Discussion following Irina Glushneva's talk

White: One comment about the Markarova monograph, which is not familiar to many of us. I have copies, and there's a wealth of information in it, about the instruments used as well as extensive tabulations of the data. We've tried unsuccessfully for two years to get the funds together to have it translated into English, but it's an extremely useful volume.

Hall: I was wondering about the Hayes 1985 Vega calibration, because there was recently a paper published in A&A by Megessier, analyzing all the different Vega calibrations. He settled on the Hayes & Latham 1975 and the Tug, White, & Lockwood 1977 calibrations as the best. Wasn't the 1985 ``calibration'' just an average of lots of others?

Garrison: That calibration came from one of the Dudley Observatory workshops, and I seem to remember there were lots of objections to it, but I can't remember what they were.

Hall: I think Megessier's objection was that Hayes 1985 was an average of calibrations from different sources, so for example, you weren't really sure what the internal errors were. So it might affect the spectrophotometric calibration a bit if you used a more uniform source like Hayes & Latham, which a lot of people use. Have you looked at how that might affect your choice of the best solar analogs?

Glushneva: Hayes analyzed all the calibrations and made a weighted mean, with maximum weight on his own, of course [laughter].

Hall: Ah, and there's the big objection, it was this average and you weren't sure which calibration it was.

Glushneva: My opinion is that it is more correct to use some absolute calibration of Vega flux and the weighted mean, as it was done in the case of the Hayes 1985 Vega calibration, than to choose one "best" calibration because of possible systematic errors present in any spectrophotometric calibration. For four calibrations used by Hayes for weighted mean in the visual range the systematic agreement is of the order of the internal error. The mean standard deviation is only 1.5% at 5000 Å. It is a very good result. However, in the UV and red, the agreement is worse. In producing the weighted mean. Hayes gave weight (2) to Hayes and Latham 1975 (HL) calibration, and to the rest weight (1). The weighted mean lies between approximately HL and Tug. et al. 1977 calibrations in the visual range, but in the UV, the Tug et al. calibration is higher than HL one, which is nearer to the weighted one.

Garrison: One comment, I noticed in your table you had the Hyades vB 64 listed as G6 V. That was a misidentification of a star by Hiltner, who took the spectra. Morgan classified that star correctly; vB 64 is really G2 V.

Glushneva:The spectral type of vB 64 G6V was taken from the paper by Campins et al. (1985) with infrared color indices. The corresponding line of Table 4 has index C, showing the author of publication. But the photometric catalog by Blanco et al. (Publ. of the US Naval Observatory, Washington, 1970) gives spectral type G0 for this star. So there are significant differences in spectral type between various publications for this star.

Garrison:The real vB 64 is very close to G2 V. Hiltner misidentified a star he thought was vB64, and Morgan classified that star correctly as G6 V. So Morgan never really classified vB64. This is important, because Hardorp made quite a case out of vB64, all based on Morgan's correct classification of the wrong star. In fact, Hardorp made it his strongest case, because he mistakenly took Morgan's classification as confirmation of his "cooler Sun."