Non-Periodic Phenomena in Variable Stars
IAU Colloquium, Budapest, 1968
CHANGES IN THE LIGHT CURVE OF BETA LYRAE 1958-1959
G. LARSSON-LEANDER
Lund Observatory, Lund, Sweden SUMMARY
Photometric results obtained during the 35 days of the international
programme on beta Lyrae are compared with observations made in 1958, mainly
at the Lick Observatory. From the minimum epochs of the two seasons
a period of 12.9355 days is obtained. The total B magnitude is found
about 0.10 mag. fainter and the colour about 0.05 mag. redder in 1959
than in 1958. A slow decrease in brightness during the 1959 campaign is
indicated. The depths of the minima are about the same as in 1958, but
the bottoms of the primary minima show great dissimilarities among
themselves. The first minimum, on August 12, 1959, was almost flat, and
the nearly constant phase had a duration of about 0.70 days. The other
two primary minima, on August 25 and September 7, 1959, had bottom
widths of similar duration, but exhibited pronounced dips shortly after
mid-eclipse. The rising branches from the primary minima are steeper
than in 1958, making the minima almost symmetric. The ensuing maxima and
the secondary minima occurred at phases about 0.24 and 0.49,
respectively, or somewhat earlier than in 1958. The B-V curve of 1959,
showing the reddening at primary eclipse, is suspected to have a narrow
"blue top" at mid-eclipse. Another feature of the colour variation is
indication of a small "red dip" at about phase 0.37.
1. INTRODUCTION
Following a proposal by the late Dr. O. Struve, Commission 42 of the
I.A.U. decided at the 1958 Moscow meeting to make beta Lyrae the subject of
an international programme of coordinated photometric and spectrographic
observations. Mainly two reasons were given for the proposal. Firstly, no
recent photoelectric light curve was available. The second and more important
reason was the existence of erratic changes, both in the total light and in
the intensity of certain spectral features. Some kind of correlation between
these secondary variations might be expected. It thus seemed desirable to make
a concentrated effort to observe the star more or less continuously during a
certain interval of time, to make possible a direct intercomparison of
successive cycles. The present writer accepted the task of coordinating
the programme.
From subsequent correspondence with Dr. Struve it was learnt that
extensive observations of beta Lyrae had been made at the Lick Observatory
during the interval 1958, June 21 to July 12. Thus, D. B. Wood and M. F.
Walker had obtained a large number of photoelectric measurements on the
U, B, V system on all nights of this interval, and spectrograms had been
obtained by S. N. Svolopoulos on almost all the nights. In addition, Struve
had taken a large number of coude spectrograms at Mount Wilson during
nights of the interval. All these observations were part of an international
programme that Struve and Walker had organized. Five observatories outside
the United States participated in the photometric part of the programme, but
only a rather limited number of observations were made by them. All the
photometric observations from this programme have been published by Wood
and Walker (1960) together with a discussion of the light curve and its
changes. The secondary light variations were compared by Struve (1959)
with changes in the intensities of the absorption lines originating in
the outer shell.
As beta Lyrae had thus already been subjected to an international
programme, the question arose if it would be worth-while to proceed with
the programme decided upon by the I.A.U. The interval of the 1958
programme included two primary minima and one secondary minimum.
According to the preliminary results, communicated by Dr. Struve, the
magnitude at the four maxima was the same, while the two primary minima
showed some striking differences. Most obvious was a difference in
width: the first minimum, centred on 1958, June 25, was found
considerably wider than the following one, on July 8. Struve further
stated that the differences were clearly correlated with differences in
the intensities of the spectral lines produced by the shell. In view of
these results Struve strongly endorsed the plan of organizing a more
extensive photoelectric study in 1959.
2. THE 1959 CAMPAIGN
In consultation with the president of Commission 42 the international
campaign was planned to cover three successive primary minima, viz.,
those of 1959, August 12.9, 25.9, and September 7.8. The limiting dates
were set at August 8 and September 11, 1959.
One of the objects of the 1959 campaign was to produce light curves
for beta Lyrae that could be directly compared with those obtained in 1958.
