The title of this talk is really just a different phrasing from one I have used at several IAU meetings on the subject of numbers and kinds of variables in globular clusters. To furnish this material, I have finished the Third Catalogue of Variables in Globular Clusters. Since many of you are coming to this Colloquium with new information, the Catalogue is in draft form with a request that corrections and additions be given me by October 2, after which the draft will go to the printer.

The First Catalogue of Variables in Globular Clusters was published at this observatory in 1939 and the Second Catalogue in 1955. In 1966 appeared the excellent Catalogue of Variables South of Declination—29° by Fourcade, Laborde and Albarracin, with splendid large prints of identification charts.

Three quarters of a century separates us from the beginning of systematic observations of variable stars in globular clusters. During the last 45 yr the name of Mrs Prof. Dr H. B. Sawyer Hogg has always been in the astronomical literature. Her activity is closely connected with investigations of variable stars in globular clusters.

The complex dependences between the presence of variable stars of different types and the properties of the globular clusters were discovered long ago. For instance, the RR Lyr variables do not occur in clusters of high metallicity. Red variables on the contrary do not occur in clusters of very low metallicity. The presence and peculiarities of variable stars have been used for the classification of globular clusters (van Agt and Oosterhoff, 1959; Castellani et al., 1970).

The present communication reports some results of observations made at Asiago on three faint globular clusters discovered on Palomar Sky Survey plates (Abell 1955).

Palomar 2 (4h43.m1, + 31 °23’, 1950; l= 170°, b= -9°). This cluster was studied some years ago by McCarthy and Treanor (1964) on infrared plates taken with the Schmidt telescope of the Vatican Observatory. After an extensive discussion the authors concluded that the object was either a peculiar globular cluster at an estimated distance of 63 kpc or an old galactic cluster at a distance of the order of 16 kpc.

Planetary nebulae are used as a tracer for halo Population II stars. A comparison of the number of planetary nebulae in the galactic pole caps (|b| ≥ 45°) with the number of planetaries in globular clusters suggests that (to within a factor of 2 or 3) the galactic halo has a luminosity LB ⋍ 2 × 108L⊙. From the work of Oort it is estimated that the galactic halo has a mass of at least so that for the galactic halo.

HD 116745 is an eleventh magnitude early F-type giant occurring in the field of the globular cluster ω Cen. Membership in the cluster appears virtually certain since the star’s radial velocity differs insignificantly from the mean cluster velocity of +238 km s-1. HD 116745 lies nearly 4 mag. above the horizontal branch in the HR diagram and is presumably in a rapid, advanced stage of evolution. Spectrograms at 22.5 Å mm-1 and 155 Å mm-1 have been obtained by one of us (R.J.D.) at the Radcliffe Observatory in South Africa using a McGee spectracon image tube. The lower dispersion spectra have been used to measure Hy profiles and the Balmer discontinuity, and one high dispersion film has provided the basis for a differential curve-of-growth analysis with respect to the Sun in order to determine some heavy element abundances. Physical data derived for the star are Mv= –3.32, θe=0.77±0.02, logg = 1.0±0.3 and Results of the curve-of-growth analysis, in which a model atmosphere was used to determine the curve appropriate to the star, yield a logarithmic iron-to-hydrogen abundance ratio with respect to the Sun of [Fe/H]= —1.2 with a provisional estimated uncertainty of about +0.2. This result is based on 28 lines of Fe i. Ti may be marginally overabundant, with a value of [Ti/Fe] = + 0.4, based on 7 lines of Ti II. Other elements and ionization states identified are Mg I, Al I, Si I, Ca I, Sc II, VI, VII, Mn I, Sr II, Zr II (Eu II).

