Short-period near-contact binary systems at the beginning of the overcontact phase

Transcription

1 Mon. Not. R. Astron. Soc. 336, (2002) Short-period near-contact binary systems at the beginning of the overcontact phase Shengbang Qian 1,2 1 Yunnan Observatory, National Astronomical Observatories, Chinese Academy of Sciences, PO Box 110, Kunming, China 2 United Laboratory of Optical Astronomy, Chinese Academy of Sciences (ULOAC), Beijing, China Accepted 2002 July 15. Received 2002 July 4; in original form 2001 November 29 1 INTRODUCTION Near-contact binaries (NCBs) displaying decreasing period are a very important source in understanding the formation and evolution of overcontact binary systems, since they are possible examples of binaries that are to become overcontact or are in the broken-contact phase as predicted by the thermal relaxation oscillation (TRO) theory of overcontact binaries (Lucy 1976; Flannery 1976; Robertson & Eggleton 1977). Members of this group have periods of less than 1 d and show an EB-type light curve where the light variation is qsb@netease.com ABSTRACT A detailed analysis of orbital period changes of seven near-contact binary stars (NCBs) (BL And, V473 Cas, XZ CMi, BV Eri, RU Eri, UU Lyn and GR Tau) with period less than 1 d has been performed and their respective O C diagrams are formed and discussed. It is found that all systems analysed show secular period decreasing. For V473 Cas, the analysis of the period change was performed based on data collected by Moschner, Frank & Bastian. For XZ CMi, its period shows some complex changes, a possible cyclic oscillation is discovered to superpose on the secular decrease that can be explained either by the presence of a third body or by magnetic activity cycles of the components. Since the third-body assumption is consistent with the photometric solution of Rafert, XZ CMi may be a truly triple system. For BV Eri, the period decrease is only supported by weak evidence. All the seven systems are short-period NCBs with AF-type primary components where both components are filling or nearly filling the critical Roche lobe. As the period decreases, the separation between both components will be reducing and thus these systems will evolve into A-type overcontact binaries. The period decrease may be caused by mass transfer or/and by angular momentum loss via magnetic braking. Combined with the published data on the other systems of the same type, a possible statistical connection between orbital period P and its rate of decrease dp/dt is obtained: dp/dt = P dyr 1. This correlation indicates that the smaller the orbital period P is, the smaller its rate of change dp/dt will be. The correlation found in this paper indicates that there may be a smooth transition from A- and F-type NCBs with period decreases to the A- and F-type overcontact binaries that have period increases, and in that sense one may postulate that the NCBs may be the progenitors of the A-type W UMa systems and will be oscillating around a marginal-contact state as predicted by thermal relaxation oscillation (TRO) theory. Key words: binaries: close stars: evolution stars: mass-loss. continuous, but with a large difference of the depths of the two minima. The primary component of an NCB is usually of spectral type A or F and the secondary can be of type G or K. Several photometric solutions (see, for example, Kaluzny 1985; Yamasaki, Okazaki & Kitamura 1984; Hilditch & King 1988; Hilditch, King & McFarlane 1988) have shown that the systems of this group could be marginally contact, semidetached or even marginally detached, and, in particular, different configurations could be obtained by different investigators for one NCB star (such as BV Eri and GR Tau). Recently, based on the collection and analysis of times of light minimum, orbital period decreases among NCBs have been surveyed. In the present work, orbital period decreasing of 7 such systems, BL And, V473 Cas, XZ CMi, BV Eri, RU Eri, UU Lyn, and C 2002 RAS