For this reason it would be most advantageous to adopt the comparison
stars used by Wood and Walker. Details on these stars, HR 6997, 8 Lyrae,
and 9 Lyrae, were kindly communicated by Mr Wood. According to this
information the colour of HR 6997 was very close to that of beta Lyrae, while
8 Lyrae was somewhat bluer and 9 Lyrae somewhat redder. HR 6997 had
therefore been used as primary comparison star in the 1958 programme, but
numerous checks with 8 Lyrae and 9 Lyrae had been made. The preliminary
results indicated that 9 Lyrae might be variable by about 0.02 mag., while
HR 6997 and 9 Lyrae appeared to be reasonably constant. On the basis of
this, it was decided to exclude 8 Lyrae from use in the 1959 campaign and the
observers were recommended to adopt HR 6997 as primary comparison star
and 9 Lyrae as check star. Further, observers using such equipment that
transformation to the U, B, V system might be feasible were asked to observe
a sufficient number of standard stars to make this transformation possible.
Not all observers who had intended to participate in the photometric
part of the campaign were successful in obtaining observations. As a
compensation, however, observations were made also by colleagues who had not
in advance announced their participation. The net photometric outcome of
the campaign is sixteen series of observations, made at fourteen observatories.
The total number of observations, from the interval of the campaign, amounts
to 2361. Some observers extended their series outside the campaign limits.
Details on the sixteen series of observations are given in Table 1, where
the participating observatories are listed in longitudinal order. The table is
self-explanatory, except for the penultimate column, giving information on
the colour systems. In cases where the instrument-filter combinations are
different from those of the standard U, B, V system, the colour regions
have been roughly indicated by the letters u, v, b, y, and r, which
stand for ultraviolet, violet, blue, yellow, and red, respectively.
Arrows following these symbols and pointing to U, B, V indicate that the
observers have also furnished data transformed to the standard system.
Particulars concerning filters and transformation formulae are given in
the papers referred to in the final column of Table 1. These same papers
also contain the original observations. Several of the observers have
desisted from publishing their observations, but kindly submitted them
to the present writer for inclusion in this final report of the campaign
(Larsson-Leander, reference 4 of Table 1).
Table 1.
Observers and instruments
Series Observatory Observers Instrument Multiplier Colour system Ref.
1 Nanking Chang, Hong, 24" reflector RCA 1P21 b (no filter) 1
Mo, Chow (silvered
mirrors)
2 Byurakan Grigoryan 16" reflector EMI 6094 u, b, y, r 2
3 Abastumani Magalashvili, 13" reflector v, y 3
Kumsishvili
4 Cracow Szafraniec 8" refractor RCA 931-A b, y 4
5 Budapest Balazs-Detre 24" reflector RCA 1P21 u, B, V, r 4
6 Stockholm Larrson- 24" refractor EMI 5060 B, V 4
Leander
7 Capodimonte Fresa 7" refractor RCA 1P21 b (no filter) 5
8 Copenhagen Gyldenkerne, 10" reflector EMI 5060 B, V 6
(Brorfelde) Jaeger
9 Hoher List Herczeg 14" reflector RCA 1P21 u, b, r->U, B, 7
V
10 Leiden, I Kwee 18" reflector EMI 6094 U, B, V, r 4
11 Leiden, II van Agt 10" refractor RCA 1P21 v, B, y 4
12 Pic-du-Midi Bouigue, 24" reflector Lallemand U, B, V 4
Pedoussant,
Rochette
13 Sidmouth Archer 7" astrograph RCA 931-A b, y-> B-V 4
14 Flower and Binnendijk 28" reflector RCA 1P21 u, v, b, y->U, 8
Cook, I B, V
15 Flower and Bookmyer 15" siderostat y->V 8
Cook, II
16 Lick Gordon 22" reflector RCA 1P21 B, V 9
1. Chang, Hong, Mo, and Chow (1959); 2. Grigoryan (1961);
3. Magalashvili and Kumsishvili (1960); 4. Larsson-Leander;
5. Fresa (1960);.6. Gyldenkerne and Jaeger (1963);
7. Herczeg (1964); 8. Binnendijk (1960); 9. Gordon (1960).
Photometric observations were made during all 35 days of the campaign.
However, because of the unequal longitudinal distribution of the
observers, gaps of about 0.5 days are frequent. Table 2 gives for each
colour the number of observations furnished by the various observers,
with the number of observing nights within parentheses.
