Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://hea-www.harvard.edu/QEDT/Papers/pg1407.ps Äàòà èçìåíåíèÿ: Sat Apr 8 01:39:36 1995 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 23:38:00 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: quasar

THE UNUSUAL QUASAR PG1407+265
Jonathan C. McDowell 1 , Claude Canizares 3 , Martin Elvis 1 , Andrew Lawrence 2 , Sera
Markoff 3;4 , Smita Mathur 1 , Belinda J. Wilkes 1
Astrophysical Journal, 1995 Sep 10, in press
ABSTRACT
PG1407+265, discovered in the Palomar­Green Survey (Schmidt and
Green 1983), was identified as a redshift one radio­quiet quasar on the basis
of a single weak line. Further observations over a wide wavelength range
confirm the identification, but reveal the object to have unusual emission line
properties. Broad Hff is the only strong emission line, with Hfi and Lyman ff
almost undetectably weak. The emission lines show a range in redshift of over
10000 km/s, systematically decreasing with ionization potential, or almost
equivalently, increasing to longer wavelengths. z = 0:94 \Sigma 0:02 is a reasonable
statement of our knowledge of the quasar's redshift. However, the object's
continuum properties are those of a normal radio­quiet quasar. We discuss a
number of possible models for the object, but its nature remains puzzling.
Subject headings: Galaxies: Quasars
1 Center for Astrophysics, 60 Garden Street, Cambridge, MA 02138
2 University of Edinburgh
3 Massachussets Institute of Technology, Cambridge MA
4 Dept. of Physics and Steward Observatory, University of Arizona

1. Introduction
In the course of our survey of quasar continuum energy distributions (Elvis et al
1994), we realized that the redshift of one of our sample objects, PG1407+265, was not
secure. The assignment of z = 0:944 by Schmidt and Green (1983) was based on a single
extremely weak line identified as Mg II –2798. Noisy IUE spectra obtained in 1983 and
1987 did not reveal any further lines. Confidence in the identification was increased with
the publication of a high quality optical spectrum by Schneider, Schmitt and Gunn (1991)
which showed weak FeII emission in addition to the Mg II line, but a 1989 IUE spectrum
of improved quality by P. O'Brien failed to show any evidence for hydrogen Lyman ff.
We have carried out a comprehensive multiwaveband study (Table 1) of this
object as part of our larger survey of quasar properties (Elvis et al 1994). Our optical
spectrophotometry showed no evidence for significant variability over a baseline of ten
years; this and the radio­quiet nature of the source make it unlikely that it is related to
BL Lac objects. The X­ray flux detected by the Einstein and ROSAT satellites is at the
level expected for a quasar, and the continuum shape is normal with an ultraviolet bump
of median strength and, as expected, an inflection at between 1 and 2 microns in the
putative rest frame. Nevertheless, the equivalent width of Lyman ff (¸ 8 š A) is extremely
low compared to other objects (20­80 š A, Francis et al 1991).
2. Observations
2.1. Optical Spectrophotometry
We obtained optical spectrophotometry of the quasar during 1988 at the
Multiple Mirror Telescope (MMT) using the Blue Channel and the Faint Object Grating
Spectrograph (FOGS). Higher resolution data were obtained through a one arcsecond slit
and the flux level was normalized using lower resolution data through a five arcsecond
slit. Mg II and CIII emission lines are detected, together with strong FeII emission. In
an attempt to observe the region around rest frame Hfi, I.W. Browne, P.N. Wilkinson
and D. Henstock kindly obtained Faint Object Spectrograph (FOS) data on the 2.5m
Isaac Newton Telescope (INT) extending to 10000A. Unfortunately the spectral region in
question sits on atmospheric absorption lines, but the presence of an Hfi line of normal
strength can be ruled out. The normalization of all the spectra is in good agreement with
CCD photometry obtained at the 0.6m telescope of the F.L. Whipple Observatory in 1986.
The combined optical spectrum is presented in Fig. 1.
2

