2 Observational data and reductions

2.1 Original data

The continuum energy distribution in the spectra of "normal" main sequence and chemically peculiar stars were taken from the catalog of stellar spectrophotometry (CSS). The observations reported in this catalog were made by Adelman, Pyper, Shore and White at Kitt Peak National Observatory, Mt. Wilson Observatory, and Palomar Observatory. The information on the telescope-instrument combination and details of the reduction procedure are given by Adelman (1978). A machine-readable version of this catalog is available at the Astronomical Data Center (ADC) and the Centre de Données Astronomiques de Strasbourg (CDS) (Adelman et al. 1989). This version includes the spectra of approximately 200 stars in the wavelength region from 3300 Å to 7100 Å. For each star, both individual and average energy distributions are presented in this catalog. In the following analysis only average energy distributions are used, because most of CP stars are well known variable stars and, probably, the calculation of the Balmer continuum slope near Balmer jump from the average energy distribution would give the average value of $T_\mathrm{eff}$ over the cycle of variability.

Recently, the final version of the Pulkovo spectrophotometric catalog of bright stars (PSC) in the range from 3200 Å  to 10800 Å  has been published (Alekseeva et al. 1996). This catalog is archived at the Pulkovo Observatory on a PC and consists of three parts: the first contains data for 602 stars in the wavelength region from 3200 Å  to 7350 Å, the second contains data for 285 stars in the wavelength region from 5000 Å  to 10800 Å  and the third contains the combine data from the preceding catalogues. All the catalogues are presented with a step of 25 Å  and with an accuracy of $1.5-2.0\%$. The information on the instrumentation, observations and details of the reduction procedure is given by Alekseeva et al. (1996). In our study only the first part of the PSC is used.

2.2 Correction for interstellar reddening

Some of the investigated stars from the catalog of stellar spectrophotometry may be reddened. Adelman (1980) estimated the effect of interstellar reddening by using the UBV photometry with the Q method of Johnson & Morgan (1953) and the synthesized u-b and b-y colors with the prescriptions of Crawford (1978). For the stars with E(B-V) > 0.03, Adelman corrected the data for the effect of interstellar reddening by using of the Schild (1977) interstellar extinction law in combination with their adopted values of E(B-V). In the cases when E(B-V) > 0.03 for the investigated CP stars, the catalog of stellar spectrophotometry includes both the observed and the corrected for reddening average energy distributions. In our study for the stars with E(B-V) > 0.03 the average energy distributions corrected for the interstellar reddening are used.

In order to estimate the effect of interstellar reddening of the investigated CP stars from the PSC the procedure described in the paper by Sokolov (1995) is used. It was controlled by the UBV photometry according to the Q method of Johnson & Morgan (1953). The values of E(B-V) < 0.03 were obtained for investigated CP stars of this catalog. Moreover, the stars under investigation being located at the distances of the order of 100 pc or less from the Sun, hence the interstellar extinction was neglected. Therefore, the observed energy distributions of the investigated CP stars from the PSC were not corrected for interstellar reddening. But, it should be noted that there is still the fundamental problem to separate the effects due to interstellar reddening from those due to the non-normal energy distributions of the CP stars.

  

Figure 1: Comparison of $T_\mathrm{eff}$ derived from the Pulkovo spectrophotometric catalog of bright stars Alekseeva et al. (1996) with $T_\mathrm{eff}$ derived from the catalog of stellar spectrophotometry (Adelman et al. 1989). The straight line denotes the relation $T_\mathrm{eff}$(CSS)=$T_\mathrm{eff}$(PSC)

2.3 Flux calibration

The catalog of stellar spectrophotometry (Adelman et al. 1989) was calibrated with the fluxes of the primary standard star Vega as presented by Hayes & Latham (1975). On the other hand, the temperature calibration in the paper by Sokolov (1995) is established by using the stars calibrated with the absolute fluxes of Vega by Hayes (1985). To test the systematic differences between such various calibrations the method proposed by Sokolov (1995) has been applied to nineteen "normal" main sequence stars from the CSS and the computed temperatures were compared with $T_\mathrm{eff}$ from the paper by Sokolov (1995). The comparison of the temperatures from two sources of the data shows no pronounced systematic discrepancies between them, as illustrated in Fig. 1. That is confirmed by the following results: the mean value of $\Delta$$T_\mathrm{eff}$ = $T_\mathrm{eff}$(PSC)$- T_\mathrm{eff}$(CSS) is 67$\pm$138 K (the $\pm$138 K uncertainty refers to the rms standard deviation around the mean, devided by the square root of the number of stars), with a linear correlation coefficient r  = 0.935, and $\alpha$ = 0.998 for the regression line slope. The direct comparison of two calibrations of the absolute energy distribution for Vega given by Hayes (1985) and by Castelli & Kurucz (1994) shows that the agreement between these calibrations is excellent (see Table 3 of Castelli & Kurucz 1994). Thus it is possible to use the catalog of stellar spectrophotometry by Adelman et al. (1989) for calculation of $T_\mathrm{eff}$ from the slope of the Balmer continuum determined between 3300 Å  and 3600 Å with temperature calibration of Sokolov (1995) without any correction for the differences between these flux calibrations.

The final list contains 68 chemically peculiar Si, Cr and SrCrEu types (CP2 stars). One manganese star (36 Lyn) was also included, because this star was investigated both by Stepien & Dominiczak (1989) and by Adelman (1985). Three stars in our list ($\beta$ Tau, $\varepsilon$ UMa, HD 200311) have a double nature. The selected stars are listed in Table 1 together with their peculiarities and the number of individual scans.