Table 2.
Number of observations and observing nights (between parentheses)
Observatory u v b y r
Nanking 442 (25)
Byurakan 53 (21) 56 (21) 55 (21) 35 (12)
Abastumani 59 (6) 59 (6)
Cracow 28 (6) 28 (6)
Budapest 44 (22) 43 (22) 42 (22) 45 (22)
Stockholm 93 (16) 93 (16)
Capodimonte 193 (28)
Copenhagen 80 (15) 92 (15)
Hoher List 53 (11) 53 (11) 53 (11)
Leiden, I 14 (6) 14 (6) 13 (6) 13 (6)
Leiden, II 8 (6) 12 (6) 12 (6)
Pic-du-Midi 12 (7) 12 (7) 12 (7)
Sidmouth 7 (7) 7 (7)
Flower and Cook, I 27 (5) 74 (6) 80 (6) 81 (6)
Flower and Cook, II 34 (3)
Lick 115 (13) 115 (13)
All 203 141 1228 643 146
The internal mean errors of the various series of observations were
estimated by two methods:
(1) From the scatter shown in magnitude differences obtained for
comparison and check stars.
(2) From the scatter in the observations of beta Lyrae at epochs when the
variable should be almost stationary in light, i.e. at maxima and
minima, or from scatter during very short intervals of time.
The pre-requisite for the use of method (1) is, of course, that the
comparison and check stars remained constant during the interval of the
campaign. As further discussed in Section 3, this was found to be true.
In several cases mean errors derived in this way were communicated by
the observers themselves.
As regards method (2), many series of closely spaced observations
exhibit a rather large scatter, which one would be tempted to interpret
as rapid fluctuations. However, intercomparisons of various overlapping
series indicate quite clearly that the main part of the "fluctuations"
merely reflects observational errors or effects introduced by variable
atmospheric extinction. This view is supported by the fact that for
series containing a sufficient number of observations, mean errors
calculated by means of method (2) agree with those from method (1).
The average values of the internal mean errors are given in Table 3. Two
series, Pic-du-Midi and Sidmouth, are missing from this tabulation,
because of insufficient data.
We end this section by noting that high-dispersion spectrograms of
beta Lyrae were obtained by Abt at McDonald Observatory on thirteen nights of
the campaign. In addition, K. O. Wright and A. McKellar at Dominion
Astrophysical Observatory obtained spectrograms on three nights, one
just after the end of the campaign. The entire spectrographic material
has been discussed by Abt (1962).
Table 3.
Internal mean error of a single observation
Series u v b y r
Nanking .018
Byurakan .02 .02 .02 .02
Abastumani .030 .034
Cracow .05 .04
Budapest .020 .018 .020 .015
Stockholm .008 .010
Capodimonte .02
Copenhagen .010 .010
Hoher List .012 .010 .008
Leiden, I .017 .010 .008 .010
Leiden, II .026 .012 .010
Flower and Cook, I .013 .015 .018 .015
Flower and Cook, II .019
Lick .01 .01
3. THE COMPARISON STARS
As recommended by the coordinator most observers used HR 6997 and 9 Lyrae
for comparison purposes, but other stars were adopted by some observers.
Table 4 lists the stars actually used in the various series, the primary
comparison star being the first one. The observers checked the constancy of
their respective primary comparison star by repeated measurements of their
second (or third) star in Table 4. From studies of plots of the magnitude
differences versus time, it is concluded that all comparison stars remained
constant, within observational accuracy. During the 1958 programme Wood
and Walker (1960) obtained occasionally discordant results for the
magnitude difference between HR 6997 and 9 Lyrae, which were ascribed to
variations of HR 6997. No such discordances, beyond the observational
errors, were noted during the 1959 campaign.
Table 4.
