What’s Their Age Again? A Blue-Straggler Merger Scenario in the γ \gamma Persei Binary System Dóra Tarczay-Nehéz, László Molnár, Rozália Ádám Abstract We used MIST isochrone fitting and a dedicated grid of stellar evolution models computed with MESA to constrain the ages of the components of the γ \gamma Persei binary system. While individual stars can be matched to the models at specific metallicities, no joint isochrone solution reproduces both the observed masses and evolutionary states. The stellar evolutionary tracks calculated by MESA reveal a clear evolutionary mismatch. The primary component of the system is in a post–main-sequence phase consistent with the red giant branch or red clump. In contrast, the lighter secondary component lies near the turn-off point of the main-sequence or is in the early phase of the subgiant branch. This discrepancy can be overcome by assuming that the γ \gamma Persei system was born as a triple and the primary component is a rejuvenated star formed through a merger of a close-by pair of main-sequence stars. We show that the merger must have occurred no later than a few hundred Myr after system formation, and the progenitor masses of the merging stars are restricted by a combination of stars that fall within a narrow band in the ( M 1 , a , M 1 , b ) (M_{1,a},M_{1,b}) plane, corresponding to M 1 , a ≃ 0.9 M_{1,a}\simeq 0.9 – 2.1 ​ M ⊙ 2.1\,M_{\odot} and M 1 , b ≃ 2.3 M_{1,b}\simeq 2.3 – 2.5 ​ M ⊙ 2.5\,M_{\odot} . 1 Introduction \gamma Persei is a known, long-periodic spectroscopic binary system, with an orbital cycle of \sim 14.6 years. Its binary nature was first mentioned by 29 ; 6 ; 7 , and the orbital period was first determined by 31 from radial velocity measurements. Although the system is bright enough to be visible to the naked eye, the first total eclipse was observed only in 1990 by 19 . The system has since been the focus of diverse research using several observational methods, which include both visual ( 20 ) and spectroscopic observations ( 32 ; 41 ; 42 ; 40 ; 38 ; 10 ) , revealing a slight metallicity deficit, compared to the Sun ([Fe/H] \simeq-0.19-(-0.20) in the literature). Moreover, 40 classified both components as luminosity class III giants, based on detailed spectroscopic data. Orbital elements of the system were calculated by 41 , confirming that the system has a highly eccentric orbit ( e=0.785 ). Based on its parallax ( 14.1252" ), the distance of \gamma Persei is measured to be \sim 70.8 pc ( 14 ) . The system is a member of the \zeta Aurigae class ( 2 ) . This class consists of double-lined binaries with a cool G–K giant primary and a hotter A-type secondary. In the case of \gamma Persei, the primary component is a luminous G-type giant with a Johnson V-band magnitude of 2.91 ( 15 ) , dominating the system’s brightness. The secondary star is significantly fainter, with an estimated V magnitude of 4.54 . The system’s total brightness is V=2.93 and B=3.63 mag, with eclipse depths of 0.28 mag in V and 0.54 mag in B ( 19 ) . These values yield to the individual color-indices of (B-V)=+0.96 mag for the primary and +0.11 mag for the companion component, consistent with a G-type and A-type classification, respectively. Following the first documented 1990 primary eclipse, the next event in 2019 was reported by 10 , enabling a precise refinement of the orbital period. The previous eclipse in 2005 occurred close to the solar conjunction, thus impossible to observe from Earth. It was later recovered and published in a recent work of 1 , based on archival data from the Solar Mass Ejection Imager (SMEI), thereby filling the observational gap between the two known events and further strengthening the orbital period determination. The 2019 eclipse was observed using DSLR CMOS photometry, which was calibrated to the 1990 Johnson B-band light curve using spectrophotometric reference data ( 10 ) . This approach allowed the DSLR measurements to be transformed onto a standard photometric scale, enabling direct comparison with the earlier eclipse and precise timing analysis. The calibration was sufficient to refine the orbital period to 5329.0\pm 0.05 days, in agreement with the previous spectroscopic data. Additional partial eclipse phases were sampled by Telescopio Internacional de Guanajuato Robótico-Espectroscópico (TIGRE) spectra and Transiting Exoplanet Survey Satellite (TESS) photometry, the latter resolving the first contact phase and totality with high cadence, ending right around fourth contact. From the eclipse geometry and projected orbital velocity ( \sim 40 km/s), the radius of the primary star was derived to be 22.7\,R_{\odot} , while the radius of the secondary component was calculated to be 3.9\,R_{\odot} . TESS observed the star later again, most notably for two consecutive sectors in late 2024, making a global asteroseismic analysis possible. A seismic mass of 3.4\pm 0.4\,M_{\odot} was determined for the primary component based on the detected solar-like oscillations, which we present in a companion paper (Ádám et al., in prep.). While \gamma Persei is a well-characterised binary system and is at the focus of extensive investigations, recent studies have drawn attention to a puzzling evolutionary mismatch between its components. As 18 pointed out, the less massive A-type companion appears to be more evolved than the main G-type component. This contradicts the expectations for a coeval binary system. To overcome this discrepancy, several competing interpretations have been raised, including early-stage merger, mass loss, dynamical capture, and episodic mass transfer during the giant’s tip-RGB phase ( 10 , see) . In this study, we examine whether a formation pathway exists in which both stars could have formed simultaneously and evolved jointly without significant disturbance, consistent with the present-day configuration of \gamma Persei. In the following, we describe the applied methods used to map the parameter space of potential co-evolutionary configurations, focusing on metallicity ( Z ), initial mass ( M ), mixing length parameter ( \alpha_{\mathrm{MLT}} ), convective overshoot, and rotational velocity ( v_{\mathrm{rot}} ), using both MIST and MESA evolutionary models (Section 2 ). Section 3 presents our results, Section 4 provides a discussion on the results, and the paper closes with a conclusion for the system’s formation history in Section 5 . Table 1: Physical Parameters of the \gamma Persei ( \gamma Per) Binary System Parameter Primary Secondary T_{\mathrm{eff}}^{1} 4970\pm 70 K 8400\pm 70 K (\log L/L_{\odot})^{1} 2.45\pm 0.06 1.83\pm 0.06 M [ M_{\odot} ] 1,2,3,4 3.5\pm 0.3 2.4\pm 0.2 R [ R_{\odot} ] 1 22.7\pm 1.14 3.9\pm 0.2 [Fe/H] 3,5 -0.19\pm 0.08 dex -0.2\pm 0.08 dex Period [days] 1 5329.0\pm 0.05 1 1 footnotemark: 1 10 , 2 2 footnotemark: 2 19 , 3 3 footnotemark: 3 1 , 4 4 footnotemark: 4 Ádám et al. (in prep.), 5 5 footnotemark: 5 43 Table 2: Summary of the input parameters of MESA simulations. Parameter Value / Range Notes Count Stellar composition and the initial parameters for the primary component Initial mass ( M_{*} ) 3.2-3.8M_{\odot} \Delta M=0.1\,M_{\odot} 7 values Initial metallicity ([Fe/H]) -0.29, -0.19, -0.09 3 values Initial helium fraction ( Y_{\text{init}} ) 0.256 Fixed Initial hydrogen fraction ( X_{\text{init}} ) 1-Z-Y Computed dynamically Stopping condition TACHeB phase Stellar composition and the initial parameters for the secondary component Initial mass ( M_{*} ) 2.2-2.6M_{\odot} \Delta M=0.1\,M_{\odot} 5 values Initial metallicity ([Fe/H]) -0.30, -0.20, -0.10 3 values Initial helium fraction ( Y_{\text{init}} ) 0.256 Fixed Initial hydrogen fraction ( X_{\text{init}} ) 1-Z-Y Computed dynamically Stopping condition tip of the Red Giant Branch Mixing and convection physics Mixing length parameter ( \alpha_{\text{MLT}} ) 1.5-2.5 \Delta\alpha_{\mathrm{MLT}}=0.2 6 values Semi-convection parameter ( \alpha_{s} ) 0.1 Convective boundary criterion Ledoux Convective overshoot scheme Exponential Convective overshoot parameters ( f , f_{0} ) See Table 3 Varies per set 3 values Opacity sources and nuclear network High- T opacity source OPAL Low- T opacity source Ferguson et al. (2005) opacity prefix ( lowT_fa05 ) Heavy element mixture Asplund et al. (2009) Solar-like abundance (’A09’) Carbon/Oxygen opacities Enabled Type2 opacities Enabled Nuclear network pp_and_cno_extras_o18_ne22 Tracks 19 isotopes Total Models 648 Table 3: Overshoot coefficients ( f , f_{0} ) adopted in this work for convective core and neighbouring radiative zones. Region f f_{0} Set A — Low Overshoot Core (H-burning) 0.012 0.002 Radiative shell 0.022 0.002 Set B — Moderate Overshoot Core (H-burning) 0.020 0.004 Radiative shell 0.030 0.004 Set C — High Overshoot Core (H-burning) 0.030 0.005 Radiative shell 0.040 0.005 2 Methods 2.1 MIST- Based Isochrone Model Setup To validate co-evolution of the binary system of the \gamma Persei, the isochrone package of the MESA Isochrones & Stellar Tracks ( MIST ) database ( 11 ; 8 ) was utilized. The package offers a comprehensive grid for stellar evolutionary modelling, providing isochrones and synthetic photometry across a broad range of physical parameters. The full metallicity ([Fe/H]) range supported by MIST spans from -4.0 to +0.5 dex. In our analysis, we varied metallicity with [Fe/H] values from -0.2 and -0.19 down to -2.0 dex to probe lower-metallicity regimes as well. Synthetic magnitudes were extracted for the Gaia photometric system, with extinction ( A_{\mathrm{v}} ) fixed at 0.0 for all models, although the database covers a range of A_{\mathrm{v}} between 0 and 6 . Isochrones for both rotational configurations available in the grid were investigated: v/v_{\mathrm{(crit)}}=0.0 , representing non-rotating stars, and v/v_{\mathrm{(crit)}}=0.4 , corresponding to moderately rotating stars. Here, v_{\mathrm{(crit)}} denotes the critical rotation velocity at which centrifugal acceleration at the stellar equator balances gravitational acceleration. Exceeding this threshold, the star would be disrupted by centrifugal forces. For the zeroth-order constraint, we adopted the standard MIST age grid, which consists of 107 discrete age points uniformly spaced in \log({\rm Age}\,[yr]) from 5.0 to 10.3 dex in 0.05 dex increments. Upon identifying the subset of isochrones that provided satisfactory fits to both stellar components, we refined the age resolution locally by resampling the grid in steps of 0.01 dex, allowing for more precise age determination for the system. 2.2 MESA Stellar Evolution Tracks In addition to the parameter space offered by the MIST grid (e.g., rotation and metallicity), we performed a set of stellar evolution calculations using the stable version of 23.05.1 of the open-source ’Modules for Experiments in Stellar Astrophysics’ ( MESA ) software package ( 33 ; 34 ; 35 ; 36 ; 37 ; 25 ) . All evolutionary tracks are available under the Zenodo doi: 10.5281/zenodo.17229941 . 2.2.1 Metallicity, Heavy Element Abundance and Stellar Mass Grid For each component of the binary system, we constructed a mass grid centered at 3.5\pm 0.3\,M_{\odot} and 2.4\pm 0.2\,M_{\odot} , respectively, with a resolution of 0.1\,M_{\odot} ( 19 ; 10 ; 1 , Ádám et al., in prep.) . This results in 7 and 5 mass points per star. The primary component was evolved up to the Terminal Age Core Helium Burning ( TACHeB , stop_at_phase_TACHeB = .true. ) phase, while the secondary was evolved to the tip of the Red Giant Branch (RGB, stop_at_phase_He_Burn = .true. ). We adopted a solar composition for both stars and varied the metallicity separately for each star: [Fe/H] = -0.19 for the primary and [Fe/H] = -0.20 for the secondary, each with a local grid of \pm 0.1 dex in 0.1 dex steps, yielding 3 metallicity values per component ( 43 ; 1 ) . These values correspond to Z\simeq 0.0089 and Z\simeq 0.0084 , respectively, assuming the relation of Z=Z_{\odot}\cdot 10^{[\rm Fe/H]}, (1) and Z_{\odot}=0.0134 ( 4 ) . The parameter Z_base was dynamically maintained equal to the initial metallicity, Z_init . 