for the RQQSOs in the PHL field at 1h +60 were
calibrated using stars in SA 94 centered at 2h 53.3m +00 20'. A
photoelectric sequence giving V magnitudes and B V colors for 54 stars
in SA 94 was published by Purgathofer in 1969. This information allows
calculation of the B magnitudes of these stars. Eight observations of
this field were made in calibrating the PHL objects. A sequence of 15
stars was used in each case. The magnitudes of these stars were
iteratively smoothed to accommodate the field response of the
telescope. The stars are identified in Fig. 10 Their designations and
old and new magnitudes are listed in Table 3. Some random zero-point
offset in the magnitudes in a RQQSO field may occur in the photographic
transfer method, but this should be small and has no effect on the
amount of variability measured. It would be relevant only when
comparing particular Rosemary Hill Observatory magnitudes with
magnitudes measured elsewhere. A photographic B magnitude is not
exactly the same as a photoelectric Johnson B magnitude, but it is the
same for most photographic observers (Hackney, 1973).
Ho v) n H H o H Ho %D o HHc HD H
Hm r rN 4 N 44 r-4
r M CD -4 C) NQ ID co N '- OD M ) r-4 04: M H
41 4! 44) ) 4!o '0 co m '0 '0 N N C)
HHr- HHH HH H -4 H H
N C)C) 444~ N C)C) ) H 44 '4
0 V 0
C.Q U >
44 r- 4j .
4-1 :4 C)
4 -) 4 '
c5-4 4 p
4:' 0U E-
44 i 4:
PHL 938 is a blue star-like object found at Oh 58m 12s and +10 56
by Haro and Luyten (1962). Spectra (Kinman, 1966) based on oxygen,
carbon and hydrogen lines gave a redshift of 1.93 with some absorption
lines suggested. Kinman photographically determined a visual
(V) magnitude of 17"16, B V color of 0'32 and U B = -088. Since no
radio source was known at this position, a special search was made by
Bolton (Kinman, 1966), who could not find any radio emission at 11 cm
down to a detection limit of 0.1 fu. Because of the absorption lines,
Burbidge et al. (1968) took more spectra, finding an emission redshift
of 1.955, an absorption in the Lyman alpha (Ly ) hydrogen line at 1.906
and a series of absorption lines due to iron and magnesium at a redshift
The comparison stars for the PHL 938 field were calibrated by
photographic transfer from the sequence in SA 94 on a plate taken the
night of January 4, 1976. These stars are identified in Fig. 11 and
their magnitudes are given in Table 4. The Florida results were based
on 13 observations, which include two pairs of exposures taken close
together in time. The magnitudes of PHL 938 are given in Table 5 and
plotted versus time in Fig. 12. They do not show any evidence of
variability. The mean magnitude is B = 16.85 with a range of 0.25
compared to an rms scatter of the comparison stars of 0o08 (Table 43).
C *- r-I
0 0 -0
0 0 4-J
1 0 u4
04 ( 0
0 0 *v
o 0- C
4, 0 >M
o <) -1
0 0 .
( 4J 0
Oi 0 0
0 0 0
.n 0 ) 4
.0 4 e
-. 0 .
0 Li 0 0
m0 0 H '0
1- C )
. O 4
10 m m ro m m '
0 0 0 rL00
o,' mmm om 4 OW
in ma ^o ocomrN C N
m mr N cr>m cco Nm o n n nmo
0 F) 4J 0 .
0 H W C
"-I 4 o 0 4 o F -o kDinr0) r- 41 0 0
40 ~ rc r i r- co 4 04
C' O a) -) q -A Cq o o) C ) 0 04
4~) a) OP O440
oll o 0 u
.01 0 w
(' U 4 0
r* l '
]> () ^CO (ri
PHL 3375 is a blue star-like object located at Ih 28m 24s and
+70 28' in the Palomar survey field of Haro and Luyten (1962). It was
suggested as a QSO candidate by Sandage and Luyten (1967), who measured
its magnitude (V = 18"02) and colors (B V = 029, U B = -0T51).