Comparison stars
Series Stars
Nanking HR 6997, 9 Lyr
Byurakan HR 6997, 9 Lyr
Abastumani 9 Lyr, HR 6997
Cracow HR 6997
Budapest HR 6997, 9 Lyr, gamma Lyr
Stockholm HR 6997, 9 Lyr
Capodimonte HR 6997
Copenhagen HR 6997, 9 Lyr
Hoher List HR 6997, 9 Lyr
Leiden, I HR 6997, 9 Lyr
Leiden, II HR 6997, 9 Lyr
Pic-du-Midi gamma Lyr, phi Lyr A, phi Lyr B
Sidmouth HR 6997
Flower and Cook, I gamma Lyr, 9 Lyr
Flower and Cook, II gamma Lyr, 9 Lyr
Lick HR 6997, 9 Lyr
The complicated procedure of obtaining the most probable B and V values
for the comparison stars from the observations made by the various
observers is omitted here. Full details are given in the writer's
comprehensive report, which is being printed in Arkiv for Astronomi.
The finally adopted standard magnitudes and colours for the three main
comparison stars, HR 6997, 9 Lyr, and gamma Lyr, are given in Table 5.
Corresponding values from 1958, as obtained by Wood and Walker (1960)
are also listed. It is seen that the 1959 V magnitudes are slightly
fainter than the 1958 values. The B-V colour of HR 6997 turned out
somewhat redder in 1959 than in 1958, while the colour found for 9 Lyr
is the same during both seasons.
Table 5.
Standard magnitudes and colours for comparison stars
1959 1958
Star Sp. V B-V V B-V
HR 6997 B8 5.452 -0.126 5.430 -.0154
9 Lyr A2 5.279 +0.058 5.254 +0.059
gamma Lyr B9 III 3.250 -0.045
4. THE DERIVATION OF LIGHT CURVES
It is known that the B-V colour of beta Lyrae is only slightly affected
by the light variations. Wood and Walker (1960) found an increase of B-V
by about 0.07 mag. at primary eclipse, but no change at secondary
eclipse. It was therefore expected that the reduction of the 1959
observations to a common system would present no particular
difficulties, even though the various series had been made in different
colour regions. However, the problem turned out to be rather more
complicated than anticipated.
Attempts to derive transformation formulae by means of the measurements
of the comparison stars proved unsuccessful, probably because of the
small colour differences involved. Instead of using this straightforward
method, it was then necessary to adopt one or several series as
standards, and to reduce the other series by means of empirical
corrections. Obviously, the standard series had to be chosen among those
stated by the observers to be on the B, V system.
With the 1959 magnitudes of the comparison stars, as given in Table 5, it was
found that the partly overlapping Stockholm, Pic-du-Midi, and Flower and Cook I
series of both B and V magnitudes agree very satisfactorily. These series,
without any corrections, were taken to define frame-works of standard
magnitudes. They were plotted on large-scale graphs, and fragmentary
light curves were obtained. These contained points at a variety of phases,
including maxima, as well as primary and secondary minima. One by one the other
series were reduced to these frameworks, and more points were
successively added to the light curves. For each series the necessary
corrections were determined at all epochs where overlaps occurred. The
runs of the corrections were studied versus magnitude and colour in
order to disclose possible non-constant terms. The order in which the
various series were taken and the corrections obtained are shown in
Tables 6 and 7.
Table 6.
Reduction of V magnitudes
Series V (beta Lyr)
Stockholm 5.452 + Delta V
Pic-du-Midi V
Flower and Cook, I 3.250 + Delta V
Copenhagen 5.452 + Delta V-0.047
Leiden, I 5.452 + Delta V-0.071
Budapest 5.452 + Delta y-0.052
Leiden, II 5.452 + Delta y-0.025
Hoher List 5.452 + Delta V-0.053
Byurakan 5.28 + Delta y-0.07
Lick 5.28 + Delta y-0.11
Abastumani 5.452 + Delta y-0.047
Flower and Cook, II 3.250 + Delta y
Table 7.