2 1248 S. Qian Table 1. Physical properties of eight NCBs. Star name Sp. Period (d) Configurations Ref. BL And A NCB with both components marginal contact (1) V473 Cas With EB light curve, configuration is unkown (2) XZ CMi F NCB with the secondary filling (3) NCB with both components nearly filling (4) BV Eri F NCB with both components nearly filling (5) NCB with both components marginal contact (6) RU Eri F3-4IV-V NCB with both components nearly filling (7) UU Lyn F3V NCB with both components nearly filling (8) GR Tau A5V NCB with both components nearly filling (9) NCB with the secondary filling (10) References in Table 1: (1) Kaluzny (1985); (2) Kholopov (1985); (3) Terrell et al. (1994); (4) Maradirossian & Giuricin (1981); (5) Baade et al. (1983); (6) Gu (2000); (7) Nakamura et al. (1984); (8) Yamasaki et al. (1983); (9) Yamasaki (1984); (10) Lázaro et al. (1995). GR Tau, are reported. Then a possible statistical correlation between the orbital period P and its rate of decrease dp/dt for NCBs is given. Finally the evolutionary states of the NCBs and a possible evolutionary link between NCBs and hotter overcontact binary systems (M 1 > 1.35 M ) are discussed. The general properties of the sample stars are shown in Table 1. 2 ORBITAL PERIOD CHANGES FOR SOME NCBs 2.1 BL Andromedae The light variability of BL And was discovered by Hoffmeister (1943). Photographic observations were published by Kljakotko (1951) and some times of light minimum were observed by Huth (1957). The first photoelectric light curves in B and V were published by Vetesnik (1967) and the data were analysed by the same author with the method of Russell & Merrill. Those observations was later reanalysed by Kaluzny (1985) with the Wilson Devinney method. It was shown that BL And is a marginal contact system where both components fill the critical Roche lobe but are not in thermal contact. Although epochs and orbital periods of BL And were given by several authors (e.g. Kljakotko 1951), it was neglected for period study. In order to see whether the orbital period is variable and to understand the properties of the period change, all available eclipse times are collected. 26 timings were tabular in the paper of Vetesnik (1967), and 49 times of light minimum were compiled at Eclipsing Binaries Minima Database (EBMD) (available at dn.html). After those collections, one charge-coupled device (CCD) timing and two photoelectric times of light minimum have been published by Nelson (2001) and by Agerer & Hubscher (2001a). The O C curve based on Kljakotko s linear ephemeris, HJD(min I ) = d E, (1) is displayed in Fig. 1. Although the O C values of visual or photographic data show large scatter (up to d), the general O C trend suggests that the period of BL And is decreasing. Assigning weights 1 to visual or photographic data and 8 to photoelectric or CCD observations, a least-squares solution yields the following quadratic ephemeris with mean error for each term: HJD(min I ) = (7) + 0 ḍ (2) E 2.60(9) E 2. (2) Figure 1. O C diagram of BL And where open circles refer to visual or photographic observations and dots to photoelectric or CCD data. Also given in solid line is its general trend described by a quadratic ephemeris. From the quadratic term, a period decrease rate of dp/dt = dyr 1 is determined. 2.2 V473 Cassiopeiae The light variation of the short-period binary star, V473 Cas, was also detected by Hoffmeister (1964). After its detection, photographic observations were obtained by Gessner & Meinunger (1973), who classified the system as an eclipsing binary with a photographic magnitude range as The first complete CCD light curve was published by Moschner, Frank & Bastian (1999). Their data showed that the light variation of V473 Cas belongs to β Lyr type. All published times of light minimum of V473 Cas were collected by Moschner et al. (1999) who derived the following linear ephemeris: HJD(min I ) = ḍ E. (3) They pointed out that the period of the binary star is variable. Based on all O C values (displayed in Fig. 2) computed by Moschner et al. (1999), a weighted least-squares solution with the same weights as those applied to BL And leads to the quadratic elements:

3 Short-period near-contact binary systems 1249 Figure 2. The same as Fig. 1, but for V473 Cas. HJD(min I ) = (8) + 0 ḍ (2) E 3.91(39) E 2, (4) and a period decrease rate of dp/dt = dyr XZ Canis Minoris Hoffmeister (1934) discovered the light variability of XZ CMi (BD ) and Lause (1938) obtained light elements and a visual light curve for the eclipsing binary star. The first photoelectric light curves in B and V were published by Wilson (1966) who pointed out that a satisfactory Russell model solution was very difficult to derive without a large amount of third light (at least 25 per cent of the total system light). Wilson s data were later analysed with the modern light-synthesis method by several authors (e.g. Mardirossian & Giuricin 1981; Terrell & Wilson 1990; Rafert 1990). The solutions of Rafert (1990) with the Wilson Devinney method indicated a large amount of third light with l 3 = 0.17±0.03 for the V light curve and l 