Spectra by Burbidge (1968) showed emission lines of oxygen, magnesium
and hydrogen at a redshift of 0.390. A radio survey by Fanti et al.
(1977) found no radio emission at 1.4 GHz to the limit of the survey
The comparison stars in the PHL 3375 field were calibrated by
photographic transfer from SA 94 using an exposure of each field on a
single plate taken on the night of December 31, 1975. These stars are
identified in Fig. 13 and their magnitudes are listed in Table 6.
Eleven observations of PHL 3375, including one pair taken one after the
other, were made at Rosemary Hill Observatory. The coverage was
somewhat hampered by Jupiter's passage through the field during August
and September of 1975. The resulting magnitudes are displayed in
Table 7. The variation of the object with time is shown in Fig. 14
PHL 3375 was slightly brighter in 1975 and 04 dimmer in 1978.
Statistics show that this object is probably variable (C.L. = 92%). It
has a mean magnitude of 17T94 with a range of 0"66. The average rms
scatter of the comparison stars was 011 (Table 43).
rl 0 -H )fl
'.) C 4
oJ r 0
M'C 41 4J
-H 0 0
0 C CD
H -4 -j - - *0- 1 wo
O -0& 0
S0 0 Cm4m
M cD m ilc L n m D
) mee H C 0 01
0 \ U) 4
0 H, N-0 00 4 _o
0 -*(0 u0 1
to m \rc-,)N-Jo -~ 4oJm 14 to
U c a r r- r. r- 4 r r m (D ri
-I tl 10
t o/ o Q o m o ^ o 'a Ln m to fto O-n
O to'0 -4
o U to 0
4' 0 tU
Located at Ih 30m 30s and +30 22', PHL 1027 was a blue object on
the Palomar survey plate studied by Haro and Luyten (1962). Photometry
by Sandage and Luyten (1967) found colors suggestive of a QSO;
V = 17?04, B V = -0"03 and U B = -0"77. An emisson line redshift of
z = 0.363 based on neon, oxygen and hydrogen lines was determined by
The comparison stars for PHL 1027 were calibrated by photographic
transfer from SA 94, using a plate taken on the night of November 21,
1976. These stars are identified in Fig. 15 and their magnitudes are
given in Table 8. The magnitudes from 13 observations of PHL 1027,
including 2 pairs, are given in Table 9 and are plotted according to the
date of observation in Fig. 16. The points scatter about a line and
show no evidence of variation. The mean magnitude of PHL 1027 is 16 83,
with a range of 0?22 compared with the average rms scatter of the
comparison stars of 0'09 (see Table 43).
PHL 3632 is listed at Ih 39m 54s and +60 10' by Haro and Luyten
(1962). Sandage and Luyten (1967) obtained the following magnitude and
colors: V = 18T15, B V = 0"13 and U B = -0'75. Since its colors
were in the QSO region of the color-color diagram (see Fig. 1), Burbidge
(1968) took a spectrum of PHL 3632. Lines due to carbon implied a
a fa c-
C0 0 >
CU).c ( -1
'0 -I > 0
a 4-V (-
rS 4J 1-1
. 1 -m
c C a) El
OJ (r- c
.0> C 'C
eO .c -'
) 4-, '-
U 04 u
0 0 4
C: 0 -..
--- m4 m 4J W .