2.2. Ultraviolet Observations
The IUE spectra were retrieved from the National Space Science Data Center and
analysed using the RDAF IDL package. The spectra were extracted and calibrated using
the standard GEX Gaussian Extraction algorithm (Urry and Reichert 1988). HST data,
taken as part of the Absorption Line Key Project (Bahcall et al 1993) were retrieved from
the Hubble Data Archive using Starview and analysed with the STSDAS package. We
adjusted the flux of the G160L observation, taken through the 0.1 arcsecond aperture, to
match the data in the other gratings, taken through the 0.25 x 2.0 arcsecond slit. We note
that the G270H spectrum was not analysed by Bahcall et al. since it covers wavelengths
longwards of Lyman ff. Weak OIV+SiIV and CIV lines are present in the HST spectrum
(Fig. 2), and (less obviously) in the IUE data. The Lyman­ff line is present, but cut up
with superimposed absorption lines. Our value of 8 š A for the equivalent width of this line
is approximate and based on an ad hoc linear continuum fit at 1300A and 1140A. (Table
2). The HST data agree well in flux (to within 15 percent) with the 1989 and 1986 IUE
data; the quasar's far ultraviolet flux seems to have increased by 30 percent since the 1983
IUE observation. This degree of variability is normal in quasars (Kinney et al 1991).
2.3. Infrared Observations
IR spectroscopy was obtained through a service observation carried out on UKIRT
(the United Kingdom Infrared Telescope). The spectrum shown in Fig. 3, which shows
a strong Hff line at an observed wavelength of 1.28 microns, was obtained on 1994 Jan
29 with the cooled grating spectrometer CGS4, using the 150 mm camera, and a 75
lines/mm grating in second order, resulting in a spectral resolution corresponding to 780
km/sec. One resolution element is approxmately equal to one detector pixel; however the
detector was stepped in increments of 1/3 of a pixel to improve sampling. The slit width
used was 3'', roughly equal to one original detector pixel. Standard nodding, calibration,
and atmospheric correction techniques were followed. Because of the large slit size, the
wavelength calibration could potentially be offset if the object was not properly centred;
any such error is certainly less than 400 km/sec, and probably much less.
JHKL photometry was obtained at the MMT in 1988, and shows the continuum
inflection characteristic of a quasar. Examination of the IRAS sky images shows that
PG1407+265 is 3 arcmin from the strong 60 micron source IRAS14070+2636, which has
been identified with the starburst galaxy Haro 40. Addscans of the IRAS survey data
show no evidence for a 60 micron detection of the quasar, but we have reported a formal,
marginal 25 micron detection (Elvis et al 1994) although this should probably be treated
with scepticism as it was not detected in any other IRAS band.
3

2.4. X­ray Observations
PG1407+265 was observed on 1992 Jan 19 with the ROSAT PSPC (Tr¨umper
1983, Pfeffermann et al 1987) for 3229 sec. The source is well detected, and is consistent
with the point spread function. We note that there is no evidence for any significant
confusing source at the position of Haro 40 (Green et al 1989), with a 3 sigma upper limit
of 1 \Theta 10 \Gamma13 erg cm \Gamma2 s \Gamma1 (0.4­2.4 keV).
We extracted counts from a 3 arcminute circle around the position of PG1407+265,
together with background in an annulus from 3.5 to 7 arcminutes, using the IRAF/PROS
package. The data were analysed with the XSPEC spectral fitting package using the
Jan 1993 (DRM 36) response matrix. We fit a single power law spectrum with Galactic
absorption (Morrison and McCammon 1983). The data are consistent with this model
and the fitted absorption was consistent with the Galactic value of 1.38\Sigma0:10 \Theta 10 20 cm \Gamma2
(Elvis, Lockman and Wilkes 1989), so we repeated the fit with absorption fixed at this
value. The best fit gave a power law energy index of 1:61 \Sigma 0:05 and a 1 keV flux of
1:0 \Sigma 0:04¯Jy, with a reduced ü 2 of 0.49 (Fig. 4). Consistent results were obtained using
the SPECTRAL package in IRAF/PROS. We also set an upper limit to the amount of
extra absorption (above Galactic) due to cold material at the redshift of the source of
NH ! 3:1 \Theta 10 20 cm \Gamma2 (or 4.9\Theta10 19 cm \Gamma2 for extra absorbers at zero redshift); although
there are no obvious edges in the spectrum, we cannot rule out the presence of an
ionized absorber. Our derived power law index is steeper than, but consistent with, the
Einstein IPC result of 1.2 +0:9
\Gamma0:2 (Wilkes and Elvis 1987); note that fits to PSPC data appear
systematically steeper than the IPC (Fiore et al 1994). The normalization, however,
is more than twice as high as the IPC value of 0.44 ¯Jy, indicating significant X­ray
variability over a timescale of a decade.
3. Results
3.1. The Redshift of PG1407+265
The initial published redshift estimate (Schmidt and Green 1983) was based on
the single MgII line. The failure to detect other lines in our intermediate signal­to­noise
spectrophotometry and in the IUE data raised the possiblity that alternate (even
non­quasar) identifications might be possible (Elvis et al 1994).
Our observation of the strong Hff line in the near infrared and our identification
of the weak CIV and OIV]+SiIV lines in the HST data provide conclusive evidence that
the redshift of the object is indeed near unity. The high quality spectrum of Steidel and
Sargent (1991) claimed to confirm the redshift as z=0.947, and identified emission redward
4

