1995A&AS..113..199S

1.2A resolution spectra over the range 3175-7575A have been obtained for the two components of the gravitational lens candidate HE 1104-1805 (z=2.31, m_B=16.7 and 18.6, separation =3.0arcsec; cf. Wisotzki et al. 1993), with the aim of setting limits on the sizes of the clouds producing the Ly-α, C IV, and Mg II absorption systems. The Ly-α absorption lines are strongly correlated in equivalent width, suggesting that the two lines of sight pass through the same clouds in all cases, and there are no significant differences in velocities within ≃10km/s between corresponding pairs of lines. From the 72 Lyα lines with W_rest>0.085A (and λ>3395A) detected at 5σ in A and 3σ in B we statistically derive a 2σ lower limit of 100h₅₀^–1kpc for the diameter of spherical Ly-α clouds, assuming that the lens redshift z_L>1, H₀=50h₅₀km/s/Mpc, q₀=0.5, and {LAMBDA}=0. Similarly, 2σ lower limits of 100, 60 and 20h₅₀^–1kpc are obtained for the region inside the clouds giving rise to lines with W_rest>0.17, 0.32 and 0.60A respectively. These values and the strong correlation between Ly-α cloud line equivalent widths are hardly compatible with the spherical mini-halo model. For the systems causing the C IV and Mg II absorption lines, we find 2σ lower limits of 28 and 22h₅₀^–1kpc respectively for W_rest>0..3A. Again the equivalent widths of corresponding lines are correlated, though some lines are not seen in both spectra; the redshift differences between corresponding pairs of lines are small. A damped Ly-α system is present in one spectrum (A) but not in the other (B), indicating a typical size of the order of the separation of the two lines of sight at that redshift (z_abs=1.6616), ∼8-25h₅₀^–1kpc depending on the actual geometry of the system. Some high-ionization lines in this system have similar characteristics in both spectra, while all low-ionization lines are much weaker in the spectrum of B. This suggests that, while the region causing the low-ionization lines may be similar in size to the damped Ly-α region, that causing some of the high-ionization lines must be considerably larger. The absorption line properties provide further support for the gravitational lens interpretation. We also discuss the differences in the continuum spectra already noted by Wisotzki et al. and the possible geometry of the lens configuration.