a C: O o
a ,-4 -) i ,- -- ,-n D E w
( 4 1
H mme amooa-0440
0 HmC'mm4m(.m-4 4
r \\ H\
M 0 L0
C- in ID IT on co m r- M 41
4 C C00
U C) co a, ) C> H -o 1( N % 4 -4 00 .1 f4
a) a)4 4
1 c Qi +
'-i 4 wU
4 m 4 NOOU)W~HH'ONNL 44 wO
C .C 3
0 C 3
(0 4 H
C4 In H
a1) a) 1
fU 0 4
0 .- 0 m
u I-- -H H
I H (n
( u m (o
r 0 0
- .0 0 M
q " = M
m U).- 0
0 0 0
4- 0 0
a O -n -
*H e u0
(0 4J l-
> 0 0
0 1 1 0 T
m oo .o o4 -
0 40 C
m ar mo c n crH om0n -n.-' 0
cl, CDr -) )C o a)m
4 0 LoHH He 0 0
*0 i-i oL 0
S. . ,
0-4o o "CN N N o VC4Lt0N _o 0
(0 mOum -Hi00'C )a)O 0 X:
(' I J 0.44 4f
0 4. 4J
o on m oi r- -cr-- m o v 0 O-I
W -A (N1 -4 C4
SU "I 0l La i
-4 0- o r-' M (o H 'U
0 0 0)4-T
>i 3N~~~m c
U 4J i-
redshift of approximately 1.5. The object was definitely a QSO but the
redshift was not as precisely measured as most of the others.
Comparison stars for PHL 3632 were calibrated by photographic
transfer from SA 94, using a plate taken on the night of December 24,
1975. These stars are identified in Fig. 17 and the magnitudes are
given in Table 10. The Florida results contain 11 observations of
PHL 3632, including two pairs. These magnitudes are given in Table 11
and are plotted in Fig. 18. The object appeared brighter in 1975 and
had faded in 1978. PHL 3632 has a mean magnitude of 17"59, with a range
of 0P57 and an average scatter of the comparison stars of 008. X2
statistics give a 98 percent confidence of variation (Table 43).
PHL 1186 was a stellar object found in the Haro and Luyten (1962)
survey because of its ultraviolet excess. Its position is 1h 47" 36s
and +90 1'. Burbidge (1967) gives its magnitude and colors as
V = 174, B V = 0 02 and U B = -083. Emission lines of neon,
oxygen and hydrogen show a redshift of z = 0.270 (Burbidge, 1968).
The comparison stars for PHL 1186 were calibrated by photographic
transfer from SA 94, using a plate taken on the night of December 24,
1975. They are identified on Fig. 19 and their magnitudes are given in
Table 12. Sixteen observations, including 4 pairs, were made at
Rosemary Hill Observatory. The resulting magnitudes are given in
Table 13. PHL 1186's variation with time is shown in Fig. 20. The
object brightened between 1974 and 1975 and remained at the higher
0 *'H -H
U fU O
0 T .0
49- C H
) c4 )v
u) 0 a)
0 4J W E
r- ) C
0 0 w
CO to 0
C -U C
a C aC
P r- r- C
W 4-4 C)
40 0 >
)( .cr- C
-4 C C
-r4 C) 'C
r-l >. -U 0)
C) 0 ,-
U) c Um N o c c n r-~ e m o me m o N ) L
S o oo o m io w i-l oO rq -4
0 l Um
4H LI0- *
* ^ cro- () coo J ) cN r c)
a en M 4 r- H4 A p 4 4 r-4 r, rU) M C) -H
w-1- r-0-l l Ii -I -r- .-l-I-4 o )o o.. 'M 4
r 0mN ONu 0) On
Ei0 0Q 4J0
c Nm CmmD04MC) )N M
Ua al3 U C)
0 WH A(D J00H 040 4
Ni W C) -4 W W W W C) U } H M
0H 0) .4
-4 D I (u
0 7 .44
444 4 '-) 4
fd i- r
CN CO O i-l>, N
0H 0 aC 4
N~O~N~m*H 0 aNj
ct~~~ J-'~ L|~i l ~
level. The mean magnitude was 17m51, with a range of 054 as compared
to an average rms scatter of the comparison stars of 0&09. The
confidence level for variability is 98 percent (Table 43).
Haro and Luyten (1962) list PHL 1194 at 1h 48m 42s and +90 2'.
