5 Selection criteria for new DIBs

Despite the high quality of BD+63$^{\circ }$ 1964's spectrum, the presence of weak spurious absorption features, including uncorrected or poorly corrected bad pixels and cosmic rays, atmospheric residuals left after the telluric correction and mismatches in stellar abundances between BD+63$^{\circ }$ 1964 and HD 37128, results in the need for strong and consistent criteria for the positive identification of new DIBs. We checked that new DIBs could not be ascribed to such spurious features.

5.1 The standard stars and synthetic stellar spectrum

Although standard stars HD 37128 and HD 205021 were used to identify and eliminate stellar lines from our target spectrum, we have additionally produced a synthetic spectrum using the TLUSTY and SYNSPEC software (a stellar atmospheres model and emergent line synthesis package) to confirm the DIB identification. The synthetic stellar spectrum together with the spectra of the standard stars HD 205021 and HD 37128 allowed the identification of stellar lines to very good accuracy in the star BD+63$^{\circ }$ 1964. Some residual absorption in the divided spectrum which coincided exactly in wavelength with strong or moderate stellar lines together with a band width comparable to the rotational velocity of BD+63$^{\circ }$ 1964 were ascribed to slight mismatches in stellar abundances between HD 37128 and BD+63$^{\circ }$ 1964 and thus were not due to bona fide DIBs in the line-of-sight.

5.2 The presence of new DIBs in other reference targets

The most important test of new DIBs are their occurence in other well known reddened stars. The large number of DIBs detected toward BD+63$^{\circ }$ 1964 may be exceptional and some DIBs might possibly only be observed in such hot and heavily reddened targets. Nevertheless the considerable strength of several DIBs suggested they might be visible in other stars. As previously mentioned HD 183143 and BD+40$^{\circ }$ 4220 were chosen as our DIB reference targets. The latter was selected due to its very reddened line-of-sight, and HD 183143 is a well known DIB source and a main target of the DIB survey by Jenniskens & Désert (1994) and Herbig (1995). HD 183143 (spectral type B7 I) suffers from some stellar line contamination and BD+40$^{\circ }$ 4220 (spectral type O7) has a faint visible magnitude of 9.05. For BD+40$^{\circ }$ 4220 we could therefore obtain a spectrum with a signal-to-noise ratio of only 88 in V but rising to 170 in the red.

All new DIB candidates were cross checked against these spectra and, after eliminating the possibility of stellar line contamination, 60 of these DIB candidates were confirmed in the spectrum of either HD 183143 or BD+40$^{\circ }$ 4220. Only those DIBs visible in either BD+40$^{\circ }$ 4220 or HD 183143 along with BD+63$^{\circ }$ 1964 were included in Table 4.