Reduction of B magnitudes
Series B (beta Lyr)
Stockholm 5.326 + Delta B
Pic-du-Midi B
Flower and Cook, I 3.205 + Delta B
Copenhagen 5.326 + Delta B -0.010
Leiden, I 5.326 + Delta B -0.041
Leiden, II 5.326 + Delta b +0.004
Hoher List 5.326 + Delta b
Budapest 5.326 + Delta b -0.33(Delta v-Delta b)+
+0.010
Lick 5.34 + Delta b -0.07
Byurakan 5.34 + Delta b -0.05
Abastumani 5.326 + Delta v +0.030
Nanking 5.337 + Delta b -0.065 Delta b -0.037
Three series, Cracow, Capodimonte, and Sidmouth, which are missing from
these tabulations, had to be omitted. The Cracow and Sidmouth series contain
rather few observations and show, when compared with other series, erratic
deviations of a considerable amount. For the more numerous Capodimonte
observations, obtained with a refractor and no filter, it was not possible
to find any single correction formula that could be used consistently
for the whole series. Further, because of large deviations, some five
observations had to be rejected in each of the Byurakan and Abastumani
series.
As shown in Table 6, all the accepted series of yellow magnitudes could
be reduced to the framework of standard V magnitudes by means of
constant corrections. It is noteworthy that none of these corrections
appear with positive sign. The largest correction, that of the Lick
series, amounts to -0.11 mag.
On the other hand, for the blue observations it was necessary, as
indicated in Table 7, to introduce a colour term for the Budapest
observations and a magnitude term for the Nanking series. The
corrections applied to the individual observations of these two series
vary from -0.007 to -0.040 mag. and from +0.031 to +0.097 mag.,
respectively. For the Hoher List blue observations the instrumental
values Delta b were used rather than values of Delta B, as calculated from
formulae derived by Herczeg (1964). The reason is that no correction
was needed in the former case, while use of Delta B made a correction
necessary. It is noted that here again the Lick series requires a rather
large negative correction, amounting to 0.07 mag.
The resulting composite light curves, showing the fit of the various
series, are not reproduced here, because of their bulky nature. They
are, however, included in the more comprehensive report
(Larsson-Leander), already referred to.
5. NORMAL POINTS, MINIMUM EPOCHS, AND PHASES
The individual observations, reduced to the B, V system as described in
the previous section, were allotted weights in accordance with the mean
errors of Table 3. The Pic-du-Midi series, not appearing in this table,
was included in the group of series of highest accuracy, the observations
of which were given unit weight. Normal points were then formed from
the observations accepted. Because of the unequal spacing of the observations
and the various degree of accuracy, the weights of the normals vary between
0.2 and 8.4. The total number of normals is 173 in V and 221 in B.
From the B light curve, which is the most complete one, the epochs of
the three primary minima are found as
I. J. D. 2436 793.47
II. 806.405
III. 819.36
where the second value has the highest accuracy. Combining this with the
minimum epoch J. D. 2436 379.532, derived by Wood and Walker (1960), a
period of 12.9355 days is obtained. Phases for the normal points were
computed using this value for the period and the observed epoch for primary
minimum II.
The B normal points and the B-V colours, derived as differences between
the B and V normals, are plotted versus phase in Fig. 1. Different symbols
have been used for the various cycles. Some few normals have been omitted,
to avoid crowding of symbols. For comparison, the runs of B and B-V according
to the 1958 Lick observations (Wood and Walker, 1960) are indicated by line
segments. These have been drawn with respect to the phases designated II
by Wood and Walker.
Fig. 1. Light and colour curves for beta Lyrae according to normal points
determined during the 1959 campaign (symbols), compared with curves
(line segments) derived from the 1958 Lick observations (Wood and Walker, 1960).
Dots refer to the interval J. D. 2436 788 - 802.5, crosses to 801.3-815.4,
open circles to 814.3-823. Line segments drawn in full refer to
J. D. 2436 375-387, dashed segments to 387-396.
It may be remarked here that the period of 12.9355 days, derived from the
1958 and 1959 minima, is somewhat longer than expected from ephemerides in
current use. Wood and Walker (1960), following J. Sahade, S.-.S Huang, Struve,
and V. Zebergs (1959), calculated phases from a formula given by
R. Prager (1931) and modified by K. Saidov (1955), namely,
Min. = J. D. 2398590.57 + 12.908006 E + 0.3919 X 10^-5 E^2-0.3 X 10^-10 E^3
The period predicted by this formula, and corresponding to the one
quoted above, is 12.9303 days. The residuals, O-C are +0.266 day for the
1958 minimum epoch and +0.439 day for the 1959 minimum II. The failure
of the formula to represent the minima was noted by Wood and Walker, and
by adjusting the phases by the corresponding amount, -0.0206 P, they
obtained the system designated II (using the period 12.93016 days, predicted
for the epoch of observations).