Photoelectric observations obtained with the 24-in. telescope of the Konkoly Observatory from 1950 to 1972 show that the amplitudes of the changes of the 0.d57 light-curve in the course of the 41-day cycle undergo considerable cyclic variations. The length of the cycles varies between 3.8 and 4.8 yr. A new cycle always begins with a large phase-shift in the 41-day period; afterwards there is generally no further shift during one 4-yr cycle. Following Julia Balázs’ hypothesis, a tentative model is suggested: The 41-day cycle is the rotation period of the star. Magnetic fields concentrated in a limited longitude zone influence the 0.d57 pulsation in such a way that variations in light amplitude and phase arise as aspect effects of the 41-day rotation period. In this way, the 4-yr cycle can be interpreted as the magnetic cycle of the star. The location of the magnetic fields strongly differs in longitude from cycle to cycle causing the observed large phase shifts at the beginning of a new 4-yr cycle.

Interest in dwarf spheroidal galaxies is motivated by a number of reasons; an important one on the occasion of this colloquium is the abundance of variable stars. The theory of stellar evolution and stellar pulsations is now able to predict from theoretical considerations characteristic properties of variable stars in the colour-magnitude diagram (Iben, 1971). By observing the variable stars in the field, and in as wide a selection of objects as possible, more insight can be obtained into the history of the oldest members of our Galaxy (the globular clusters) and of the dwarf spheroidal galaxies in the Local Group. It is worthwhile to explore the spheroidal galaxies as observational tests for the theoretical predictions of conditions in space away from our Galaxy. The numbers of variable stars in the dwarf spheroidal galaxies are such that we may expect well-defined relations to emerge once reliable magnitude sequences have been set up, the variable stars found, and their periods determined. Six dwarf spheroidal galaxies are presently known in the Local Group within a distance of 250 kpc. In Table I, which lists members of the Local Group, they are at the low-luminosity end of the sequence of elliptical galaxies (van den Bergh, 1968).

RR Lyrae variables play a prominent role in many of the problems of globular clusters, and from several points of view. In the first place they can be considered as pretty good indicators of population and distance; although they do not form a completely homogeneous set of stars, the knowledge of their mean absolute magnitude gives a powerful means of establishing distances within and outside the Galaxy, and hence of determining the form and size of the Galaxy itself. Moreover, the number of RR Lyrae stars in clusters, the relative frequency of RRc and RRab, types, the length of the transition periods, the array of colors, when correctly interpreted, give important information on the degree of evolution, age and chemical composition of the clusters. Placed as they are in a peculiar region of the H — R diagram of Population II, the RR Lyr variables can be used as a good test of the theories of advanced evolution or the models of pulsating stars.

115 new variable stars have been detected in a one square degree field centered on the globular cluster NGC 6401. Two of them (No. 157 and 164) seem to belong to the cluster.

The globular cluster NGC 4590 has been observed at Haute Provence Observatory (photographic photometry, 80-cm reflector, F/6) and at the European Southern Observatory (image-tube, 152-cm reflector, F/15).

• – A sequence of red magnitudes, mr (λett ≅ 6400 Å) has been established around the cluster. It comprises 28 stars with magnitudes between 11.66 and 16.50.

• – Five new variable stars were discovered in a one square degree field, centered on the cluster.

• – Four other variable stars (No. 39, 40, 41, and 42) were discovered in the cluster.

NGC 6981 =M72 is a globular cluster with a late-type integrated spectrum (G3) and an ultraviolet excess of δ(U–B)=0.15 ±0.05, which places the cluster in the class of intermediate-to-high metal abundance clusters such as M3. The colour-magnitude diagram shows a horizontal branch populated on both sides of the variable star region. Previous work on the periods of the variables has indicated that the cluster is of Oosterhoff type I, with ≈0.55 day. Some results from B and V photometry of 21 RR Lyrae variables are described. Original measurements of about 50 candidate variables were made, including some possible new variables found by blinking plate pairs, but crowding and background photometric problems reduced the number of variables for which reasonable quality data could be obtained with the available 100-in plate material.

