Non-Periodic Phenomena in Variable Stars
IAU Colloquium, Budapest, 1968
IRREGULAR VARIATIONS OF RADIO SOURCES
M. V. PENSTON and R. D. CANNON
Royal Greenwich Observatory, Hailsham, Sussex, England
At the Royal Greenwich Observatory, we have been carrying out a program
to study the optical variations of quasars and other radio sources. For
the past three years we have been using our 26" f10.5 refractor for
photographic photometry of these objects. This seems an opportune time
to give a general account of the situation to variable star observers
since if one is set up to observe variable stars, one is then just as
well set up to study quasars and it may be of interest to state what the
current position is.
We have already (Penston and Cannon, 1968) given an account of our
observations of quasars before 1967 November 5 = JD 2439799.5. Then we
included an extensive review of the literature on optical variations of
quasars. Since then other papers on this topic have appeared notably
from Kinman et al. (1968) on the detailed behaviour of 3C345 and from
Barbieri and Erculiani (1968) on the observations of quasar variability
made at Asiago.
At Herstmonceux we have been observing all quasars with V or m_pg less
than 17m in an attempt to survey such objects for variability, and have
been trying to obtain one plate per month of each object. We feel this
is a better use of our resources, given our relatively poor weather at
Herstmonceux, than trying to obtain detailed light curves in a few cases as
Kinman and his collaborators have been doing at Lick. Presently we have about
20 objects with enough data to say something about the presence or
absence of variability. We find that all quasars vary, most by about
0.1m to 0.2m in a year. Our results to date (1968 Sept. 29 JD 2440128.5)
are illustrated for PKS 0403-13 and 4C49.22 in Figures 1 and 2. These
slow rises and falls could be similar to the 13 year quasi-periodic
variations (Smith and Hoffleit, 1963) of 3C273 but it is too early to
draw firm conclusions as yet. Further light-curves have been given
elsewhere (Penston and Cannon, 1968). On the other hand a small
- proportion which we estimate at 10 per cent - of the quasars, which we
call optically violently variable (OVV) quasars, behave in a different
manner, varying by more than a magnitude and sometimes changing from
night to night by appreciable amounts. As well as being distinguished by
their optical behaviour they are characterized by flat radio spectra and
radio variability at high frequencies. Table I gives a list of bright OVV
quasars with approximate magnitudes, amplitudes of variation and comments.
Fig. 3 illustrates the variation of a typical OVV quasar 3C454.3.
Table I
Properties of the brighter OVV quasars
Name 3C 273 3C 345 3C 454 * 3 3C 446 3C 279
B Magnitude 12-13 15-17 15-17 15-18 16-18
Amplitude 0.7m 2m 2m 3m 2m
Comments 13-year 80 day Angione(3) Variation see Kinsman's
quasi-period(1). period(2) reports 2m characterized light curve
May not be OVV. flare in 1953. by "anti- given by
flares"(4' 5) Schorn et al.(6)
(1)Smith and Hoffleit (1963); (2)Kinman et al. (1968); (3)Angione (1967);
(4)Penston and Cannon (1968); (5)Cannon and Pension (1967); (6)Schorn et al. (1968).
Fig. 1. Light curve for the quasar PKS 0403-13. There is no magnitude
sequence for this object and the brightness is presented in terms of
mean iris readings. A difference of one in mean iris reading corresponds
to between 0.1m and 0.2m.
Fig. 2. As Fig. 1 but for the quasar 4C49.22
Fig. 3. Light curve for the OVV quasar 3C454.3.
We turn now to the variation of other radio sources first found
independently by Oke (1967) and Sandage (1967a) for the N-galaxy 3C371.
At Herstmonceux we have been taking plates of this and other radio
galaxies in search of optical variation. In order that our photometry
should not be influenced by the non-stellar nature of these objects, we
have been taking our plates deliberately out of focus to spread out the
star-images to be the same size as the galaxies. In fact it is the
nuclei of the galaxies that vary and in the case of the N-galaxies these
appear stellar on our plates and the correction is not important.
Subsequent investigation of old plates at Herstmonceux and Harvard has
produced a light curve from 1890 to date. The Harvard results have
already been published by Usher and Manley (1968) and in Fig. 4 we show
their results plotted with our results from Herstmonceux, complete to
date. Another case for which old plates have been examined to show
variability is the similar N-galaxy 3C390.3; we have already published
our results for both old and modern plates (Cannon et al. 1968) and
Sandage (1967b) has demonstrated the variability of 3C390.3 independently.
Fig. 4. Light curves for the N-galaxy 3C371 from old and modern plates.
Circles represent plates from Herstmonceux or Greenwich, squares plates
from Harvard(10), triangles photoelectric observations by Sandage(12) and
crosses other plates. Open circles or squares are eye-estimates, while
fills symbols show the plates were photometered.