of MgII as due to the –2950 FeII feature. However the feature they claimed as CIII]
(observed wavelength 4019.4A, their table 1) is inconsistent with this redshift; the feature
corresponds to a weak unidentified line in the composite spectrum of Francis et al (1991),
and is not as strong in our MMT data. The true CIII]+Al III feature is the weaker one
visible in their spectrum at approximately the correct rest wavelength.
The wavelengths of the peaks of each emission line were measured and used to
derive redshift estimates (Table 2). We quote nominal `1 sigma' uncertainties which are
one third of our eyeball estimates of the maximum uncertainty in the peak location. As
is typical for quasars, the redshifts of the low ionization lines are larger than those of the
high ionization lines (Gaskell 1982, Wilkes 1984, Espey et al 1989). However the range
in redshift for PG1407+265 is an order of magnitude larger than typically seen; 10200 \Sigma
1200 km s \Gamma1 between Hff and CIV compared with ¸1000 km s \Gamma1 in the earlier studies.
The wavelength scales of the various instruments should be much better than this; the
optical spectra were taken with the slit vertical to avoid refraction effects, while this is
not an issue for the ultraviolet spectra, and refraction is small enough in the infrared to
compensate for the larger slit width. Uncertanties in the velocity scale due to slit width
and refraction effects should be less than 500 km/s in all cases.
Figure 5 shows the various emission lines on a common velocity scale with zero
point set by Mg II. We see a significant trend of the velocity shift to increase with
the ionization potential of the emission line (Fig 6a). We note that a trend with rest
wavelength is almost equivalent (Fig 6b). We suggest that z=0.94\Sigma0.02 is a reasonable
statement of our knowledge of the quasar's redshift. Note in Figure 5 that there is a
narrow absorption feature in Lyman ff coincident with the velocity of the Hff emission
line, implying some overlaid neutral hydrogen component.
We also measured approximate line widths and fluxes. CIV –1549 and the broad
component of Hff both have full width half max (FWHM) around 7000 km/s. MgII has a
similar component but with a broad red­wing `shelf' of blended FeII UV multiplets leading
to a formal FWHM of around 12000 km/s. CIII] –1909 appears to be equal in flux with
Al III –1858, although our spectrum has low signal­to­noise in this region.
3.2. Line properties
The emission lines in PG1407+265 appear unusually weak by eye. How extreme
are these equivalent widths? Table 2 gives the details of our measured values, including
estimates of the integrated rest frame luminosity in each line. Typical emission line
properties of quasars are given in Table 12.4 of Osterbrock (1989), and for a subset of lines
in Francis et al (1991). While the Osterbrock data is not based on any specific sample,
the latter paper gives average properties of the well­defined Large Bright Quasar Survey.
Further, Francis (1993) gives histograms of equivalent width for the bright ultraviolet
5