3 = 0.11 ± 0.04 for the B light curve in units of total flux of the system, while Terrell & Wilson (1990) derived a good fit without third light. However, as pointed out by Rafert (1990), the Terrell & Wilson attempt to assess the need for third light was made in later runs and their solution is therefore probably less reliable. Terrell, Gunn & Kaiser (1994) published new light curves and a solution of their data. Their solution does not indicate the presence of third light. It is shown that XZ CMi is a NCB with the secondary component filling its critical Roche lobe. The near-contact configuration and the possible presence of a third body both made XZ CMi a very interesting system to study. Wilson (1966) collected 20 times of light minimum and displayed an O C diagram. After his compilation, 94 timings were collected at the EBMD and one photoelectric timing was published by Ogloza et al. (2000). The (O C) 1 value of those eclipse times is calculated with the following light elements given by Wilson (1966): HJD(min I ) = ḍ E. (5) The corresponding (O C) 1 curve are shown in Fig. 3. As displayed in this figure, the orbital period is changeable. Assuming a secular period decrease, and with the same weights as those used for BL And and V473 Cas, a least-squares solution yields the following parabolic ephemeris: Figure 3. The same as Figs 1 and 2, but for XZ CMi. HJD(min I ) = (10) + 0 ḍ (5) E 2.66(37) E 2, (6) and a rate of decrease in the orbital period dp/dt = dyr 1. The residuals [(O C) 2 values] of all photoelectric and CCD data, from the parabolic ephemeris, are calculated and are listed in Table 2. The corresponding (O C) 2 diagram is shown in Fig. 4 where a periodic change may exist. This implies that apart from the long-time decrease in the orbital period, there is a possible periodic component to superpose on it. With least-squares method, the following equation is obtained: (O C) 2 = (1) + 0 ḍ 0051(5) sin[ E (±2. 4)], (7) which can yield a best fit to the residuals (solid line in Fig. 4). This formula clearly suggests a periodic variation with a semi-amplitude of d and a period of 30.5 yr. If the periodic change is caused by the presence of a third body, then by considering a total mass of 3.0M, the orbital parameters of the assumed body for several different values of i are listed in Table 3. On the other hand, XZ CMi contains an F0 primary star and a cooler secondary star. Table 2. Photoelectric and CCD data for XZ CMi. JD Hel. Meth. E (O C) 1 (O C) 2 Ref pe (1) pe (1) pe (2) pe (2) pe (2) pe (2) pe (3) pe (4) pe (4) CCD (5) CCD (6) CCD (7) pe (8) References in Table 2: (1) Wilson (1966); (2) Gimenez & Costa (1979); (3) BAV-M 36; (4) Terrell et al. (1994); (5) AAVSO 2; (6) Agerer & Heubscher (1998); (7) Agerer & Heubscher (1999); (8) Ogloza et al. (2001).

4 1250 S. Qian Figure 4. (O C) 2 residuals in days of XZ CMi after subtracting of the bestfitting parabola given in equation (6). Solid line refers to the theoretical solution of an assumed third body. Table 3. Orbital parameters of a possible third body in XZ CMi. Parameters Values Units A d T 30.5 yr a 12 sin i 0.88 A.U. f (m) M m 3 (i = 90 ) 0.20 M m 3 (i = 70 ) 0.21 M m 3 (i = 50 ) 0.26 M m 3 (i = 30 ) 0.41 M m 3 (i = 10 ) 1.40 M a 3 (i = 90 ) A.U. a 3 (i = 70 ) A.U. a 3 (i = 50 ) A.U. a 3 (i = 30 ) A.U. a 3 (i = 10 ) A.U. The cyclic change may be caused by the variation of the gravitational quadruple momentum via the cyclic magnetic activity of the components. 2.4 BV Eridani The eclipsing variable BV Eri was discovered by Hoffmeister (1933). Later, the first photoelectric light curves were obtained by Baade et al. (1979). In 1983, the same authors published the first radial velocity curve and reported the absolute parameters of the system (Baade et al. 1983). Recently, BV Eri has been observed photoelectrically by Gu et al. (1994), they collected all available times of light minimum and gave a revised ephemeris. After the compilation of Gu et al. new photoelectric timings were published by Agerer & Heubscher (1996, 1998b, 1999). By using the ephemeris obtained by Gu et al. (1994): HJD(min I ) = ḍ E, (8) the O C values of these timings are computed. The corresponding O C curve is displayed in Fig. 5. As shown in the figure, the orbital period of the system is variable. With the same weights as those used for BL And, V473 Cas, and XZ CMi, a least-squares solution leads to the following quadratic ephemeris: Figure 5. The same as Figs 1,2,and3,butforBVEri. HJD(min I ) = (33) + 0 ḍ (66) E 1.92(44) E 2. (9) From this ephemeris, the period decrease rate is determined to be dp/dt = dyr 1. However, since the O C values, especially those of visual and photographic data, show large scatter, new times of light minimum are needed to check the present period change. 