2.2.2 Opacity Tables In this work, the stellar energy transport efficiency was characterized using the built-in MESA opacity tables. For high-temperature conditions, the Opacity Project at Livermore (OPAL) data 23 ; 24 was employed, specified by the kap_file_prefix = ’a09’ parameter. Conversely, the low-temperature regime was parametrized by the inclusion of opacity data from 13 , by using kap_lowT_prefix = ’lowT_fa05_a09p’ . We used specialized opacity tables — namely those enriched in carbon and oxygen ( kap_CO_prefix = ’a09_co’ ) and the Type 2 opacities ( use_Type2_opacities = .true. ) It should be emphasized, however, that their overall influence on the evolutionary trajectories of the stars is negligible. 2.2.3 Convection Core Overshoot and the Mixing Length Parameter Convection was modeled across all computations using the Mixing Length Theory (MLT) coupled with the time-dependent convection ( MLT_option = ’TDC’ ). Following the approach of e.g., 51 ; 50 and 26 , the Ledoux criterion ( use_Ledoux_criterion = .true. ) was employed to define convective boundaries, utilizing a semi-convection parameter ( \alpha_{s} ) of 0.1 ( alpha_semiconvection = 0.1 ). We systematically investigated the MLT parameter ( \alpha_{\mathrm{MLT}} ), varying its value from 1.5 to 2.5 with a resolution of 0.2 (six distinct values). For each \alpha_{\mathrm{MLT}} setting, 3 distinct overshoot prescriptions were tested using an exponential overshoot scheme ( overshoot_scheme = ’exponential’ ). This scheme was applied below the hydrogen-burning shell ( overshoot_zone_type(1) = ’burn_H’ ) and above the ”non-burning” regions ( overshoot_zone_type(2) = ’none’ ). Following the most recent work of 52 and 47 , we used the three distinct overshoot prescriptions for all models. The applied input parameters are summarized in Tables 2 and 3 . 2.2.4 Nuclear Reaction Network For the analysis of intermediate-mass stars with initial masses between 2 and 4\,M_{\odot} , we adopted the extended nuclear network pp_and_cno_extras_o18_ne22 . Reaction rates in this network are taken from the Nuclear Astrophysics Compilation of Reaction Rates ( 3 , NACRE; ) and the Joint Institute for Nuclear Astrophysics REACtion LIBary (JINA REACLIB) database ( 9 ) . The network follows 19 isotopes involved in the proton–proton chain, the carbon–nitrogen–oxygen cycles (including hot CNO branches), and additional burning channels involving heavier nuclei. The tracked isotopes are: 1 H, 2 H, 3 He, 4 He, 7 Li, 7 Be, 8 B, 12 C, 14 N, 14 O, 16 O, 18 O, 19 F, 18 Ne, 19 Ne, 20 Ne, 22 Ne, 22 Mg, and 24 Mg. 2.2.5 Spatial and Temporal Resolution The spatial and temporal resolution of the models can be adjusted by modifying MESA ’s default refinement parameters, namely mesh_delta_coeff and time_delta_coeff ( 28 , see, e.g.,) . These coefficients scale the internal adaptive mesh refinement (AMR) scheme used by MESA . In the present work, we retained the default resolution settings, with mesh_delta_coeff = 1.0 and time_delta_coeff = 1.0 . For improved numerical stability and energy conservation, the stellar energy equation was solved using the dedt formulation ( energy_eqn_option = ’dedt’ ). In addition, the make_gradr_sticky_in_solver_iters option was activated to enhance convergence at the interfaces between convective and radiative regions ( 36 ; 37 , see, ) . 2.3 Matching Ages for Different Evolutionary Tracks In order to identify which combinations of model configurations yield a consistent age for both stellar components, we used a dedicated Python ( 49 ) script ( 30 ; 48 ; 22 , based on pandas, numpy, and matplotlib;) , to compare evolutionary tracks and extract age-matching solutions. Table 4: Age and mass ranges where the two components of the system intersect MIST isochrones (within the observational errors), grouped by metallicity. Ages are given in Myr. [Fe/H] Primary: t ranges [Myr] Primary: M ranges [ M_{\odot} ] Secondary: t ranges [Myr] Secondary: M ranges [ M_{\odot} ] -4.00 8 900–20 000 0.696–0.882 1 410–1 580, 5 010 1.471–1.527, 1.025 -3.50 7 080–20 000 0.699–0.943 1 260–1 410, 4 470–5 010 1.524–1.580, 1.021–1.062 -3.00 – – – – -2.50 – – – – -2.00 1 000–20 000 0.652–1.707 794–891 1.745–1.812 -1.50 724–2 820 1.235–1.958 708–871 1.799–1.932 -1.00 282–794 2.080–2.909 708–794 1.894–1.983 -0.20 178–355 3.165–3.904 631–759 2.170–2.338 -0.19 178–355 3.171–3.909 631–708 2.247–2.343 -0.10 – – 631–759 2.216–2.370 The analysis proceeded through the following steps: • First, it identifies where each stellar component intersects the calculated evolutionary track beyond the pre-main-squence phase within the observational uncertainties listed in Table 1 . • In cases where no track point falls explicitly within the observational error box, but the track segment crosses it, the script estimates the intersection age via linear interpolation between the segment endpoints. This allows us to identify age-matching solutions even when no direct data point lies within the uncertainty region. • Finally, the script compares the resulting age distributions for both stars and identifies model configuration pairs that yield the smallest age difference. These are considered the best candidates for a physically consistent binary solution. As an additional validation step, we applied a threshold of 0.1 dex in \log_{10} (age) to identify model configuration pairs with sufficiently consistent ages. This can be calculated as follows: \Delta_{\log}=\log_{10}\left(\frac{t_{2}}{t_{1}}\right). (2) This ensures that the age estimates derived from two stellar evolution models for the components of the \gamma Persei system differ by no more than 0.1 dex (corresponding to \sim 25%), thereby allowing us to retain only those configurations in which both stars can be considered coeval within the adopted tolerance. 