More recently a new ephemeris,
Min. = J. D. 2433 289.47185 + 12.928481 E + 0.3556 X 10^-5 E^2 - 0.648 X 10^-10 E^3
has been derived by Wood and J. E. Forbes (1963) from a leastsquare
solution of 465 minimum epochs. The period predicted for the mean epoch
of the 1958 and 1959 observations is 12.93029 days. The residuals, O-C,
are -0.049 for the 1958 epoch, but +0.12, +0.125, and +0.15 days,
respectively, for the three 1959 minima. Although these latter residuals
are smaller than the dispersion, 0.17520 days, obtained in the solution,
they are obviously to be regarded as significant.
Of course, the difference between the periods used for calculating
phases for the 1958 and the 1959 observations, is much too small to
invalidate in any noticeable degree the comparison of light and colour
curves, presented in Fig. 1.
6. DISCUSSION OF LIGHT AND COLOUR CURVES
As shown by Fig. 1, the over-all magnitude of beta Lyrae in the blue region
was about 0.10 mag. fainter in 1959 than in 1958, and the B-V colour
appears to have been about 0.05 mag. redder. The difference in the
visual region is thus about 0.05 mag. It is recalled in this connection
that most of the 1959 series of observations, if taken uncorrected,
indicate still fainter magnitudes (cf. Tables 6 and 7). Disregarding for
the moment the disturbance in the form of an extra dip shortly after
zero phase, it is also obvious from Fig. 1 that the depth of the primary
minimum was about the same both in 1958 and 1959. The secondary minimum,
on the other hand, was possibly somewhat shallower in 1959.
The most striking difference between the 1958 and 1959 light curves is
the change in asymmetry. In 1959 the rise from primary minimum was much
steeper than in 1958. This is probably to some extent connected with the
phases of the ensuing maximum and of the secondary minimum. In 1:958 the
maximum, M 1, following primary minimum arrived at about phase 0.27,
while in 1959 M 1 came at a phase somewhat earlier than 0.25. A similar
phase shift is apparent for the secondary minimum, which in 1958 arrived
at phase 0.51 and in 1959 at phase 0.49.
Data on the phases of the two maxima, M 2 and M 1, and of the secondary
minimum are given in Table 8, according to independent measurements on
both the B and V light curves from 1958 and 1959. The cycles are counted
according to the ephemeris of Wood and Forbes (1963). The first two cycles
of Table 8 thus correspond to the interval observed by Wood and Walker,
while the three later ones are those of the 1959 campaign. Besides the
systematic differences mentioned above, we note a large difference in
the M 2 phase for the two 1958 cycles. This, of course, is due to the
abnormally faint magnitudes found during the first of the 1958 observing
nights (cf. Fig. 1).
Table 8.
Phases of maxima and secondary minimum
Cycle* M 2 M 1 Sec. min
E B V B V B V
239 -0.219 -0.230 0.276 0.272 0.508 0.512
240 -0.258 -0.258 0.274 0.265
271 -0.247 -0.253 0.247 0.246 0.492 0.490
272 -0.244 -0.254 0.236 0.228 0.491 0.489
273 -0.240 -0.251 0.235 0.232
* Cycles are counted according to the ephemeris of Wood and Forbes (1963).
The asymmetry of the primary minimum was first noted by J. Stebbins
(1916) from his observations made at the Lick Observatory in 1915. From
a mean light curve, covering three cycles, he found M 2 at phase 0.267
and M 1 at -0.228. As pointed out by Wood and Walker (1960) the
difference in slope of the declining and rising branches was at that
time found much larger than in 1958. It appears from Fig. 1 that in 1959
the slope of the two branches was almost the same, or even somewhat
steeper for the rising branch, if the comparison is extended to phases
more distant from the minimum epoch than 0.10 P.
At phases corresponding to the shoulders of the principal minimum, say
from phase 0.80 to 0.90 and from 0.10 to about 0.20, the various cycles
deviated systematically from each other both in 1958 and 1959. The deviations
in 1958, which affected the upper width of the primary minimum, were found
by Wood and Walker. As shown in Fig. 1 the 1959 deviations are exaggerated
because of a slight progressive change in the total magnitude of the system.