This field and that of PHL 1186 overlap partially. The objects are
17 minutes of arc apart. Sandage and Luyten (1967) give V = 17m5,
B V = -0o07, U B = -0'85 and z = 0.298. Spectra by Burbidge (1968)
show emission lines of neon, oxygen and hydrogen at a redshift of
z = 0.299 and an absorption line which is probably Ly. but might be due
to magnesium. A radio search by Fanti et al. (1977) at 1.4 GHz failed
to detect a radio source at this position with a detection limit of
Comparison stars for PHL 1194 were calibrated by photographic
transfer from SA 94, using a plate taken on the night of December 9,
1974. These stars are identified in Fig. 21 and their magnitudes are
given in Table 14. Results from 13 observations, including 3 pairs, are
tabulated in Table 15. Variation is evident in Fig. 22. The graph
shows PHL 1194 brightening over the period 1974 to 1977 and much fainter
in 1978. The mean magnitude is 17 47, with a range of 0.65 and an
average rms scatter of the comparison stars of OT10 (see Table 43).
Statistics give a confidence level greater than 99.9 percent that this
object is variable.
0 > .
.o 0 (
00 C (l
. on C
(U U14* l
U) -1 0
. 0 >.0
lZ 1) 0
0 4J i-
.O l f ) E-
0 a U
Li H1 *
E-i 0 -
0 W I-
0o 0 U
a mnHHoccHQoo4. m --
o N ooodo o m m o m -
w0 M(nm' M) ON M 000
M 0-C*NA* 0 0 ( N0
a) 0 C.1
(0 t)) 4
0r rmm rm rl r a mO O 0 >
SE-E 0 u
11 1 uC
rql a ) m c D o co rc 0 r0 co m 0, -A -
ia NOO CN OO.4QOOCCN~t 00 4JO
0 0 V-0 4
>1 rl ar +r
"- 4 4 0 0
40r 0 N OO OO O r4' 4
"-4 .0 .
PHL 1222, located at ih 51m 12s and +40 48', was one of the blue
objects found by Haro and Luyten (1962). Photometry by Sandage and
Luyten (1967) gives a V magnitude of 17 63, B V = 0T41 and
U B = 078. Burbidge (1968) determined the emission line redshift to
be 1.910. An absorption line, probably Ly is also present. Fanti
et al. (1977) detected no radio emission at 1.4 GHz greater than 20
mfu. The detection level here is somewhat larger due to the presence of
a nearby radio source.
Comparison stars for PHL 1222 were calibrated by photographic
transfer from SA 94, using a plate taken on the night of December 24,
1975. These stars are identified in Fig. 23 and their magnitudes are
recorded in Table 16. Thirteen observations of PHL 1222, including 3
pairs, were made at Rosemary Hill Observatory. The resulting magnitudes
are given in Table 17 and they are plotted against time in Fig. 24. The
brightness of the object appears to be steady with a possible slight
decline in 1978. No convincing evidence of variation is seen. The mean
magnitude is 17 72, with a range of 0 29 and an rms scatter of the
comparison stars of 007 (Table 43).
One of the objects found by the Haro and Luyten (1962) survey,
PHL 1226 at Ih 51m 48s and +4 34', aroused extra interest due to its
close proximity to a galaxy, IC 1746. Photometry by Sandage and Luyten
C -H 1
04 4j 4Jr
4J C4 0 C)
0 r) >
Qa -4 4-)
. 0 A -
A In -E
0 U0 )
f (4l -n Cn
F 0 0
C-I <) u)
* U )l
)0 c 41
C) C> ) 0 m
O ) r-H0U) o )r.mm 0 H
u r 4
'4 r -ro =
n n o) c 0- Q 0
(Mm 0 00
El 0) c In -
H.1-I MH' c ,r (n 0
M O N O
r~ U ?
'N0 Q mo
4 4 LU .- U) +M
0 1 O0 -O
U) 0 0 .
U) U) 0 0
U c-I l) WD '. r-l U-r U) H 0,- .u
0r -lNNN 0 U) 0
4 0, -H
V- GD 0)
r0 C 0
U Io -U
. .0 C
j (0 0
en e il
S4J -4 0
U) ~ NP o 4 4 'u4 40
3o C) I L
or a .