2.3.1 The Goodness-of-Fit Metric To quantify how well individual MESA evolutionary models reproduce the observed properties, we define a scalar goodness-of-fit metric: \varepsilon . This measures the distance between the model predictions and the observational constraints in the three-dimensional parameter space of the effective temperature, luminosity, and radius. This metric is computed only for those model points that were identified as matching with the observations according to the procedure described in Section 2.3 . \varepsilon reads as \varepsilon=\sqrt{\left(\frac{\Delta\log T_{\mathrm{eff}}}{\sigma_{\log T}}\right)^{2}+\left(\frac{\Delta\log L}{\sigma_{\log L}}\right)^{2}+\left(\frac{\Delta R}{\sigma_{R}}\right)^{2}}, (3) where \Delta\log T_{\mathrm{eff}} , \Delta\log L , and \Delta R are the differences between the calculated model parameters and the observed quantities, and \sigma_{\log T} , \sigma_{\log L} , and \sigma_{R} are the corresponding observational uncertainties. 3 Results In this section, we present our results on matching the ages of the two components of the \gamma Persei binary system. 3.1 MIST Isochrone Fitting As a first approach, we examined the stellar parameters of the two stars in the system using MIST isochrones. The results of this analysis is listed in Table 4 . While the secondary component of the system can be consistently matched to MIST isochrones across the full range of metallicities explored, the primary component cannot be fitted in all cases. In particular, for metallicities of \mathrm{[Fe/H]}=-0.1 , -2.5 , and -3.0 , the primary component of the \gamma Persei system cannot be fitted on the given isochrones from the MIST database. Furthermore, a joint fit for both components is only achievable for metallicities \mathrm{[Fe/H]}=-1.0 and \mathrm{[Fe/H]}=-1.5 . In the case of \mathrm{[Fe/H]}=-1.5 , both stars can be placed on a single isochrone with ages in the range 728-871 Myr. In this case, the inferred masses are 1.96 – 1.84\,M_{\odot} for \gamma Persei 1 and 1.93 – 1.80\,M_{\odot} for \gamma Persei 2. Similarly, for \mathrm{[Fe/H]}=-1.0 , a common isochrone solution is found at ages 708-794 Myr, yielding masses of 2.17 – 2.08\,M_{\odot} for the primary and 1.98 – 1.89\,M_{\odot} for the secondary. However, in both cases the inferred stellar masses and metallicities differ substantially from the values reported in the literature (see Table 1 ). Consequently, although formal isochrone solutions exist for certain metallicities, they are astrophysically implausible. 3.2 MESA Evolution Track Fitting As the MIST isochrone fitting did not yield any configuration for the system that is consistent with the observed masses and metallicities, we constructed a grid of stellar models with MESA , adopting the parameters described in Section 2.2 (see Table 2 ). Table 7 lists the results for this analysis. In general, independent of the adopted physical parameters, we consistently find that the age of the primary component lies in the range of approximately 200-320 Myr. In contrast, the secondary component is significantly older with solutions spanning \sim 32-1000 Myr. Considering the model pair that minimizes the age difference between the two components, we find ages of t_{1}=390 Myr for the primary and t_{2}=547 Myr for the secondary component. This corresponds to an absolute age difference of 156 Myr. Using the definition introduced in Eq. ( 2 ), this yields a logarithmic age difference of \Delta_{\log}\simeq 0.15\penalty\ \mathrm{dex}, (4) which corresponds to a relative age difference of \sim 40 %. Therefore, even the closest age pairing identified in our model grid exceeds the adopted tolerance of 0.1 dex, and cannot be considered coeval under our selection criterion. Figure 1: Evolutionary tracks for the \gamma Persei primary component, colour–coded by stellar age. The three subpanels illustrate representative models that place the star in distinct post-main-sequence evolutionary phases. From left to right: the end of the subgiant branch; the red–clump phase; either the upper red–giant branch or the early asymptotic–giant branch. The position of the primary component is shown in cyan color with observational uncertainties. Figure 2: Evolutionary tracks for the \gamma Persei secondary component, colour–coded by stellar age. The three subpanels illustrate representative models that place the star in distinct evolutionary stages. From left to right: subgiant phase; a model located in the vicinity of the main-sequence turn-off point; a main-sequence phase. The observed position of the secondary component is shown with its measurement uncertainties (green triangle), and the main-sequence turn-off point is marked for each model with a red circle. 3.3 Evolutionary State of the Two Components The locations of the two components of the \gamma Persei system were examined in detail using the MESA evolutionary tracks described in Section 2.2 . Representative evolutionary tracks are shown in Figures 1 and 2 , with the observationally determined positions of the two stars overplotted. Appendices A and B contain the full set of evolutionary tracks for the whole explored parameter-space (see Figures 6 – 8 and 9 – 11 ). For the primary, we consistently find that its position in the Hertzsprung–Russell diagram (HRD) is near or beyond the red giant branch. Depending on the adopted model parameters, the star is placed onto the HRD either: (i) near the luminosity dip at the end of the subgiant branch, (ii) in the red clump phase, corresponding to stable core helium-burning, or (iii) on the red giant branch or in the early phase of the asymptotic giant branch (see Figure 1 ). Their position on the HRD shows only a mild dependence on the adopted input parameters (see Figures . In general, lower metallicity or larger overshoot values tend to shift the primary component toward the red clump, while higher metallicity or smaller overshoot favour a location near the subgiant-branch dip. However, across all combinations of \alpha_{\rm MLT} , [Fe/H], and overshoot, the primary remains in a post–main-sequence phase (see Figures 6 – 8 ). Considering the secondary component, the dependence on the input parameters is similarly modest. However, we found that only