2.5 RU Eridani The light variation of RU Eri was detected by H. S. Leavitt (Pickering 1907) and it was further observed by Zinner (1913) and Dugan, Piercel & Wood (1951). The first photoelectric observations were obtained by Nakamura & Tanabe (1953). The complete photoelectric light curves in B and V were carried out by Sarma & Sanwal (1981) and they determined the period of the system to be d. The first radial velocity curve for the primary component was given by Nakamura, Yamasaki & Kitamura (1984), and combined with the photometric solution based on Sarma & Sanwal s photoelectric observations, the absolute parameters of RU Eri are obtained. Their study shows that RU Eri is a nearcontact binary with both components almost filling their critical Roche Lobes. The orbital period of the system has been studied by several authors. Nakamura & Tanabe (1953) suspected that the period is variable; however, the O C curve formed by Sarma & Sanwal (1981) indicated no significant period changes. 56 times of light minimum published between 1912 to 1984 were collected by Nakamura et al. (1984). They pointed out that the period is constant in this time interval. After the compilation of Nakamura et al. (1984), many eclipse timings were compiled at the EBMD. In order to see the general behaviour of the change in the orbital period, the O C values of those timings are calculated with the ephemeris of Sarma & Sanwal (1981): HJD(min I ) = ḍ E. (10) The O C plot is displayed in Fig. 6 where circles refer to visual and photographic observations, and solid dots to photoelectric data. Although the O C values, especially those of the visual and photographic data (up to 0 ḍ 03 as displayed in Fig. 6), show large scatter, a general trend in the O C curve may exist. With the weighted least-squares method, the following ephemeris:

5 Short-period near-contact binary systems 1251 Figure 6. The same as Figs 1, 2, 3, and 5, but for RU Eri. HJD(min I ) = (7) + 0 ḍ (11) E 3.0(4) E 2, (11) is obtained. If only the photoelectric and CCD times of light minimum are used (listed in Table 4), we can obtain HJD(min I ) = (4) (8) E 7.1(8) E 2, (12) which can give a good fit to the photoelectric timings (solid line in Fig. 7). The quadratic term corresponds to a period decrease rate of dp/dt = dyr UU Lyncis UU Lyn was detected to be an eclipsing binary by Geyer et al. (1955) from his photographic observations. Strohmeier, Knigge & Ott (1963a) published a complete photographic light curve and found the period of the system to be d. The spectral type was determined by Götz & Wenzel (1962) to be A4. The first photoelectric light curves and radial velocity curves were published by Yamasaki, Okazaki &Kitamura (1983). They reported a spectral type of F3V for primary and K type for the secondary, and from the analysis of the photoelectric light curves and the radial velocity curve they determined the absolute parameter of the system. Their study showed that UU Lyn is a near-contact binary with both components almost filling their critical Roche Lobes. Table 4. Photoelectric and CCD data for RU Eri. JD Hel. Meth. E O C Ref pe (1) pe (2) pe (2) pe (2) pe (2) pe (2) pe (2) pe (3) CCD (4) References in Table 6: (1) Nakamura & Tanabe (1963); (2) Sarma & Sanwal (1981); (3) Wolf et al. (1982); (4) Paschke (1996). Figure 7. O C plot in days for RU Eri formed by all photoelectric and CCD data. In order to see the general trend in the change of the orbital period, all published times of light minimum are analysed. 50 visual and photographic minimum times have been collected at the EBMD and two photoelectric times of light minimum were published by Yamasaki et al. (1983) and by Van Cauteren & Vils (2000). With the following ephemeris determined by Flin (1981): HJD(min I ) = ḍ E, (13) the O C values of these times of light minimum are calculated and are displayed in Fig. 8. Although showing large scatter, a general trend in the O C curve is apparent and the least-squares solution leads to the quadratic ephemeris: HJD(min I ) = (4) + 0 ḍ (3) E 6.46(2) E. (14) From this ephemeris, the rate of decrease in the orbital period is determined to be: dp/dt = dyr GR Tauri The variable star, GR Tau (BD ), was discovered as an eclipsing binary by Strohmeier, Knigge & Ott (1963b). The first Figure 8. The same as Figs 1, 2, 3, 5 and 6, but for UU Lyn.

