Previous
abstract 
Next
abstract
Optical spectra of QSO 1937-1009 ($z_{em} = 3.78$, V $\simeq 17.5$) with the W. M. Keck telescope HIRES echelle spectrograph show deuterium in the Lyman$\alpha $ and Lyman$\beta $ lines of an absorption systems at 3.57. The H~I Lyman series lines are narrow ($b \simeq 20$ km~s$^{-1}$) and resolved all the way to Lyman-19. They absorb all the flux in their cores, and there is strong Lyman continuum absorption, showing that the H~I column density is $log N(H~I) \geq 17.8$. Weak metal lines in the system, C~IV, CII, SI~IV and Si~II all show the same profiles which are adequately described by two clouds separated by 13 km~s$^{-1}$. A simultaneous fit all the Lyman series lines, especially the weaker high order lines, indicates a total H~I column of about $10^{18}$, with 80\% in the lower redshift component. The same fit gives D/H $\simeq 2 \times 10^{-5}$ (preliminary), which corresponds to a high cosmological baryon density of $\Omega_b \simeq 0.03~h_{100}^{-2}$.
Both clouds are moderately ionized, with H~I/H $\simeq 0.01$. The metal abundances are about [C/H] $\simeq -3.1$ and [Si/H] $\simeq -2.7$ in the component with the most H~I and D~I, and [C/H] $\simeq -2.0$ and [Si/H] $\simeq -1.7$ in the other cloud. Both have [C/Si] $\simeq -0.3$, like low abundance halo stars, and their abundances are much too low for stars to have removed significant deuterium.
These results are consistent with D/H of (1 -- 2) $\times 10^{-5}$ from the interstellar medium, but not with Songalia et al. (1994, Nature, 368, 599) who reported D/H $= 2.6 \times 10^{-4}$, at least ten times more D, in a different QSO absorber. Either their absorption line was H rather than D, or primordial nucleosynthesis was dramatically inhomogeneous.