lines which may be used to measure how unusual PG1407+265 actually is, although their
spectral signal to noise is too low to give reliable measurements for individual objects at
the low equivalent widths we are probing.
The OIV] equivalent width, W(OIV]), is relatively normal in PG1407+265, only
a factor of two low. However W(Lff) is in the bottom three percent of the Francis et al
distribution (typical values 30­100 š A), as is the CIV strength (typically 20­50 š A), while
W(MgII) is in the lowest 4 percent of its corresponding distribution (typical values
50­80 š A). Quasar Hff equivalent widths are still hard to find, with results from several
groups (Baker et al 1994, Baker 1994) indicating that typical values are in the range
250­1200 š A, consistent with the early data of Baldwin (1975). However we note that three
objects in the sample of Espey et al (1989) have values between 140 and 150 š A, comparable
to our value of W(Hff)=126 š A. We may summarize by saying that most of the line ratios in
PG1407+265 are not exceptional within the large uncertainties, but the equivalent widths
for the principal lines are anomalously weak by factors of 3 to 10. In contrast, the FeII
multiplets (both optical and UV) appear to be unusually strong.
3.3. Continuum properties
In contrast to the lines, the continuum properties of PG1407+265 are remarkably
normal. We present the dereddened, rest frame infrared to X­ray energy distribution in
Fig. 7, and a close­up of the optical to ultraviolet region in Fig. 8. The radio­loudness
(log of 5 GHz to optical B band flux) is 0.28, within the normal range (Kellerman et
al 1989, Elvis et al 1994) for radio­quiet objects. The 8 mJy, 5 GHz radio source is
unresolved, although there is a probably unrelated weak source one arcminute away
(Kellerman et al 1994). The blue bump strength is log (L(1000­2000 š A)/L(0.8­1.6¯m)) =
0.69, indistinguishable from the median value of 0.66 for the UVSX sample of Elvis et al
(1994), of which it is a member (see Fig.6, which shows the 68 percentile ranges of that
sample). The absolute continuum level in our 1988 MMT observations agrees with the
Neugebauer et al (1987) Palomar observations made in 1980 to within 15 percent, as seen
in a number of other objects in the Elvis et al survey. IUE and HST observations are
consistent with mild variations typical of quasars (Kinney et al 1991). The presence of an
inflection in the energy distribution between H and K, the Einstein IPC X­ray spectral
index of 1.2 and ROSAT index of 1.6, and the low radio flux (Kellerman et al 1989), are
all consistent with a normal, non­variable, radio­quiet quasar. The factor 2 variability in
the X­ray flux is somewhat unusual but not unprecendented for high luminosity quasars
(Zamorani et al 1984). There is therefore no evidence for any underlying variable, beamed,
blazar­like continuum component which may reduce the strength of the emission lines in
some flat­spectrum radio quasars.
6

4. Discussion
The emission line spectrum of PG1407+265 is unlike that of any other radio­quiet
quasar we have seen, e.g. out of ¸ 100 in Elvis et al (1994), Kinney et al (1991),
Boroson and Green (1993). In Fig. 8, we contrast the 1000­8000 š A rest frame spectrum of
PG1407+265 with comparable data for the ordinary, low redshift quasar PG1211+143.
The difference in the amount of energy in the ultraviolet lines is dramatic. We now consider
several possible explanations for the weak lines and differential redshifts of PG1407+265.
1. Not a quasar: Prior to the detection of Hff, we had speculated that the object
might not be a quasar at all. That line and the overall multiwavelength continuum
properties rule out that possibility.
2. A beamed continuum: The lines are washed out by a beamed, non­thermal component
(a jet along the line of sight). This solution, which would be unprecedented in a
radio­quiet object, is ruled out by the lack of continuum variability, and the normal
shape of the energy distribution.
3. Low BELR covering factor: The broad emission line region has a lower covering
factor, intercepting less of the continuum radiation. This simple explanation
accounts for the low equivalent widths of the emission lines which originally drew our
attention to the object, but does not address the bizarre CIV and OIV line shifts.
4. Abnormal photoionizing continuum: The weakness of the lines is intrinsic and due to
an abnormal photoionizing continuum as seen by the line emitting region. Although
the continuum we see is normal, we may be looking at a special angle. If there is
obscuration between the central source and most of the BELR (except along our
line of sight) this might occur, although we have not tried to reproduce the observed
spectrum with a photoionization code. Since objects like PG1407+265 are clearly
rare (a few percent at most), explanations involving special geometry are reasonable.
5. Time lags: The UV continuum has recently brightened and the lines have not yet
caught up. The lack of optical and UV variability on a ten year timescale casts
doubt on this, although the factor two increase in the X­ray since 1981 is worth
noting. This explanation would imply an effective BELR radius large compared with
ten light years, while scaling with the expected L 1=2 dependence from NGC5548
(Clavel et al 1991) would predict a much smaller BELR radius, ¸ 1 light year.
6. Abnormal BELR physical properties: The weakness of the lines is intrinsic and due to
abnormal cloud properties in the broad emission line region. Before the detection of
Hff the possibility of `hydrogen depleted' clouds was tenable (if physically unlikely).
The presence of Hff requires some other peculiarity: the size or density of the clouds
7