models within the 2.2\text{--}2.4\,M_{\odot} range align with the position of the star (see, e.g., Figure 9 ). Higher-mass tracks (e.g. 2.6\,M_{\odot} ) do not fit the measurements of the star for any combination of \alpha_{\rm MLT} , [Fe/H], or overshoot (see Figures 9 – 11 ). This yields a more precise mass determination for this component of the system, effectively tightening the observational limits for the given set of model assumptions. Nevertheless, all three solutions suggest that the primary component of the system is a post-main-sequence star and must be at least in the red giant branch phase, independently of whether it is currently undergoing core helium-burning (red clump) or is located slightly before or after this phase. In contrast, the secondary component of the \gamma Persei system is in a remarkably different evolutionary stage. The results of the MESA evolutionary tracks indicate that the position of this component is either close to the turn-off point of the main-sequence or on the very early phase of the subgiant branch (see Figure 2 ). The combination of these evolutionary states implies a substantial discrepancy between the ages inferred for the two components under the assumption of single-star evolution. While the secondary component requires an age sufficient to reach helium ignition, the primary component remains near the main-sequence turn-off or early subgiant phase. This evolutionary mismatch persists across the explored range of stellar parameters. 3.4 The Most Plausible Model Configurations Based on the \varepsilon Metric We calculated the most plausible models for each component by utilizing the goodness-of-fit metric \varepsilon in the \log T_{\mathrm{eff}}-\log L-R phase defined in Equation 3 . For the primary component, the best-fitting models yield \varepsilon\simeq 0.09 – 0.095 . All of these solutions correspond to the red clump phase (see Figure 1 ), with masses M\simeq 3.4 – 3.5\,M_{\odot} and metallicities at [Fe/H] =-0.29 , placing the star in the core helium–burning phase. The most relevant configurations are summarized in Table 5 . Figure 3: The four best-fitting models for the primary component of the \gamma Persei system. Each model corresponds to the red-clump phase of stellar evolution. The yellow point corresponds to the best-fitting model, while the purple point (with error bars) presents the place of the primary component. Table 5: Best-fitting MESA models for the primary component, ranked by the goodness-of-fit metric \varepsilon . All configurations correspond to red clump (core He-burning) models. M [Fe/H] \alpha_{\mathrm{MLT}} t R \log T_{\mathrm{eff}} \varepsilon [ M_{\odot} ] [Myr] [ R_{\odot} ] [K] Low overshoot models 3.4 -0.29 1.7 279 22.617 3.6960 0.090 3.5 -0.29 1.7 246 22.625 3.6968 0.092 3.5 -0.29 1.7 246 22.582 3.6967 0.095 3.5 -0.29 1.7 247 22.698 3.6958 0.095 Figure 4: The four best-fitting models for the secondary component of the \gamma Persei system. The panels correspond to different evolutionary stages: the two left-hand panels show models near the main-sequence turn-off (the lower-left panel is still on the late main-sequence), while the upper-right panel represents a subgiant model. The yellow point corresponds to the best-fitting model, while the purple point (with errorbars) presents the place of the primary component. Table 6: Best-fitting MESA models for the secondary component, ranked by the goodness-of-fit metric \varepsilon . The models cover late main-sequence, turn-off, and early post–main-sequence phases. M [Fe/H] \alpha_{\mathrm{MLT}} t R \log T_{\mathrm{eff}} \varepsilon [ M_{\odot} ] [Myr] [ R_{\odot} ] [K] Medium overshoot models 2.3 -0.30 1.7 656.3 3.893 3.9245 0.088 2.2 -0.30 2.5 707.7 3.900 3.9240 0.097 2.3 -0.30 1.7 658.8 3.912 3.9242 0.125 2.3 -0.30 1.5 660.9 3.914 3.9243 0.137 Considering the secondary component of the \gamma Persei system, the best-fitting models cover a range of \varepsilon\simeq 0.088-0.137 . However, the evolutionary phase cannot be uniquely determined by the \varepsilon metric, as the best-fitting solutions cover late main-sequence, turn-off, and subgiant configurations (see Figure 4 ). Table 6 collects the calculated stellar parameters for the four best-fitting models for this component. 4 Discussion The discrepancy between the ages of the two stars suggests that the primary component is a rejuvenated merger product, which passed through the evolutionary phase commonly referred to as a blue straggler ( 46 ; 45 , see, e.g., the recent review of) . Blue stragglers are stars that appear younger than their neighboring stars in clusters ( 21 ; 39 ; 5 ; 12 , see, e.g.,) . Thus, their calculated age is an “apparent age” and originates from the replenishment of hydrogen in its core during the merger and their new location on the HRD. 44 pointed out that the rejuvenation can be significant depending on the masses and the evolutionary stages of the merging components. They showed that the rejuvenation can cover a range from at least a few per cent up to 80% and more of the main-sequence lifetime of stars. This scenario may apply to the primary component of the \gamma Persei system. While the primary component holds a rejuvenated age, the secondary component reflects the true evolutionary age of the system and acts as an anchor to determine the age of the system. Following this idea, based on the results of the stellar evolution models, we define: • True age of the system ( T_{\rm true} ): inferred from the secondary component, yielding \sim 750 – 900\, Myr. • Apparent age of the merger product ( t_{\rm app} ): inferred from the primary component, yielding \sim 280 – 350\, Myr. We quantify the rejuvenation using an effective rejuvenation fraction, R_{\rm eff}=1-\frac{t_{\rm app}}{T_{\rm true}}, (5) which directly measures the fractional reduction in apparent age relative to the true system age. This quantity is an analogue of the rejuvenation defined in terms of fractional main-sequence ages by 16 and 44 . 