The fact that 3C390.3 was brighter in the past provides a "continuity"
argument between quasars and N-galaxies indicating that quasars are at
cosmological distances. In fact 3C371 was also brighter in the past and
sor the same argument applies. In fact in the period 1890-1900 3C371 was
within half a magnitude of being as intrinsically bright (assuming the
Hubble law) as the quasar 3C48. If 3C371 had been observed in 1890-1900
it seems probable that it would have classified as a quasar.
The fact that these two objects were bright and variable prompted a search
by one of us (Penston, M. V., 1967) for similar objects that were known
variables and were listed as such in the General Catalogue of Variable Stars
(Kukarkin, 1958). It appeared that the interesting Seyfert galaxy 3C120 was
discovered to be variable by Hanley and Shapley (1940) in 1940 and had
been designated BW Tau. Since then Kinman (1968) has confirmed that this
object is still varying today - our few plates at Herstmonceux fit his
light-curve well. Another radio source was found to be identified with a
known variable star by J. Schmitt (1968). He showed that the radio
source VRO 42.22.01 was the variable star BL Lac (discovered by
Hoffmeister in 1930). On the Palomar charts this object appears as a
blue slightly fuzzy object and a plate taken at the prime focus of the
98" Isaac Newton Telescope shows that it has an ultra-violet excess,
typical of quasars and N-galaxies.
The fact that both BW Tau and BL Lac are radio galaxies and are listed
as irregular variables in the General Catalogue of Variable Stars
suggests that all 2000 irregular variables in that catalogue should be
examined to see if any of them are nebulous on the Palomar Charts. A
further radiogalaxy that varies appreciably in the optical region is the
Seyfert galaxy NGC 4151 as was shown by Fitch, Pacholczyk and Weymann
(1967). We have a few plates at Herstmonceux which also show
variability. The galaxies mentioned above show large variations and form
a group similar to the OVV quasars in many ways but other cases of
smaller variability have been found, in particular 3C109 and III Zw
1727+50 by Sandage (1967b). Table II gives a list of known OVV galaxies
together with their properties.
Table II
Properties of the OVV Galaxies
Name 3C 371 3C 390 * 3 3C 120 VR 42.22.01 NGC 4151
B Magnitude 13m-15m 14 1/2m-16m 14m 13m-16m 12m
Amplitude 2m 1.5m 0.7m 3m ~1m
Comments brighter in 1890-1900 BW Tau BL Lac
than present day
We wish to thank Drs. T. D. Kinman and C. Barbieri and Mr. R. Angione
for permission to quote their results prior to publication. We would
like to thank our colleagues who have assisted us with our observations.
Miss Rosemary Brett gave us valuable help in measuring our plates and
preparing the diagrams. We are particularly grateful to Mr. C. A.
Murray, Dr. D. Lynden-Bell and the Astronomer Royal for their support
and encouragement.
REFERENCES
Angione, R., 1967, private communication.
Barbieri, C. and Erculiani, Laura A., 1968, private communication.
Cannon, R. D. and Penston, M. V., 1967, Nature 214. 266.
Cannon, R. D., Penston, M. V. and Penston, Margaret J., 1968, Nature 217, 340.
Fitch, W. S., Pacholczyk, A. G. and Weymann, R. J., 1967, Astrophys. J. 150, 167.
Hanley, C. M. and Shapley, H., 1940, Harvard Bull. No. 913.
Hoffmeister, C., 1930, Mitt. Sternw. Sonneberg, No. 17.
Kinman, T. D., et al. 1968, Astr. J. (in press).
Kukarkin, B. V. and Parenago, P. P., 1958, General Catalogue of Variable Stars
(2nd edition), Moscow.
Oke, J. B., 1967, Astrophys. J. 150, 15.
Penston, M. V., 1967, Inf. Bull. Var. Stars No. 255. (IBVS N°.255)
Penston, M. V. and Cannon, R. D., 1968, R. Obs. Bull. (in press).
Sandage, A., 1967 a, Astrophys. J. 150, L9.
Sandage, A., 1967 b, Astrophys. J. 150, L177.
Schmitt, J., 1968, Nature 218, 663.
Schorn, R. A., Epstein, E. E., Oliver, J. P., Soter, S. L, and Wilson, W. J., 1968,
Astrophys. J. 161, 126.
Smith, H. J. and Hoffleit, D., 1963, Nature 198, 650.
DISCUSSION
Feast: Do you regard the sources of large and low amplitudes as distinct
species?
Penston: 1) There are several possible reasons why OVV Quasars should be
different. They may be
1. distinguished by such a property as total mass,
2. an evolutionary stage for AU Quasars,
3. an intermittent phase in the activity of any individual Quasar,
4. an effect of the direction from which we view the Quasars
(i. e. an aspect effect).
2) The OVV Quasars are indeed distinguished by their radio properties.
Rosino: 1. Which criteria do you use for the reality of small variations?
2. How much are affected the observations by the atmospheric extintion?
Penston: If I may answer your second point first. We have used a refractor
which acts as an extremely efficient filter to the ultra-violet. Thus
effects of atmospheric extinction do not worry us much. To your first
point we devised a statistical test to detect variability and also our
results are in good agreement with those of Kinman in several cases.