The very oldest and metal-poor stars are of special importance in the study of the early galactic history. Our detection of truly unexpected amounts of beryllium in the extreme Pop II star HD 140283 (Gilmore et al. 1991, ApJ 378, 17) was received with considerable interest from Big-Bang nucleosynthesis theorists. The reason was that no process "after" the Big Bang was then known to be able to produce such large amounts of beryllium, and only an "inhomogeneous Big Bang" might possibly be able to do so. Later studies by us and others, of beryllium in more stars, and also of the element boron, strongly suggest that the source of the beryllium and boron was local to the Milky Way, rather than the Big Bang itself. (Our other contributions are found in Gilmore et al. 1992, Nature 357, 379-383; Edvardsson et al. 1994, A&A 293, 75 (based on HST spectra); Nissen et al. 1994, A&A 285, 440-450.) In a collaboration led by Ramon Garcia López (Tenerife) we obtained more HST time for studies of boron in two more extreme pop II stars, and together with a re-analysis of 7 other stars in the literature shows that the B/Be abundance ratio in the Early galaxy was constantly about 20. This result confirms that galactic spallation reactions were responsible for these elements in the early galaxy (Garcia López et al. 1997, ApJ 500, 241).

In a project lead by my PhD student Patrik Thorén, we have received 8 orbits of Hubble Space Telescope (HST) time to use the new STIS spectrograph for a study of the boron abundance in the extreme Pop II star HD 140283. Earlier studies of boron in halo stars (including our own detailed HST study of HD 140283, Edvardsson et al. 1994, A&A 293, 75) have exclusively utilized a single B I absorption line near 2497 Å. As a much-needed check of these results, we will study the 2090 Å B I line. This study also serves to investigate whether the 2090 Å line is free of line-blends, so as to be useful for future studies with improved instrumentation of the abundance ratio between the two boron isotopes 10 and 11. This ratio is also sensitive to which mechanism(s) were important in the early galactic production of light elements.

Our analysis shows that we can not detect the boron line in the noisy spectrum. All we can do is to determine an upper limit of the total boron abundance. We therefore plan to apply for more time at the HST to ascertain detection of the line, and make an important consistency check.