6 1252 S. Qian photoelectric light curves and the radial velocity curves were published by Yamasaki et al. (1984) who reported a spectral type of A5V and a revised period of P = 0 ḍ From the analysis of their photoelectric observations they determined the first photometric elements of the system. Later, photoelectric observations of GR Tau were published by many authors (e.g. Faulkner 1986; Hanzl 1990, 1991; Wunder et al. 1992; Mullis and Faulkner 1991; Fang et al. 1994; Lázaro et al. 1995; Agerer & Heubscher 1996, 1998a, 1999b, 2001b; Agerer, Dahm & Hubscher 2001) and one can find many times of light minimum in these literatures. Recently, a detail photometric study by Lázaro et al. (1995) with the W-D method has shown that GR Tau is a near-contact binary with the secondary component filling its critical Roche Lobe. The period of GR Tau has been reported to be variable by Fang et al. (1994). With all the published photoelectric timings collected from the aforementioned literature, the following revised ephemeris: HJD(min I ) = (16) + 0 ḍ (17) E, (15) is obtained. The residuals of all times of light minimum from this ephemeris are calculated and are plotted in Fig. 9 where open circles refer to visual and photographic data and dots denote photoelectric observations. Since the O C values of the visual and photographic timings show large scatter, starting from the first photoelectric timing, only the photoelectric and CCD data are displayed in this figure and are used for present period study. Although they display large scatter, a general trend in the O C curve may exist. With weight 1 to visual and photographic observations, weights 3 and 8 to secondary and primary times of light minimum observed with photoelectric method, the quadratic ephemeris: HJD(min I ) = (40) + 0 ḍ (5) E 2.49(16) E, (16) is determined. The quadratic term of this quadratic ephemeris corresponds to a period decrease rate of dp/dt = dyr 1. If only the photoelectric times of light minimum are used, and with the same weights as used before for the secondary and the primary eclipse times, the following quadratic ephemeris is derived: HJD(min I ) = (2) (3) E 4.74(13) E 2. (17) Figure 9. O C curve of GR Tau. Starting from the first photoelectric minimum, only photoelectric data are plotted. Figure 10. O C plot in days for GR Tau formed by all photoelectric data. Large solid dots represent primary times of light minimum and small solid dots denote the secondary ones. The coefficient of the ephemeris indicates a period decrease of dp/dt = dyr 1 which is the same order as that derived by using all data. As shown in Fig. 10, this ephemeris can give a good description to the photoelectric timings (solid line in Fig. 10). 3 DISCUSSIONS AND CONCLUSIONS The orbital periods of 7 AF-type short-period NCBs are studied based on the analysis of their O C observations. The secular period decreases of all systems are discovered and the decrease rate of each system is determined. This kind of change in the period is very typical for some NCBs, other example such as CN And (Samec et al. 1998a), FT Lup (Lipari & Sisteró 1986), RT Scl (Duerbeek & Karimie 1979), AK CMi (Samec, McDermith & Carrigan 1995), V1010 Oph (Lipari & Sisteró 1987), AG Phe (Cerruti 1994), and others also show such kind of period variation. The period of XZ CMi shows some complex variations, a possible periodic change is found to superpose on the long-time decrease. The periodic variation can be explained either by the light-time effect via the presence of an assumed third body or by the variation of the gravitational quadruple momentum via magnetic activity cycles of the cool components. By the analysis of the photoelectric data with Russell model, a large amount of third light in XZ CMi was proposed by Wilson (1966) and was later confirmed by Rafert (1990) with the Wilson Devinney method. Those results suggest that XZ CMi may be a truly triple system, which is in agreement with the period change of the system. In Section 2.3, orbital parameters of the third body are determined and are shown in Table 3. Rafter s photometric solution showed that the amount of third light is l 3 = 0.17 ± 0.03 for the V light curve and l 3 = 0.11 ± 0.04 for the B light curve in units of total flux of the system. If the third body is a main-sequence star, the parameters listed in Table 3 indicate that the orbital inclination of the third body should be very small (i < 15 ), which is much smaller than that of the eclipsing pair (i = 81. 8). If this is in the case, we can conclude that the third body is captured by the eclipsing binary star. The triple system is not like planetary systems (and galaxies) as formed by contracting spinning gaseous clouds. The situation of XZ CMi is the same as that of the Algol-type eclipsing binary system S Equ (Qian & Zhu 2002a). However, as that discussed by Qian &

