5.4. Colour Indices and the Coupled Evolution of Double Galaxies
Examining the various characteristics of galaxies in pairs we have repeatedly sought correlations between these properties which could not be attributed to selection effects. These attempts extend as well to the colour indices of double galaxies. The first such attempt at examining the colours of members of paired systems was performed by Holmberg (1958). For a sub-sample of 32 pairs with photographic colour index B - V he found a correlation coefficient of +0.80 ± 0.6. For a long time the Holmberg effect received attention from neither observers nor theoreticians. After almost two decades, many photoelectric measurements of double galaxies were performed by Tomov (1978, 1979), and Vardanian and Tomov (1980). Demin et al. (1981) reduced these data to the standard (U, B, V)T0 of the RCBG (de Vaucouleurs et al., 1976). These authors confirm the existence of the Holmberg effect using two colour indices (B - V)T0 and (U - B)T0, with the following correlation coefficients: +0.79 and +0.81 respectively for EE pairs, +0.71 and +0.86 for SS pairs, and +0.51 and +0.63 for ES pairs. Among the 105 pairs in the compilation of photoelectric data by Demin et al. (1981) there are 68 isolated double systems with f < 100 from our catalogue. The mean values of the colour indices for these, and their dependence on morphological type, are presented in Table 25. The same data are shown for the mean values of field galaxy colours, according to the RCBG. The numbers in parentheses indicate galaxy types coded in the RCBG system.
Further analysis of the photoelectric observations by Tomov was performed by Demin et al. (1984), who found that the distribution of pair members according to colour index could not be accounted for solely by the known correlation with morphological type. NOTE: This had already been concluded by Holmberg. This holds even in the presence of a colour correlation in the selection of EE pairs. Demin and coauthors examined three sub-catagories of double systems: EE, ES, and SS pairs. They noted the properties of the distribution of colours for each sub-sample and found separate explanations for each. We will repeat here their basic conclusions, along with a short commentary. Elliptical galaxies located in EE systems are almost indistinguishable in colour from other elliptical galaxies, so that anomalies in chemical composition appear very unlikely. It is known that the colour index of elliptical galaxies depends on their luminosity (Sandage and Visvanathan, 1978), so the run of colours for members of the EE systems might be explained by the observed correlation between their luminosities. However, Demin (1984) showed that the colour-luminosity correlation for E galaxies in pairs is practically absent. With rare exceptions elliptical galaxies do not contain significant populations of hot young stars. The main factor influencing their colour is the chemical composition of the red giant stars contributing the bulk of the visible luminosity of the galaxy. Therefore, the most important factor in setting the colour indices for galaxies in EE pairs must be the similarity of their chemical compositions. This coupling of chemical contents in double galaxies might be due to similar epochs of first star formation, or to similar protogalactic environments.
Mixed ES double systems exhibit interesting colour asymmetries. While the spiral components of mixed pairs are no different in colour from members of SS pairs of the same Hubble classification, the elliptical galaxies in mixed pairs are generally bluer than the components of EE pairs. Earlier, Smirnov and Komberg (1980) drew attention to the fact that elliptical galaxies with a significant gas content occur predominantly as members of small systems with neighbouring spiral galaxies, and proposed the possible transfer of gas from the gas-rich spiral galaxies to the elliptical, with the accompanying formation of blue stars. This idea can explain the correlation of colours in galaxies in ES pairs. The more blue stars and their progenitor gas in the spiral members of pairs, the greater the chance of producing star formation in the elliptical members. Such gas introduced into elliptical galaxies in pairs should have a greater effect in the very closest interacting systems. However, the probability of encountering emission lines in E galaxies in mixed pairs shows no evidence of such an effect (see section 5.3). In order to clarify the occurrence of active gas transfer in ES systems it would be necessary to obtain accurate photoelectric and spectrophotometric observations of these objects and to confirm their expected morphological classification by obtaining large scale images.