may be unusual, or the distance of the BELR gas from the central object may be
abnormal, leading to an unusual ionization parameter.
7. FeII dominated continuum: the emission lines have a normal equivalent width
with respect to the true continuum but are masked by strong FeII emission as in
IRAS07598+6508 (Lawrence et al 1988). This might work for some of the optical
lines but cannot contribute to the weakness of CIV and Lyman ff.
8. Dust absorption: The weakness of the short wavelength lines is due to patchy
reddening between us and the quasar nucleus. This is unlikely since the continuum
source shows no evidence of reddening: the ultraviolet continuum is strong and the
soft X­rays show no significant absorption in excess of Galactic. The absence of a
definite IRAS detection implies (Fig. 6) that the object is not exceptionally far
infrared bright, unlike some of the Fe II emitters studied by Lawrence et al (1988),
and so most of the continuum cannot be reprocessed by dust. Any explanation
based on reddening must therefore explain how the lines but not the continuum are
affected.
9. Line absorption: PG1407+265 is a weak BAL quasar in which the absorption lines
are eating away at the emission. Note that the Lyman ff region (Fig 3) appears
to contain narrow absorption lines, although there is no evidence for any Lyman
continuum break so NH Ÿ 10 17 cm \Gamma2 . Patchy absorption leaves the continuum
source unaffected while most of the BELR emission is soaked up in a BAL region
lying between us and the high ionization region of the BELR generating CIV and
OIV]. This explanation can address the observed redshift differences. The velocity
of the BAL clouds is such that the red wing and center of the BELR lines are
preferentially absorbed. The remaining blue wing gives an artificially low redshift.
The low ionization emission lines, which are generated further from the nucleus,
suffer less or no red BAL absorption and give the true emission redshift, which is
nearer to z=0.95. The problem with this picture, as with dust absorption, is the low
probability that the bulk of the emission lines would be absorbed without affecting
the continuum source.
10. Gravitational amplification: The equivalent widths are low because the continuum
source is magnified by a stellar mass microlens along the line of sight. Dalcanton
et al (1994) set probabilities on the cosmological density of lensing objects based
on the absence of weak equivalent width objects in some quasar samples; to cast
their argument slightly differently, if microlensing has any significant effect on the
quasar population, we should expect it to show up in a few high luminosity, bright
objects with weak equivalent widths. PG1407+265 is a factor 25 more luminous
than the median PG quasar from Schmidt and Green (1983), and only a magnitude
less luminous that the multiple­image lens PG1115+080. This interpretation would
predict that the continuum is likely to fade on a timescale of decades (for a 1
8

solar mass lens, Dalcanton et al 1994). However, the apparent strength of the FeII
emission, which would be unlensed, counts against this explanation. A quantitative
measurement of the strength of FeII(opt), not really possible with our current
data due to the low signal to noise near Hfi, would be important in evaluating the
possibility of lensing. Deep CCD imaging could reveal the presence of any lensing
galaxy at z ¸ 0:5.
5. Conclusion
We have presented observations of an unusual quasar, PG1407+265. The quasar
has a normal non­variable, radio­quiet continuum energy distribution but the high
ionization emission lines are extremely weak and strongly blueshifted with respect to
the low ionization lines. We have discussed nine possible explanations for the unusual
spectrum, none of which is immediately convincing. Higher signal­to­noise observations of
the lines and detailed modelling will be required to resolve the mystery of PG1407+265.
We acknowledge support from STScI grant AR­3686.01­91A, NASA grants
NAGW5­2201 (LTSA), NAS5­30934 (RSDC), and use of IPAC's National Extragalactic
Database (NED), the Goddard Space Flight Center's IUE data analysis facility and
HEASARC X­ray archive, and of the Hubble Data Archive.
9