4.1 Minimum Rejuvenation Requirements The minimum rejuvenation corresponds to the limiting case in which the merger occurred very recently, such that the present-day apparent age closely reflects the post-merger apparent age. This represents a conservative lower limit, as it neglects the short post-merger thermal relaxation phase. The latter can be done as the thermal relaxation is expected to occur on the Kelvin–Helmholtz timescale, t_{\rm KH}\simeq\frac{GM^{2}}{RL}. (6) For the primary ( M\simeq 3.5\,M_{\odot} , R\simeq 22.7\,R_{\odot} , L\simeq 10^{2.45}\,L_{\odot} ), this yields t_{\rm KH}\sim 5\times 10^{4}\text{--}10^{5}\,\mathrm{yr}, (7) which is negligible compared to the inferred system age and the main-sequence evolutionary timescale. The rejuvenation factor ( R ) can be constrained between the lower and upper bounds inferred from T_{\rm true} and t_{\rm app} : • Lower bound: T_{\rm true}=750\,\mathrm{Myr} and t_{\rm app}=350\,\mathrm{Myr} , corresponding to the younger end of the calculated true-age range and the older end of the inferred apparent-age range, yielding R_{\mathrm{eff,low}}=1-\frac{350}{750}\approx 0.53. (8) • Upper bound: T_{\rm true}=900\,\mathrm{Myr} and t_{\rm app}=280\,\mathrm{Myr} , corresponding to the older end of the calculated true-age range and the younger end of the inferred apparent-age range, yielding R_{\mathrm{eff,up}}=1-\frac{280}{900}\approx 0.69. (9) Thus, a rejuvenation of at least 50 – 70\% is required to explain the observed properties of the primary component with the system age inferred from the secondary component. The calculated range of the rejuvenation factor is consistent with the results of e.g., 44 . 4.2 Timing of the Merger Event Assuming a representative rejuvenation efficiency of R=0.8 , which corresponds to the approximate upper limit expected from main-sequence (or MS+MS) mergers ( 44 ) , we can constrain a window for the epoch of the merger ( T_{\rm merg} ). The apparent age of the rejuvenated star after the merger is reduced by a factor R_{\rm eff} . Subsequently, after the thermal relaxation, the star evolves along the main-sequence as a “simple”, non-disturbed star. The apparent, present-day age of the merged star is given by t_{\rm app}=T_{\rm true}-R_{\rm eff}\,T_{\rm merg}, (10) where T_{\rm true} is the actual age of the system, T_{\rm merg} is the time of the merger, and t_{\rm app} is the apparent age calculated from observations. Solving for T_{\rm merg} yields T_{\rm merg}=\frac{T_{\rm true}-t_{\rm app}}{R_{\rm eff}}. (11) Using the calculated ranges for the system age and apparent age of the primary component (see Table 7 ), one can constrain a time window for the epoch of the merger as T_{\rm merg,min}=\frac{T_{\rm true,\ min}-t_{\rm app,\ max}}{R_{\rm eff}}\approx\frac{750-350}{0.8}\approx 500\,\mathrm{Myr}, (12) T_{\rm merg,\ max}=\frac{T_{\rm true,\ max}-t_{\rm app,\ min}}{R_{\rm eff}}\approx\frac{900-280}{0.8}\approx 775\,\mathrm{Myr}. (13) Thus, the merger must have occurred no later than \sim 500 – 775\,\mathrm{Myr} ago. Assuming that the system is 750-900 Myr old, the merger likely took place \sim 150-200 Myr after formation, while the progenitor stars were still on the main-sequence. 4.3 Constraining the Progenitor Masses for the Primary Component Assuming that the present-day primary component was produced by a MS+MS merger, we can constrain the masses of the progenitor merging stars. This can be obtained by using the current age of the system from the secondary component of the \gamma Persei system and the estimated lifetime of the merging main-sequence stars ( t_{\mathrm{MS}} ). The latter can be approximated as t_{\rm MS}\sim\frac{M}{L}, (14) where M is the stellar mass and L the luminosity ( 27 , e.g.,) . Assuming the standard mass-luminosity relation on the main-sequence: L\sim M^{3.5}, (15) the lifetime of the main-sequence becomes t_{\rm MS}\sim 10^{10}\,{\rm yr}\left(\frac{M}{M_{\odot}}\right)^{-2.5}. (16) Assuming that the merger occurred recently, but sufficiently long ago for the remnant to have completed its thermal relaxation, the age of the system can be used to constrain the maximum masses of the progenitor stars. Given an estimated system age of T_{\rm true}\simeq 750 – 900\,{\rm Myr} , the corresponding main-sequence turn-off point mass ( M_{\mathrm{TO}} ) can be calculated from Equation 16 , which yields: M_{\rm TO}\simeq 2.3\text{--}2.5\,M_{\odot}. (17) This provides an upper limit on the masses of the progenitor stars for the close binary. Any star exceeding M_{\rm TO} would have already left the main sequence by the time of coalescence, preventing a MS+MS merger at the present epoch. The present-day mass of the primary component is M_{1}=3.5\pm 0.3\,M_{\odot} . Assuming a modest mass loss during the merger ( \eta\simeq 0.9 – 0.95 ), the total mass of the progenitor binary is constrained by M_{1,a}+M_{1,b}=\frac{M_{1}}{\eta}\simeq 3.4\text{--}4.2\,M_{\odot}. (18) As both progenitors must remain below the turn-off mass ( M_{\rm TO}\simeq 2.3 – 2.5\,M_{\odot} ), the two masses cannot vary independently. For the lowest allowed total mass ( M_{1,a}+M_{1,b}=3.4\,M_{\odot} ), the extreme case corresponds to a point that lies near the turn-off point, while the other one reaches the absolute minimum that fulfills Equation 18 assuming that M_{\mathrm{1,b}}\geq M_{\mathrm{1,a}} : \begin{split}M_{1,a}^{\mathrm{lower}}\simeq 0.9\text{--}1.1\,M_{\odot}\\ M_{1,b}^{\mathrm{lower}}\simeq 2.3\text{--}2.5\,M_{\odot}.\end{split} (19) Assuming the highest allowed total mass for the system ( M_{1,a}+M_{1,b}=4.2\,M_{\odot} ), we can constrain the progenitor masses with the turn-off point from its conservative upper limit as M_{\rm TO}=2.5\,M_{\odot} constrained by M_{\mathrm{1,b}}\geq M_{\mathrm{1,a}} : \begin{split}M_{1,a}^{\mathrm{upper}}\simeq 1.7\text{--}2.1\,M_{\odot},\\ M_{1,b}^{\mathrm{upper}}\simeq 2.1\text{--}2.5\,M_{\odot}.