This change is further substantiated by Table 9, giving B and V
magnitudes for the maxima and minima observed in 1958 by Wood and Walker
(1960) and in 1959 during the international campaign. Smoothed light curves
have been used in all cases. Note that the magnitudes of the 1959 primary
minima refer to the faintest portion of the minima, the dips around phase 0.02
are thus included. From the magnitudes at the maxima observed in 1959, it
appears that during the interval of the campaign the brightness of the system
decreased rather regularly by about 0.05 mag. This is a second evidence for
the existence of slow magnitude variations, the first one being simply the
over-all difference between the 1958 and 1959 magnitudes. Of course, such
variations are contributing to the scatter obtained when observations from
several cycles are combined to a mean light curve.
Table 9.
Magnitudes at maxima and minima
Cycle* M 2 Prim. min. M 1 Sec. min.
E B V B V B V B V
239 3.29 3.36 4.21 4.19 3.30 3.35 3.77: 3.85:
240 3.29 3.35 4.22: 4.20: 3.31 3.35
271 3.37 3.36 4.31 4.26 3.39 3.39 3.80 3.81:
272 3.40 3.41: 4.40 4.34 3.40 3.41 3.82 3.85
273 3.41 3.42 4.37 4.31 3.41 3.41
* Cycles are counted according to the ephemeris of Wood and Forbes (1963).
The bottoms of the three primary minima observed during the 1959 campaign
exhibit striking dissimilarities. The first minimum (E = 271) seems to have
been almost flat, with perhaps a very small dip, in blue light at about
phase 0.02. The duration of the constant, or nearly constant phase was about
0.06 P. If considered flat, the minimum magnitudes are B = 4.30 and V = 4.26,
and the depths of the minimum are Delta B = 0.92 and Delta V = 0.88.
These depths are only slightly larger than the values, Delta B = 0.91 and
Delta V = 0.84, found by Wood and Walker (1960) for the two 1958 minima.
The other two of the 1959 primary minima show pronounced dips shortly after
zero phase. At the second minimum (E = 272) the dip amounts to Delta B = 0.10
and is centred at phase 0.015. At the third minimum (E = 273) the dip is
Delta B = 0.07, and the centre falls at phase 0.025. As shown by the V
magnitudes in Table 9, the two dips appear slightly smaller in the yellow
light curve. For phases 0.02 to 0.03 these pronounced differences between
the first cycle and the two following ones are clearly shown already in
the light curve based on the Lick observations alone (Gordon, 1960).
The bottoms of the secondary minima appear rounded and no peculiarities
are noted. The depths are about Delta B = 0.41 and Delta V = 0.43, which may
be compared with the Delta V = 0.47, given by Wood and Walker (1960).
Turning now to the B-V colours, we note that the 1958 and 1959 mean
curves are closely parallel to each other. The reddening during primary
minimum is well shown also in the 1959 curve. There is, however, in the
1959 data some indication that the system grew somewhat bluer very close
to mid-eclipse, say between phases -0.015 and 0.015. Such a slight top
in the colour curve is not incompatible with the 1958 B-V data, and a
wider top is certainly apparent in U-B (cf. Fig. 7 in Wood and Walker,
1960). The existence of this top in the colour curve seems to be
supported by the six-colour observations by M. J. S. Belton and H. J.
Woolf (1965). These observations were made at Lick Observatory on twelve
selected nights 1961, spaced over an interval of two months. Segments of
light and colour curves were obtained, which were joined to mean curves,
using the ephemeris of Wood and Forbes (1963). The resulting U-V curve,
as drawn by Belton and Woolf, has a peculiar shape at primary eclipse.
In contrast to V-I and other similar, long-wave colours, the minimum of
the U-V curve occurs much earlier than mid-eclipse. This is because the
U-V was found "redder" around phase 0.94 (Wood and Forbes' system) than
at the next phase observed, around 0.02, while no further measurements
were made until phase 0.17. It is implied that the measurements around
phase 0.02 refer to the colour top found from the 1958 and 1959
observations.