REFERENCES
Bahcall, J., et al, 1993, ApJS 87, 1.
Baldwin, J., 1975, ApJ 201, 26.
Baker, A.C., et al, 1994, MNRAS 270, 575.
Baker, J.C., 1994, Ph.D. Thesis, University of Sydney.
Boroson, T., and Green, R.F., 1993. ApJS 80, 109.
Clavel, J., et al, 1991, ApJ 366, 64.
Dalcanton, J.J., Canizares, C.R., Granados, A., Steidel, C.C., and Stocke, J.T., 1994, ApJ
424, 550.
Elvis, M., et al, 1994 ApJ 95, 1.
Elvis, M, Lockman, F.J., and Wilkes, B.J., 1989. AJ 97, 777.
Espey, B., et al, 1989, ApJ 342, 666.
Fiore, F., et al, 1994 ApJ 431, 575.
Francis, P. J., et al, 1991. ApJ 373, 465.
Francis, P. J, 1993. ApJ 405, 119.
Gaskell, C.M., 1982, ApJ 263, 79.
Green, P.J., Ward, M.J., Anderson, S.F., Margon B., De Grijp M., and
Miley, G.K., 1989. ApJ 339, 93.
Kellerman, K.I., et al., 1989. AJ 98, 1195.
Kellerman, K.I., et al., 1994. AJ 108, 1163.
Kinney, A., et al., 1991. ApJS 75, 645.
Lawrence, A., et al, 1988, MNRAS 235, 261.
Morrison, R. and McCammon, D.A., 1983 ApJ 270, 119.
Osterbrock, D., 1989. Astrophysics of Gaseous Nebulae and Active Galactic Nuclei (Univ.
Sci. Books: Mill Valley, CA).
Pfefferman, E., et al, Proc SPIE, 733, 519.
Schmidt M., and Green, R.F, 1983. ApJ 269, 352.
Steidel, C.C., and Sargent, W.L.W., 1991. ApJ 382, 433.
Trumper 1983, Adv. Space Res. 2, No. 4, 241.
Wilkes, B.J., 1984, MNRAS 207, 73.
Wilkes, B.J., and Elvis, M., 1987. ApJ 323, 343
10

Table 1: Log of Observations
Telescope
1
Instrument
1
Date
Exp
(min)
Spectral
range
Observers
ID
2
Einstein
IPC
1981
Jan
17
26
0.2­3.5
keV
CfA
I05381
IRAS
Survey
1983
12­100¯m
[Survey]
VLA
1983
Nov
5
GHz
Kellerman
et
al
FLWO
0.6m
CCD
1986
May
14
BVRI
Wilkes
MMT
IR
Photom.
1988
Apr
7
JHKL
Willner
MMT
FOGS
1988
Apr
9
4500­8300š A
Wilkes
MMT
Blue
Channel
1988
Jun
6
3200­6200š A
Wilkes
IUE
SWP
1983
Apr
356
1250­2000š A
Snijders
SWP19858
IUE
LWP
1986
Dec
400
2000­3100š A
Bechtold
LWP09733
IUE
LWP
1989
May
28
412
2000­3100š A
O'Brien
LWP15616
IUE
SWP
1989
May
28
395
1250­2000š A
O'Brien
SWP36351
ROSAT
PSPC
1992
Jan
19
54
0.15­2
keV
Canizares
RP700359
HST
FOS/H19
1992
Mar
9
18
1590­2300š A
Bahcall
et
al
Y0RV0C03T,04T
HST
FOS/H27
1992
Mar
9
18
2200­3250š A
Bahcall
et
al
Y0RV0C05T
HST
FOS/L15
1992
Mar
9
11
1170­2500š A
Bahcall
et
al
Y0RV0C06T
UKIRT
CGS4
1994
Jan
29
8
1.15­1.36¯m
UKIRT
Service
INT
FOS
1994
Feb
12
10
3650­10000š A
Browne,
Wilkinson,
Henstock
1.
See
main
text
for
explanation
of
abbreviations.2.
Sequence
number
or
other
designation
in
archive.
11