\end{split} (20) Figure 5: Allowed combinations of the progenitor masses (M_{1,a},M_{1,b}) . The blue region shows the parameter space permitted by the total-mass constraint ( 3.4 – 4.2\,M_{\odot} ) and the requirement that both stars remain below the turn-off mass ( M_{\rm TO}\simeq 2.3-2.5\,M_{\odot} ). The light red shaded regime denotes the region where M_{1,b}>M_{1,a} , as assumed throughout the analysis. The red points indicate the extreme allowed configurations discussed in the text: the lower-mass limit ( M_{1,a}^{\rm lower}=0.9\,M_{\odot} , M_{1,b}^{\rm lower}=2.5\,M_{\odot} ), the upper-mass limit ( M_{1,a}^{\rm upper}=1.7\,M_{\odot} , M_{1,b}^{\rm upper}=2.5\,M_{\odot} ), and the equal-mass upper case ( M_{1,a}=M_{1,b}=2.1\,M_{\odot} ). Only mass pairs that lie within the diagonal band satisfy all physical constraints. The constraints in Equation 18 , together with the requirement that both progenitors remain below the turn-off mass, do not define two independent mass intervals for (M_{1,a},M_{1,b}) . Instead, the allowed solutions occupy a narrow diagonal band in the (M_{1,a},M_{1,b}) plane, bounded by the total-mass limits ( 3.4 – 4.2\,M_{\odot} ) and by the condition M_{1,b}\geq M_{1,a} . As illustrated in Fig. 5 , only mass pairs lying within this band are physically permitted. We note that the merger efficiency is not fixed but depends on the evolutionary state, mass ratio, and internal structure of the progenitor stars. Hydrodynamic simulations show that MS+MS collisions typically eject only a few per cent of the total mass, with higher mass loss occurring for more evolved or more equal-mass encounters ( 17 , see, e.g.,) . 5 Conclusions In this work, we investigated whether the present-day configuration of the \gamma Persei system can be explained under the assumption of coeval formation and standard single-star evolution, using the observational constraints reported in the literature ( 19 ; 40 ; 10 ; 1 ; 43 , e.g.,) and in the companion paper presenting the asteroseismic analysis of the primary (Ádám et al., in prep.). Isochrone fitting based on the MIST database demonstrates that, although formal coeval solutions exist for certain metallicities, these solutions require stellar masses and chemical compositions that are inconsistent with the directly measured system parameters. In particular, joint isochrone solutions systematically favor significantly lower masses and more metal-poor compositions than those that can be found in the literature. We therefore conclude that MIST -based coeval solutions are astrophysically implausible for the \gamma Persei system. To further investigate the coevality of the system, we constructed a grid of stellar evolution models using MESA . Across all the applied model parameters, we consistently found an evolutionary mismatch between the two components. The primary component is found to be in a post-main-sequence evolutionary phase, i.e., on the red giant branch, or most likely, in the red clump. In contrast, the secondary component lies close to the turn-off point of the main-sequence or at the very beginning of the subgiant branch. To overcome this discrepancy, we conclude that the primary component of the system is a rejuvenated (former blue straggler) object formed through a merger of two main-squence stars ( 16 ; 44 ; 46 , e.g.,) . In this scenario, the present-day, the secondary component reflects the true age of the system, while the merger product appears significantly younger due to its placement on the HRD and the replenishment of hydrogen in its core. This suggests that the \gamma Persei system formed as a triple-star system, where an inner, closer binary merged during the main-sequence phase, and the outer component evolved mainly undisturbed. Using the evolutionary state of the non-merged star as an age anchor, we calculated the true age of the system to be \sim 750 – 900 Myr. The merger is constrained to have occurred \sim 500 – 775 Myr after system formation, and the total mass of the progenitor binary is limited to 3.4 – 4.2\,M_{\odot} . Furthermore, we conclude that the progenitor masses do not span independent intervals for M_{1,a} and M_{1,b} , but must instead lie within the narrow diagonal band shown in Fig. 5 . Within this band, assuming that M_{1,b}\geq M_{1,a} , the primary component is constrained by the turn-off mass as an upper limit and M_{\mathrm{TO}}/2 as a lower limit, M_{1,b} falls within a mass range of M_{1,b}\simeq 1.7 – 2.5\,M_{\odot} ). To the contrary, the lower mass companion lies within the range M_{1,a}\simeq 0.9 – 2.1\,M_{\odot} , depending on the total mass of the system. Acknowledgements. This research was supported by the ‘SeismoLab’ KKP-137523 Élvonal grant of the Hungarian Research, Development and Innovation Office (NKFIH), and by the LP2025-14/2025 Lendület grant of the Hungarian Academy of Sciences. The simulations were performed on the high-performance computing machine at HUN-REN CSFK CSI acquired through the EC Horizon2020 project OPTICON (Grant Agreement No. 730890). This research made use of NASA’s Astrophysics Data System Bibliographic Services, as well as of the SIMBAD and VizieR databases operated at CDS, Strasbourg, France. This paper was written with the assistance of a large language model (e.g., Gemini by Google, Microsoft Copilot, and ChatGPT by OpenAI) for linguistic refinement. All content and any errors remain the sole responsibility of the first author. Appendix A MESA Model Grid for the Primary Component Calculated MESA evolutionary tracks for the primary component of the \gamma Persei system. Figures 6 , 7 , and 8 present the high-, moderate-, and low overshoot models with stellar masses increasing from top to bottom and metallicites increase from left to right panels. Each evolutionary track is color-coded by the \alpha_{\mathrm{MLT}} parameter. The cyan circles denote the primary component of the \gamma Persei system derived from observational data. Figure 6: Calculated MESA evolutionary tracks for the primary component of the \gamma Persei system for high overshoot models. Stellar masses increases from top to bottom and metallicites increase from left to right panels. Each evolutionary