7 Table 5. Orbital period decrease rates dp/dt for some NCBs. Short-period near-contact binary systems 1253 Star name Mass ratio Orbital period (d) dp/dt (d yr 1 ) Reference VW Boo Qian and Zhu (2002b) V473 Cas present BE Cep Samec et al. (1999) GR Tau Present CN And Samec et al. (1998a) UU Lyn Present FT Lup Lipari and Sisteró (1986) BV Eri Present RT Scl Duerbeck and Karimie (1979) AK CMi Samec et al. (1995) XZ CMi Present BO Peg Qian (2001a) RS Ind Marton et al. (1990) RU Eir Present BF Vir Qian et al. (2000) V1010 Oph Lipari and Sisteroó (1987) AV Hya Qian (2000) BL And Present AG Phe Cerruti (1994) V388 Cyg Young et al. (2001) TT Her Milone et al. (1989) Zhu (2002a), third light may be the result of a numerical artefact in the solution of W-D code, since it is usually strongly correlated with many parameters. To check the presence of the third body, new photometric and spectroscopic observations and a careful analysis of those data are required. The rates of the period decrease (dp/dt) of the present studied systems are listed in the fourth column of Table 5. Also shown in the same column are the period decrease rates for other NCBs collected from literatures, which are listed in the order of orbital period increasing. Of the 21 sample stars, the range of orbital period is from 0 ḍ 34 to 0 ḍ 91 with a mean value at 0 ḍ 57. Those listed in the second column of this table are the mass ratio of the sample stars. No photometric and spectroscopic studies of V473 Cas were published. The mass ratio of VW Boo was given by Rainger, Bell & Hilditch (1990), and those of BE Cep, GR Tau, CN And, UU Lyn, FT Lup, BV Eri, RT Scl, AK CMi, XZ CMi, BO Peg, RS Ind, RU Eri, BF Vir, V1010 OPh, AV Hya, BL And, AG Phe, V388 Cyg and TT Her were from Samec et al. (1999), Lázaro et al. (1995), Van Hamme et al. (2001), Yamasaki et al. (1983), Lipari & Sisteró (1986), Gu (1999), Hilditch & King (1986), Samec et al. (1998b), Terrell et al. (1994), Yamasaki & Okazaki (1986), Marton et al. (1990), Nakamura et al. (1984), Russo & Sollazzo (1981), Leung & Wilson (1977), Qian et al. (2000), Kaluzny (1985), Cerruti (1996), Young et al. (2001) and Milano et al. (1989), respectively. The sample stars can be divided into four groups: (1) semidetached systems (e.g. RT Scl, V388 Cyg, and TT Her) that have their primary components filling the critical Roche lobe; (2) semidetached ones with lobe-filling secondary components (e.g. CN And, RS Ind, BF Vir and AV Hya); (3) systems with both components nearly close to the critical Roche lobe (e.g. RU Eri and UU Lun); and (4) the remaining group with both components in marginal contact (e.g. VW Boo, FT Lup and BL And). However, it should be noted here that for some NCBs their configurations are difficult to determined exactly. A given system can be divided into several groups and different investigators usually obtain different configurations even by using the same data. This may be caused by the photometric disturbances and asymmetries on the light curve via mass and energy transfer between the components. The dp/dt of the NCBs are displayed graphically against P in Fig. 11 where diamonds refer to systems [usually members of groups (1) and (4)] that displayed evidence for primary to secondary mass transfer (PSMT), which is indicated by a primary-filling configuration or by a hotspot on the secondary, and solid dots refer to other systems. For systems belonging to groups (2) and (3), the period decrease may not be reasonably explained by mass transfer between the components. The primary components of the NCBs are usually A- and F-type, while the secondary ones are G- or K-type cool stars. The period decreasing may result from angular momentum loss (AML) via magnetic braking. It is shown from Fig. 11 that the dp/dt of NCBs with PSMT are usually larger than those of the other systems. This indicates that apart from AML, PSMT in these Figure 11. A possible correlation between dp/dt and P for NCBs with decreasing period. Diamonds refer to systems with PSMT and solid dots to the others. See text for detail.