Finally, the B-V colour curves from both 1958 and 1959 indicate a slight
dip at about phase 0.37. It would perhaps have remained unnoticed, if it
had not been for the much more pronounced dip, at the same phase, that
appears in the V-I curve given by Belton and Woolf (1965).
7. CONCLUDING REMARKS
The aim of the present paper has been to report very briefly on the
photometric result of the 1959 international campaign, and to compare these
results with the photometric data obtained in 1958 (Wood and Walker, 1960).
The 1959 data have added to the complexity of the system. As a whole the
system appears to be slightly variable in light and colour. The rising branch
of the light curve from primary minimum was found much steeper in 1959
than in 1958, and the ensuing maximum and the secondary minimum occurred at
somewhat earlier phases. The bottoms of the three primary minima observed
in 1959 have different shapes; two of them show pronounced dips just after
mid-eclipse. New features of the B-V colour curves are indications that
the system exhibits a "blue top" at mid-eclipse and a "red dip" at phase 0.37,
i.e. at about the beginning of the secondary eclipse.
The interpretation of the photometric and spectrographic data of beta
Lyrae, and the changes reported by various observers, in terms of a coherent
model, presents formidable problems. Probably the tentative model proposed
by Huang (1963) offers the most promising possibilities. The invisible
secondary component is assumed imbedded in an opaque semi-stable disk in the
orbital plane, inclined to the line of sight. Eclipses, caused by the opaque
disk and the B8 star, may be expected to show a variety of changes, depending
upon slightly variable size and opacity of the disk.
In any case, beta Lyrae is a system in rapid evolution, showing ample
evidence of mass loss and mass transfer. Already the star most often studied,
except for the Sun, it still represents a challenge. Future cooperative
photometric and, if possible, spectrographic programmes would certainly be
of the highest value. Much greater emphasis than in 1959 should, however,
be placed on the standardization of the photometry.
The writer is indebted to all those who participated in the 1959
programme and for their kindness to make their observations available.
He apologizes for his long delay in furnishing the final results.
REFERENCES
Abt, H. A., 1962, Astrophys. J., 135, 424.
Belton, N. J. S. and Woolf, N. J., 1965, Astrophys. J., 141, 145.
Binnendijk, L., 1960, Astr. J., 65, 84.
Chang Chia-hsiang, Hong Hen-jung, Mo Ching-er and Chow Hsin-hai, 1959, Acta astr. Sin., 7, 198.
Fresa, A., 1960, Mem. Soc. astr. ital., 31, 365.
Gordon, Katherine C., 1960, Publ. astr. Soc. Pacific, 71, 1960.
Grigoryan, K. A., 1961, Soobsch. Bjurak. Obs., 29, 51.
Gyldenkerne, K. and Jaeger, J. R., 1963, Publ. mindre Medd. Kbh. Obs., No. 177, 6.
Herczeg, T., 1964, Veroff. astr. Inst. Univ. Bonn, No. 69.
Huang, S. S., 1963, Astrophys. J., 138, 342.
Larsson-Leander, G., Ark. Astr., (in print).
Magalahsvili, N. L., and Kumsishvili, Ya. I., 1960, Abastumansk. astrofiz. Obs. Gore
Kanobili Bjul., 25, 91.
Prager, R., 1931, K1. Veroff. Univ. Sternw. Berlin-Babelsberg, 3, No. 10. 125.
Sahade, J., Huang, S.-S., Struve, O., and Zebergs, V., 1959, Trans. Am. Phil.
Soc., 49, Part 1.
Saidov, K., 1959, Astr. Cirk. Isdav. bjuro astr. Soobshch, Kazan, No. 158, 12.
Stebbins, J., 1916, Lick Obs. Bull. 8, 186.
Struve, O., 1959, Publ. astr. Soc. Pacific 71, 441.
Wood, D. B. and Forbes, J. E., 1963, Astr. J., 68, 257.
Wood, D. B, and Walker, M. F., 1960, Astrophys. J., 131, 363.
DISCUSSION
Bakos: In your paper you mentioned that the B-V does not change during
the secondary minimum. From your light-curve I notice a small change
to the blue at the time of secondary minimum.
Larsson-Leander: It is very small, so it may not be real
Sahade: Were there spectra taken during the campaign?
Larsson-Leander: Yes, in the U. S. A.