track is color-coded by the \alpha_{\mathrm{MLT}} parameter. The cyan circles denote the primary component of the \gamma Persei system derived from observational data. Figure 7: Same as Figure 6 for moderate overshoot models. Figure 8: Same as Figure 6 for low overshoot models. Appendix B MESA Model Grid for the Secondary Component Calculated MESA evolutionary tracks for the secondary component of the \gamma Persei system. Figures 9 , 10 , and 11 present the high-, moderate-, and low overshoot models with stellar masses increasing from top to bottom and metallicities increase from left to right panels. Each evolutionary track is color-coded by the \alpha_{\mathrm{MLT}} parameter. The green triangles denote the secondary component of the \gamma Persei system derived from observational data, while the red circle represents the main-sequence turn-off point for each evolutionary track. Figure 9: Calculated MESA evolutionary tracks for the secondary component of the \gamma Persei system for high overshoot models. Stellar masses increase from top to bottom and metallicities increase from left to right panels. Each evolutionary track is color-coded by the \alpha_{\mathrm{MLT}} parameter. The green triangles denote the secondary component of the \gamma Persei system derived from observational data, while the red circle represents the main-sequence turn-off point for each evolutionary track. Figure 10: Same as Figure 9 for moderate overshoot models. Figure 11: Same as Figure 9 for low overshoot models. Table 7: Age and mass ranges where the two components of the system intersect MESA evolutionary tracks (within the observational errors). Ages are given in Myr. Primary component Secondary component [Fe/H] 1 t_{1} [Myr] M_{1} [ M_{\odot} ] R_{1} [ R_{\odot} ] [Fe/H] 2 t_{2} [Myr] M_{2} [ M_{\odot} ] R_{2} [ R_{\odot} ] High overshoot models \alpha_{\rm MLT}=1.5 -0.29 295–316 3.2–3.3 22.041 – 23.661 -0.30 708–776 2.2–2.6 3.732 – 3.789 -0.19 288–316 3.3–3.4 22.006 – 23.739 -0.20 724 2.3 3.840 – 3.840 -0.09 282–309 3.4–3.5 22.270 – 24.007 -0.10 691 2.4 3.949 – 3.949 \alpha_{\rm MLT}=1.7 -0.29 275–316 3.2–3.4 21.323 – 22.778 -0.30 691–831 2.2–2.6 3.713 – 4.064 -0.20 269–316 3.3–3.5 21.253 – 22.929 -0.20 708–891 2.2–2.3 3.760 – 3.850 -0.10 288–309 3.4–3.5 21.504 – 23.130 -0.10 676 2.4 3.837 – 3.928 \alpha_{\rm MLT}=1.9 -0.29 269–288 3.3–3.4 21.695 – 23.422 -0.30 676–794 2.2–2.3 3.920 – 4.049 -0.20 245–282 3.4–3.6 21.227 – 23.677 -0.20 645–871 2.2–2.6 3.708 – 4.095 -0.10 245–282 3.5–3.7 21.280 – 22.982 -0.10 660 2.4 3.754 – 3.838 \alpha_{\rm MLT}=2.1 -0.29 263–316 3.2–3.4 21.646 – 23.990 -0.30 660–794 2.2–2.3 3.854 – 4.001 -0.20 239–363 3.2–3.6 21.369 – 24.189 -0.20 631–851 2.2–2.6 3.722 – 4.092 -0.10 239–389 3.2–3.7 21.209 – 24.163 -0.10 602–812 2.3–2.5 3.789 – 4.095 \alpha_{\rm MLT}=2.3 -0.29 282–309 3.2–3.3 22.093 – 24.100 -0.30 660–776 2.2–2.3 3.759 – 4.094 -0.20 257–363 3.2–3.5 21.287 – 24.179 -0.20 616–831 2.2–2.4 3.718 – 4.067 -0.10 235–389 3.2–3.7 21.231 – 24.089 -0.10 588–891 2.2–2.5 3.720 – 4.093 \alpha_{\rm MLT}=2.5 – – – – -0.30 660–758 2.2–2.6 3.727 – 4.007 -0.20 316–316 3.2–3.2 21.329 – 24.149 -0.20 616–812 2.2–2.4 3.908 – 3.996 -0.10 263–389 3.2–3.5 21.329 – 24.149 -0.10 588–871 2.2–2.5 3.715 – 4.064 Med overshoot models \alpha_{\rm MLT}=1.5 -0.29 219–235 3.5–3.6 22.324 – 23.771 -0.30 660–758 2.2–2.6 3.732 – 4.034 -0.20 213–245 3.5–3.7 22.438 – 23.880 -0.20 616–831 2.2–2.4 3.720 – 4.026 -0.10 213–245 3.6–3.8 21.403 – 24.080 -0.10 588–891 2.2–2.5 3.700 – 4.091 \alpha_{\rm MLT}=1.7 -0.29 204–323 3.2–3.7 21.349 – 24.158 -0.30 645–758 2.2–2.6 3.707 – 4.029 -0.20 204–229 3.6–3.8 21.370 – 24.115 -0.20 616–812 2.2–2.4 3.887 – 4.012 -0.10 213–229 3.7–3.8 21.592 – 22.994 -0.10 575–871 2.2–2.5 3.716 – 4.081 \alpha_{\rm MLT}=1.9 -0.29 204–331 3.2–3.7 21.423 – 24.181 -0.30 741 2.2 3.962 – 3.962 -0.20 200–355 3.2–3.8 21.208 – 23.908 -0.20 602–794 2.2–2.6 3.777 – 4.092 -0.10 209–316 3.4–3.8 21.309 – 24.199 -0.10 575–758 2.3–2.5 3.700 – 4.063 \alpha_{\rm MLT}=2.1 -0.29 200–331 3.2–3.7 21.226 – 24.129 -0.30 724 2.2 3.939 – 3.939 -0.20 194–355 3.2–3.8 21.275 – 24.134 -0.20 602–776 2.2–2.6 3.706 – 4.064 -0.10 209–355 3.3–3.8 21.348 – 24.126 -0.10 562–741 2.3–2.5 3.708 – 4.079 \alpha_{\rm MLT}=2.3 -0.29 224–331 3.2–3.5 21.428 – 23.985 -0.30 708 2.2 3.876 – 3.967 -0.20 209–355 3.2–3.7 21.214 – 24.143 -0.20 588–776 2.2–2.4 3.702 – 4.094 -0.10 204–380 3.2–3.8 21.264 – 24.191 -0.10 562–741 2.3–2.5 3.715 – 4.088 \alpha_{\rm MLT}=2.5 – – – – -0.30 708 2.2–2.6 3.900 – 3.900 -0.20 269–295 3.2–3.3 21.392 – 24.165 -0.20 588–758 2.2–2.4 3.711 – 4.014 -0.10 251–380 3.2–3.5 21.392 – 24.165 -0.10 550–724 2.3–2.5 3.704 – 4.097 Low overshoot models \alpha_{\rm MLT}=1.5 -0.29 178–309 3.2–3.8 21.448 – 24.178 -0.30 708 2.2–2.6 3.732 – 3.935 -0.20 186 3.8 21.455 – 21.535 -0.20 676–758 2.2–2.3 3.954 – 4.028 – – – – -0.10 550–724 2.3–2.5 3.715 – 4.052 \alpha_{\rm MLT}=1.7 -0.29 178–323 3.2–3.8 21.213 – 23.175 -0.30 691 2.2–2.6 3.727 – 3.818 -0.20 229–316 3.3–3.6 21.258 – 24.181 -0.20 660–741 2.2–2.3 3.931 – 4.003 -0.10 224 3.7 24.063 – 24.063 -0.10 550–708 2.3–2.5 3.708 – 4.065 \alpha_{\rm MLT}=1.9 -0.29 178–331 3.2–3.8 21.353 – 24.055 -0.30 691 2.2 3.824 – 3.824 -0.20 186–323 3.3–3.8 21.421 – 23.872 -0.20 660 2.3–2.6 3.969 – 3.969 -0.10 209–288 3.5–3.8 21.203 – 23.877 -0.10 550–708 2.3–2.5 3.709 – 4.084 \alpha_{\rm MLT}=2.1 -0.29 175–331 3.2–3.8 21.238 – 24.195 – – – – -0.20 182–323 3.3–3.8 21.234 – 24.171 -0.20 645 2.3–2.6 3.914 – 4.005 -0.10 195–316 3.4–3.8 21.201 – 24.147 -0.10 616–691 2.3–2.4 3.734 – 4.076 \alpha_{\rm MLT}=2.3 -0.29 200–331 3.2–3.6 21.308 – 24.113 -0.30 676 2.2 3.839 – 3.839 -0.20 182–355 3.2–3.8 21.237 – 24.196 -0.20 645 2.3 3.945 – 3.945 -0.10 190–380 3.2–3.8 21.201 – 24.135 -0.10 602–691 2.3–2.4 3.703 – 4.091 \alpha_{\rm MLT}=2.5 -0.29 263–269 3.2 22.848 – 23.551 -0.30 602–676 2.2–2.6 3.791 – 3.791 -0.20 224–355 3.2–3.5 21.447 – 24.115 -0.20 631 2.3 3.860 – 3.977 -0.10 219–380 3.2–3.6 21.245 – 24.200 -0.10 602–691 2.3–2.4 3.701 – 4.073 Table 7: continued.