8 1254 S. Qian systems also contributes to the period decrease. For overcontact binary systems, the studies of Qian (2001a,b) have shown that mass ratio is a key variable for orbital period change. The parameters listed in Table 5 indicate that there is no significant correlation between the mass ratio and the period change for NCBs. Two evolutionary paths to overcontact binaries were discussed by Hilditch et al. (1988). One was from detached system directly evolving into initially shallow overcontact and the other was via a case A mass transfer to semidetached, then to overcontact systems. The components of the NCBs listed in Table 5 are main-sequence stars. They may be the results of a Case A evolution in which a magnetic stellar wind has resulted in drastic AML. The secular period decreases of the NCBs indicates that the AML is continuous, and a overcontact configuration would seem inevitable. Thus the present sample stars may evolve into overcontact binaries via the second path of Hilditch (1988). As the orbital period is decreasing, the shrinking of the critical Roche lobe can cause the formation of a common convective envelope (CCE). Once the CCE is formed, these NCBs become to overcontact binary stars where both components share the CCE with its nearly uniform surface brightness resulting from energy transfer between the two stars and thus the systems display EW-type light curve. Therefore, these NCBs are at the beginning of the overcontact phase. They are very important source for understanding the formation of CCE and for studying the dynamical evolution of close binary star. As shown in Fig. 11, a possible statistical correlation between orbital period P and its rate of decrease dp/dt may exist. With the least-squares method, the equation dp/dt = 5.3(±1.4) 10 7 P + 1.3(±0.7) 10 7 (18) is obtained where P is in unit of d and dp/dt in 10 7 dyr 1. The equation tell us that the smaller P is, the lower dp/dt will be. The general trend of the correlation may indicate that once the NCBs are evolved into overcontact binary systems the periods will increase. Recently, the period changes of 22 A- and F-type overcontact binaries (M 1 > 1.35 M ) have been investigated by Qian (2001b). It is shown that, apart from the AW UMa with the shortest q(q 0.072), the periods of the others are increasing. The correlation found in this paper leads one to speculate that there may be a smooth transition from A- and F-type NCBs that have period decreases to the A- and F-type overcontact binary stars with period increases, and in that sense one may postulate that the NCBs may be the progenitors of the A-type W UMa systems. To explain the period variations of W UMa-type overcontact binary systems, an evolutionary scheme was proposed by Qian (2001a). The scheme assumed that the change of the depth of overcontact can cause the variation of magnetic activity that changes the AML rate (see also Vilhu 1981 and Smith 1984). This model predicted that cooler overcontact binaries (M 1 < 1.35 M ) will oscillate around a critical mass ratio because a period increase can cause a decrease of depth of overcontact and result in a rather higher AML rate, and finally the period will decrease again. For the hotter overcontact binary systems (M 1 > 1.35 M ), since the AML rate is lower, a period-increasing system is expected. Once overcontact is broken, an NCB will be formed and the system will oscillate around a marginal-contact state as predicted by TRO (Lucy 1976; Flannery 1976; Robertson & Eggleton 1977). ACKNOWLEDGMENTS This work was supported by the Chinese Natural Science Foundation (No ) and by the National Key Fundamental Research Project through grant G The author is greatly indebted to the anonymous referee for valuable comments and suggestions. REFERENCES Agerer F., Heubscher J., 1996, Inf. Bull. Variable Stars, 4383 Agerer F., Heubscher J., 1998a, Inf. Bull. Variable Stars, 4562 Agerer F., Heubscher J., 1998b, Inf. Bull. Variable Stars, 4606 Agerer F., Heubscher J., 1999a, Inf. Bull. Variable Stars, 4711 Agerer F., Heubscher J., 1999b, Inf. Bull. Variable Stars, 4712 Agerer F., Hubscher J., 2001a, Inf. Bull. Variable Stars, 5016 Agerer F., Hubscher J., 2001b, Inf. Bull. Variable Stars, 5017 Agerer F., Dahm M., Hubscher J., 2001, Inf. Bull. Variable Stars, 5017 Al-Naimiy H. M. K., Fleyeh H. A., 1989, Ap&SS, 151, 29 Baade D., Duerbeck H. W., Karimie M. T., 1979, Inf. Bull. Variable Stars, 1568 Baade D., Duerbeck H. W., Karimie M. T., Yamasaki A., 1983, Ap&SS, 93, 69 Cerruti M. A., 1994, Inf. Bull. Variable Stars, 4052 Cerruti M. A., 1996, Acta Astron., 46, 429 Duerbeek H. W., Karimie M. T., 1979, Inf. Bull. Variable Stars, 1617 Dugan R. S., Piercel N. L., Wood R. B., 1951, Princeton Contr., 25 Fang M. J., Zhai D. S., Cao Z. D., Zhang R. X. Zhang