Pairs consisting of spiral galaxies also show evidence of coupled evolution. Ardeberg and Bergvall (1977) noted that spiral galaxies in interacting pairs have bluer colours than do spirals in general and that pair members share similar colours. The presence of a blue excess in double spirals was confirmed by Sharp and Jones (1980), Bergvall (1981a) and Adams et al. (1980). The last authors, along with Smirnov and Komberg (1980), compared these properties with the idea of Larson and Tinsley (1978) about bursts of star formation which could periodically recur in SS pairs, depending on the type of interaction. Incorporating the mean value of the excess blue light, Sharp and Jones (1980) concluded that the total rate of star formation was increased in double galaxies by 1.5 to 2 times, in agreement with White and Valdes (1980) and Karachentsev (1981c).
Examining pairs of spiral galaxies, Demin et al. (1984) drew attention to several tendencies. The colour differences between components decrease on passing from wide to narrow pairs. The more tightly coupled values of the colour index are also found for small values of the radial velocity difference. Thus, in the majority of contact pairs with y < 100 km/s, the component colour indices do not differ by more than 0.1 magnitude. This property is easily explained if the correlation of colours in spirals in double galaxies is produced by the interaction. According to Demin et al. (1984) the tightly coupled colour indices of double spirals arise primarily from nearly simultaneous increases of the star formation rate in both systems. The greatest blue excess occurs in galaxies of type Sa (8) , where the number of blue stars in comparison to that in other galaxies may be larger because of its smaller original population, so that the colour amplitude produced by the burst of star formation would be more noticeable. Smirnov and Komberg (1980) showed that such a burst of star formation should be accompanied by an increase in the velocity dispersion of interstellar gas clouds in the interacting galaxies, which would decrease further star formation and put galaxies into an `anti-burst' phase. This postulated transition of interacting doubles between phases of increased and decreased star formation is probably the main mechanism accounting for the properties of the observed distributions of the colour index for spirals in double systems. The increased number of Markarian galaxies in pairs would, in this case, represent the extremes of induced star formation.
Larson and Tinsley (1978, 1980) and Silchenko (1982) calculated evolutionary tracks for galaxies of various types in the two colour (U - B, B - V) diagram after bursts of star formation of various intensities. According to these results, the characteristic timescale for such a burst to fade is 2 × 108 years, which is of the same order as the typical orbital period of a pair and the period of rotation of its components. The near match of these timescales with one another may introduce strong resonances among spiral galaxies, an evolutionary effect which has received little theoretical study. It is apparent that the tracks of synchronized evolution for double galaxies in colour-colour diagrams should reveal not only their gaseous contents but also the orbital parameters, such as eccentricity and distance of closest approach. Therefore, the construction of more usable models of interaction and the resulting star formation in pairs should lead to more precise determinations of the character of their orbital motions.
In summary, we may say that the properties of the colour indices of double galaxies contain important clues to the character of their evolution. In pairs of elliptical galaxies the distribution of their colours is most probably of a relic nature, due to their similar chemical compositions, and in pairs with spiral components the distribution indicates the dominant role played by coupled star formation processes. Note that our ability to deal with this question increases with the accuracy of the photoelectric measurements. The data of Tomov (1978, 1979) by comparison with the RCBG have colour measurement errors (B - V) 0.10m and (U - B) 0.15m. Increasing the accuracy by factors of three to five may produce more precise conclusions about the evolution of galaxies in double systems. Great possibilities have now been opened up by the existence of simultaneous multi-channel spectral instruments. For example, Beaver et al. (1974) used a 40-channel spectral scanner to do surface photometry of faint tails between members of interacting pairs (3% of the brightness of the night sky). Unexpectedly, these observations showed red colours in the tail (B - V = + 1.00m), which indicate a composition of old stellar populations. These kinds of sensitive photoelectric observations are important for understanding the role of disk and spheroid populations of stars in interactions.
8 According to data in the RCBG the mean colour of Sa galaxies in pairs is slightly redder than according to Demin et al. (1984). This may be due to small systematic differences in the colour systems. Back.