S ) o co mN r~ r) CC co rc 0 3
OOOOO000 000 000
4.4 . .l m r
) o omHHuo 4 O
0 Qon o o L 0
0 N MM-4C) CNNNN N Ma
( m o m m co CD c,
47. (q 1) 4 0 C )c cn m r n C n a 0 .-
M . 0
r-) r- v) 00 0 o
a). c4 4o 30 C, rq -4 -4 o H4 D)
-I rrl -4 -
CV ero H o !
U3 4C) 40
(1967) gave its magnitude at 17"5 and its colors as B V = 0O04 and
U B = -0?72. Burbidge (1968) measured its redshift as z = 0.404.
Burbidge et al. (1971) reported the existence of a faint 19th-magnitude
object between the QSO amd the galaxy. This object proved to be
transitory. A radio search by Fanti et al. (1977) at 1.4 GHz failed to
detect any radio emission to the detections limit of 20 mfu.
The comparison stars for the PHL 1226 field were calibrated by
photographic transfer from SA 94, using a plate taken on the night of
January 11, 1976. These stars are identified in Fig. 25 and their
magnitudes are given in Table 18. Twelve observations were made of PHL
1226, including 2 pairs. The resulting magnitudes are given in
Table 19. Figure 26 shows the behavior of PHL 1226 with time. The
object brightened in 1974 and declined from 1975 to 1977, possibly
rising again in 1978., The mean magnitude is 17"12, with a range of
0.36 and an average rms scatter of the comparison stars of 0O08.
Statistics show a 98 percent confidence of variation.
On the night of November 7, 1977, a strange, possibly non-stellar
object of approximately 18T7 appeared near the QSO on the side opposite
the galaxy. This object appeared only on this exposure and possibly, on
the same night, at the edge of the field of PHL 1222, which overlaps
This sample contains 7 objects from the Sandage and Luyten (1967)
field at Ih +60 and one nearby object, PHL 938, with the same colors
(Table 1). These 8 objects have a spread in V magnitude from 17 04 to
18M15. Their B V colors range from -0"07 to +0m41 and their U B
colors from -0m51 to -0O88. There are redshifts from z = 0.27 to
z = 1.93.
Variability results are shown in Table 43. Three objects (PHL 938,
PHL 1027 and PHL 1222) have shown no real evidence of variation during
this time period. One object (PHL 3375) whose variability has a
confidence level (C.L.) greater than 90 percent is probably variable,
but variability is not usually considered established unless the
confidence level exceeds 95 percent. Three objects (PIL 3632, PHL 1186
and PHL 1226) are variable (C.L. >95%) and one (PHL 1194) is strongly
variable (C.L. >99%). No relation between variability and magnitude,
color or redshift is apparent in this sample.
FIELD AT 13h +36
The comparison stars for the RQQSOs in the Braccesi field at 13h
+360 were calibrated using stars in M 3 located at 13h 40m +28.60. A
photoelectric sequence of outer stars in M 3 was published by Sandage
(1970), giving V magnitudes and B V and U B colors. This allows
calculation of B magnitudes for these stars. Eleven observations were
made of the area of M 3. A sequence of 14 stars was read on each plate
and the magnitudes were iteratively smoothed to give the best fit for
this photographic system. These 14 stars are identified in Fig. 27.
Their designation and magnitudes are listed in Table 20.
BSO 1 was one of the original blue "interlopers" found by Sandage
(1965) and Sandage and Veron (1965). They reported a redshift of
1.241. Braccesi et al. (1968) measured its magnitude and colors,
finding V = 16"98, B V = 0'31, U B = -0'78, and infrared excess
(Iex) of -0T30. Absorption lines were also visible in the spectrum.