I have recently prepared for publication a study of 100 field RR Lyraes which were measured on the u, k, b, y, β system described by Jones (1971). It is shown that the following indices

β1 = β − 0.125(k − b) + 0.103(b−y)

(u − b)3 = (u − b) − 0.93(b − y) − 0.62(k − b) − 1.95(β − 2.75)

(k − b)2 = (k − b) − 0.826(b − y) + 0.25(β − 2.75)

correlate strongly with θeff, logg and [Fe/H] respectively. Figures in that paper show the tightness of the correlations.

It has been shown by the author (Geyer, 1967), and independently by Dickens and Woolley (1967), that there occur horizontal branch stars well within the RR Lyrae variable gap of the c–m-diagram of ω Centauri. Variability of these objects is not known, and presumably is unlikely, because the cluster is so well studied for variable stars. For the observationally best studied globular clusters M3, M13, M15 and M92 (Sandage, 1969), this phenomenon is not found. The author’s original published c–w-diagram of NGC 5139 showed about 10 stars with (B–V)≥0.2 and V-magnitudes comparable with those of RR Lyrae stars of the cluster. Unpublished photoelectric measurements of some of these stars, carried out in 1968 with the 40-in photometric reflector of ESO by the author, support this finding. Also the study of blue horizontal branch stars by Newell, Rodgers and Searle (1969) of the Herst-monceux catalog of ω Centauri (Woolley et al., 1966) confirms these results, the importance of which for the understanding of the horizontal branch stars and RR Lyrae variables is obvious.

Two-color photographic photometry has been carried out on RR Lyrae variables in M14. The position of these variables in the color-magnitude diagram is shown, and it is noted that there is a lack of horizontal branch stars to the red of the variable star gap. Color-amplitude, period-amplitude and color-period relations for the RR Lyrae variables measured are shown.

The period of a variable star is very sensitive to extremely minor changes in luminosity, mass, radius, surface temperature, etc. of the star. With new theoretical calculations of the interior structure of stars in the late stages of evolution now available, the investigation of instabilities in the periods of the RR Lyrae variables in globular clusters becomes urgent.

New observations of the variables in M15 have been made with the Yale 40-in. reflector for the study of periods and period changes.

Since the counting of cycles from the time of Bailey’s observations to the present day cannot always be made unambiguously, there is a corresponding uncertainty in the nature and amount of the period changes. Good lightcurves are obtained even when the periods used for their reduction are incorrect because the number of cycles in a day, or in a month, or in a year, is wrong by one. When this happens, however, period changes are likely to be indicated, and these will be spurious.

Nevertheless, it can be shown that the periods of most variables have not been constant during the interval 1900–1970.

The situation could be improved if a relatively small number of plates were to be obtained each year for several decades, preferably with the same instrument. Individual accuracy could be improved by the use of the electronic camera.

The full contents of this paper will be published elsewhere.

Much has been said at this colloquium to the effect that observed changes in the periods of variable stars are so fast or so abrupt or so quick to change sign that they cannot possibly be attributed to stellar evolution. If you are wondering how this report can dare to make a contradictory claim read the title carefully. It does not promise that there really is evidence for evolution in Omega Centauri. It only proposes to look at the period changes with evolution in mind. We may all decide that the evidence is strong or that it is weak or perhaps that there is no evidence at all.

NGC 1466 (α1950 = 3h44.m6, δ1950= -71°45’) is a globular cluster which appears to be situated between the two Magellanic Clouds. Previous estimates (Gascoigne, 1966) put it at roughly the same distance from us as the LMC, so it is regarded as a member of the Cloud system. It is globular in appearance, and its colour-magnitude diagram confirms this classification. It has a fairly well-developed horizontal branch, and was found by Wesselink (1970) to be quite rich in variables. The metallicity index, Q, (van den Bergh, 1967) has a value of -0.36 for NGC 1466 (Andrews and Lloyd Evans, 1971). This would rank it with M5 and NGC 6171 as a cluster of intermediate metal content. This comparison is consistent with the value of Δ V for the cluster, which, at 2.m6, is representative of the Δ V values of globular clusters of intermediate metal abundance in the Galaxy.