Table 2: PG1407+265: Measured Rest Frame Emission Line Properties
Line Instrument – obs Redshift EW Luminosity Typical EW 1
( š A) (10 44 erg s \Gamma1 ) ( š A)
Lyff –1215 IUE ­ ­ ! 10 30--100
HST 2370 0.95\Sigma0.015 8\Sigma3 10:
OIV]+SiIV –1400 IUE 2690 0.92\Sigma0.01 9.1 \Sigma 3 4.2
HST 2675 0.912\Sigma0.004 4.3 \Sigma2 2.2
CIV –1549 IUE 2960 0.91\Sigma0.01 3.6\Sigma2.5 1.5 20--50
HST 2980 0.924\Sigma0.004 4.6\Sigma2 1.9
CIII]+Al III MMT/Blue 3700 0.94\Sigma0.03 9.9 3.5
––1909,1858
MgII –2798 MMT/Blue 5435 0.943\Sigma0.003 23.5 4.6 50­80
MMT/FOGS 5450 0.948\Sigma0.0012 23.2 4.8
INT 5445 0.946\Sigma0.002 11.6 2.2
SS91 2 0.947 24.4
Hfi –4861 INT ­ !40 !2.2
Hff –6563 UKIRT 12850 0.958\Sigma0.001 126.0 4.1 250­1200
1) Typical equivalent widths for the samples of Francis (1993), Espey et al (1989), Baker
et al (1994), Baker (1994).
2) Steidel and Sargent (1991).
12

Urry, C.M., and Reichert, G., 1988, NASA IUE Newsletter 34, 95.
Zamorani, G et al, 1984. ApJ 278, 28.
This preprint was prepared with the AAS L A T E X macros v3.0.
13

Fig. 1.--- Optical spectrum of PG1407+265. The plot shows the combined observa­
tions from the MMT Blue channel and FOGS instruments and the INT. Expected
locations of emission lines and selected Fe II multiplets are indicated. The Blue
channel and INT data have been multiplied by a constant 12 percent factor to agree
with the FOGS data.
14

Fig. 2.--- HST spectrum of PG1407+265. The plot shows the combined observations
from the G270H, G190H and G160L gratings.
15

Fig. 3.--- Infrared spectrum of PG1407+265, obtained as a UKIRT service observation
on 1994 Jan 29.
16

Fig. 4.--- ROSAT PSPC pulse height spectrum of PG1407+265. The upper panel
shows the extracted counts (crosses) and the fitted power law model (solid line) while
the lower panel shows the ratio of the counts to the model in each channel. The power
law model includes Galactic absorption of 1.38\Theta10 20 cm \Gamma2 .
17

Fig. 5.--- PG1407+265 emission lines plotted on a common velocity scale. The zero
point corresponds to the Mg II redshift. There are several Mg II spectra superposed
(MMT Blue, FOGS, and the INT data), and two HST exposures superposed for
Lyman ff. Squares in Mg II and Hff boxes correspond to broad band photometry
measurements.
18

2000 4000 6000
­15000
­10000
­5000
0
5000
0 20 40 60 80
­15000
­10000
­5000
0
5000
Fig. 6.--- PG1407+265 estimated line velocities versus (a) ionization potential, (b)
rest wavelength. The velocity zero point corresponds to the Mg II redshift. The
dashed lines in (a) connect the components of line blends, since those points are not
independent.
19

Fig. 7.--- Rest frame spectral energy distribution for the quasar PG1407+265. Observations
have been corrected for galactic foreground reddening with E(B­V)=0.03. The X­ray data
points show power law fits to the Einstein and ROSAT data with 90 percent slope errors
indicated. Solid and dashed lines indicate median quasar spectral energy distribution and
68 percentile deviations from Elvis et al (1994).
20

Fig. 8.--- Optical and ultraviolet rest frame energy distribution for PG1407+265, compared
with that of PG1211+143 (data from Elvis et al, 1994). The plots cover the same range
in logarithmic flux, so lines of identical equivalent width will appear the same. Line
identifications are indicated.
21