Further photometry of this field was done by Braccesi et al. (1970). It
is number 9 on that list. A more precise optical position is given by
Sm N D m m -1 -4 am -0 0) 0 o m(
o rO Co o -H C ( a O
Ln Ln a, u) Lo D a, t- r-co w, w, m o m O
H r r H H- H H H H --I H H r-H -I
H (N (I M a, r-I a, a o H (N IC| N, N a, 5N
H m r- m m H
H H r-~-4 r-( -- --
0 *0 C
44 > 4,
n -H 0
' >i 4
": a "
e. * *
.*S "* .
* * a
wU a) C
4- tl 0
S4-1 a 4
J- '0 0)
cn 0) a)
1 OO yU
r^ v l (
w o m 'Aco 00Nr m
co omr H
0' m amwa, m'o o o m' 03 C)m 00O
N N N m m m m omm Oi
m m- o m mT m cN q u C m m mD 0 N -
n H( 0 cm) o oD N o c>V N 0 "0
r: c) '0 44
r- o S m0
1 c> c Dn m c -i r, i i r- o C4 cr ca% o) N o r- ij 41 W 04
4- 0 r-4 cH r-4 NN0 0c0 o 0 -4 4-44-0,4D-
m C q -i r c 4 rN r, HCN r r m *H rQ r o n3
0 N H N NN "-.m 0 r al
.0 0 m
Braccesi et al. (1973) as 12h 46m 28.7s and +370 46' 50". It was not
detected in the radio by Katgert et al. (1973) with a detection limit of
10 mfu at 1.4 GHz.
Comparison stars for BSO 1 were calibrated by photographic transfer
from M 3, using a plate taken on the night of January 25, 1977. The
magnitudes of these comparison stars are given in Table 21 and they are
identified in Fig. 28. Resulting magnitudes of 17 observations,
including 2 pairs, are given in Table 22 and plotted in Fig. 29. BSO 1
has shown a slow, steady increase in magnitude over this 4-year
period. The mean magnitude is 17m80, with a range of 0m56 compared to
an rms scatter of the comparison stars of 013. (Since plate effects
become more important for the fainter images, the average rms scatter of
the comparison stars is expected to be greater in the fields of the
fainter objects.) Statistics show BSO 1 to be variable at the 99
percent confidence level (Table 43).
Found by Braccesi et al. (1968), B 46 has a V magnitude of 17m83,
B V of 0'36, U B of -0'87 and lex of -1m3. Its redshift is 0.271.
It is object AB 11 in the listing of Braccesi et al. (1970). A refined
optical position of 12h 46m 29.6s and +340 40' 49" is given by Braccesi
et al. (1973). It was not detected in the radio by Katgert et al.
(1973) with a detection limit of 10 mfu at 1.4 GHz, or by Colla et al.
(1970) with a limit of 20 mfu at 408 MHz.
Comparison stars for B 46 were calibrated by photographic transfer
from M 3 using a plate taken on the night of July 25, 1976. These stars
O 0 .0
0 4 '0
o 0 m
0 0 0U
D W 4 4J
.o o m
l 00 0M
O 0 .
m -r E-4
o- oo or:co o ino o r-o w m w
- 00000000000000 C 0 0 0 0
) 0m C
[-r- r- r-
- N- rN
H H H H
SN O r co coN
0 ONm 0 Nm 0
HHmH H H Hmm HHHH
N H 0 l 0 H n
co 0 0 00 C, rN m
LVor-m D m
SN o00 00
'r N NN
0 N N H 01 CN
0000 N N H
m C c mo m
>-l iCi 1 04 CM r4 C
Nm r- 0
l^ 1 l^) 1
0- CD' 0
Pr *rf *w
CN N CN m D o IT 'T IN 'N N N 0 m
M NNN NN 0 NNNM M M H M N 0 H N
S0 00000 N N N N CD 000 0
n N- N Hz Hr N- 0 (q 0 NI Nr 0n 0 N D r- NN
to 4 4-
' C C
* 0 O
O e tO
u e n
4J m -
>. > C
tri (pfl 0
i-l P (
U~ -P a