J. T., Zhang X. B., 1994, Inf. Bull. Variable Stars, 4085 Faulkner D., 1986, PASP, 98, 690 Flannery B. P., 1976, ApJ, 205, 208 Flin P., 1981, in Rudnicki K., Flin P., eds, Rocznik Astronomiczny Obserwatorium Krakowskiego Paňstwowe Wydawnictwo Naukowe, Kraków,p.90 Gessner H., Meinunger I., 1973, Veröffentlichungen der Sternwarte in Sonneberg, 7, 619 Geyer E., Kippenhahn R., Strohmeier W., 1955, Kleine Veräff. Remeis- Sternw. Bamberg, No. 11 Gimenez A., Costa V., 1979, Inf. Bull. Variable Stars, 1643 Götz W., Wenzel W., 1962, Mitt. Veränderl. Sterne, 702 Gu S. H., 1999, A&A, 346, 437 Gu S. H., Liu Q. Y., Yang Y. L., Zhang Z. S., 1994, Inf. Bull. Variable Stars, 4076 Hanzl D., 1990, Inf. Bull. Variable Stars, 3423 Hanzl D., 1991, Inf. Bull. Variable Stars, 3615 Hilditch R. W., King D. J., 1986, MNRAS, 223, 581 Hilditch R. W., King D. J., 1988, MNRAS, 231, 397 Hilditch R. W., King D. J., McFarlane T. M., 1988, MNRAS, 231, 341 Hoffmeister C., 1933, Astron. Nachr., 247, 281 Hoffmeister C., 1934, Astron. Nachr., 253, 195 Hoffmeister C., 1943, Astron. Nachr., 274, 39 Hoffmeister C., 1964, Mitteilungen über Veränderliche Sterne Sonneberg, 2, 99 Huth H., 1957, Sonn. Veröff., 4, 149 Kaluzny J., 1985, Acta Astron, 35, 327 Kholopov P. N., 1985, Fourth edition of the General Catalogue of Variable Stars. Nauka, Moscow Kljakotko M. A., 1951, Perem. Zvezdy Prilozh., 8, 75 Lause F., 1938, Astron. Nachr., 266, 237 Lázaro C., Niarchos P., Rovithis P. E., Rovithis-Livaniou, Arévalo M. J., Antonopoulou E., 1995, AJ, 110, 1796 Leung J. C., Wilson R. E., 1977, ApJ, 211, 853 Lipari S. L., Sisteró R. F., 1986, MNRAS, 220, 883 Lipari S. L., Sisteró R. F., 1987, AJ, 94, 792 Lucy L. B., 1976, ApJ, 205, 208 Mardirossian F., Giuricin G., 1981, A&A, 96, 415 Marton S. F., Grieco A., Lapasset E., Sistero R. F., Claria J. J., Cerruti M. A., 1990, A&A, 237, 73 Milano L., Barone F., Mancuso S., Russo G., Vittone A. A., 1989, A&A, 210, 181 Moschner W., Frank P., Bastian U., 1999, Inf. Bull. Variable Stars, 4669

9 Mullis C. R., Faulkner D. F., 1991, Inf. Bull. Variable Stars, 3593 Nakamura T., Tanabe H., 1953, Ann. Tokyo Astron. Obs. 2nd Ser., 3, 175 Nakamura Y., Yamasaki A., Kitamura M., 1984, PASJ, 36, 277 Nelson R. H., 2001, Inf. Bull. Variable Stars, 5040 Ogloza W., Drozdz M., Zola S., 2000, Inf. Bull. Variable Stars, 4877 Paschke A., 1996, BBSAG Bull., 112 Pickering E. C., 1907, Harvard Coll. Obs. Circ., 135 Qian S. B., 2000, Ap&SS, 274, 711 Qian S. B., 2001a, MNRAS, 328, 635 Qian S. B., 2001b, MNRAS, 328, 914 Qian S. B., 2002, A&A, 387, 903 Qian S. B., Zhu L. Y., 2002a, ApJS, 142, 139 Qian S. B., Zhu L. Y., 2002b, ApJ, 568, 1004 Qian S. B., Liu Q. Y., Yang Y. L., 2000, Chinese Astron. Astrophys., 24, 331 Rafert J. B., 1990, AJ, 100, 1253 Rainger P. P., Bell S. A., Hilditch R. W., 1990, MNRAS, 246, 47 Robertson J. R., Eggleton P. P., 1977, MNRAS, 179, 359 Russo G., Sollazzo C., 1981, Ap&SS, 78, 141 Samec R. G., McDermith R. J., Carrigan B. J., 1995, Inf. Bull. Variable Stars, 4186 Samec R. G., Laird H., Mutzke M., Faulkner D. R., 1998a, Inf. Bull. Variable Stars, 4616 Samec R. G., Carrigan B. J., Gray J. D., French J. A., McDermith R. J., Padgen E. E., 1998b, AJ, 116, 895 Samec R. G., Gothard N. W., Parker B. C., Savage T. W., Anderson S. D., Faulkner D. R., 1999, Inf. Bull. Variable Stars, 4753 Short-period near-contact binary systems 1255 Sarma M. B. K., Sanwal N. B., 1981, ApSS, 74, 41 Smith R. C., 1984, QJRAS, 25, 405 Strohmeier W., Knigge R., Ott H., 1963a, Veröff. Remeis-Sternw. Bamberg, 5, 16 Strohmeier W., Knigge R., Ott H., 1963b, Veröff. Remeis-Sternw. Bamberg, 5, 17 Terrell D., Gunn J. B., Kaiser D. H., 1994, PASP, 106, 149 Terrell D., Wilson R. E., 1990, PASP, 102, 646 Van Cauteren P., Wils P., 2000, Inf. Bull. Variable Stars, 4872 Van Hamme W., Samec R. G., Gothard N. W., Wilson R. E., Faulkner D. R., Branly R. M., 2001, AJ, 122, 3436 Vetesnik M., 1967, Bull. Astron. Inst. Czech., 18, 208 Vilhu O., 1981, Ap&SS, 78, 401 Wilson R. E., 1966, AJ, 71, 32 Wolf G. W., Kern J. T., Hayes T. L., Chaffin C. R., 1982, Inf. Bull. Variable Stars, 2185 Wunder E., Wieck M., Kiling B., Gulmen O., Tunca Z., Evren S., 1992, Inf. Bull. Variable Stars, 3760 Yamasaki A., Okazaki A., 1984, PASJ, 36, 175 Yamasaki A., Okazaki A., 1986, PASP, 98, 1325 Yamasaki A., Okazaki A., Kitamura M., 1983, PASJ, 35, 131 Yamasaki A., Okazaki A., Kitamura M., 1984, PASJ, 36, 175 Young W. K., Ho-il K., Woo-Baik L., Kyu-Dong O., 2001, MNRAS, 325, 707 Zinner E., 1913, Astron. Nachr., 195, 454 This paper has been typeset from a TEX/LATEX file prepared by the author.