We have obtained a high signal-to-noise ( 30 \leq S/N \leq 70 ) high resolution ( FWHM = 20 km s ^ { -1 } ) spectrum of the radio-quiet QSO HS 1700+6416 ( z _ { em } = 2.72 ) with the echelle spectrograph on the KPNO 4m telescope .
We detect 13 metal systems in the optical spectrum of this QSO , including six systems with associated optically thin Lyman limit absorption in the HST spectrum obtained by Reimers et al .
We use the apparent column density technique and profile fitting to measure the heavy element column densities and to evaluate the impact of unresolved absorption saturation .
Profile fitting indicates that four of the \ion C4 systems are narrow with b < 8 km s ^ { -1 } , which implies that these absorbers are relatively cool and probably photoionized .

The dense cluster of \ion C4 doublets at 2.432 < z _ { abs } < 2.441 shows the weak line of one \ion C4 absorber apparently aligned with the strong line of a different \ion C4 doublet , i.e. , line locked , for two pairs of \ion C4 absorbers .
Line locking has been detected previously in z _ { abs } \approx z _ { em } absorbers where radiation pressure is likely to play a role , but it is surprising in this case since this \ion C4 complex is displaced by \sim 24000 km s ^ { -1 } from the QSO emission redshift .
This may be the remnant ( or precursor ) of a broad absorption line ( BAL ) outflow .
However , it is possible that these alignments are chance alignments rather than true line locking .

The high ion column density ratios in the multicomponent Lyman limit absorber at z _ { abs } = 2.3150 suggest that the ionization conditions in this absorber differ significantly from the conditions in the gaseous halo of the Milky Way .
From photoionization models we derive [ Si/H ] \geq –0.95 and [ Al/H ] \geq –0.96 for the stongest component of this absorber .
These are conservative lower limits derived from lower ionization stages only ; photoionization models in agreement with the observed low and high ionization stages require [ M/H ] \approx –0.45 .
In contrast , Vogel & Reimers derive [ N/H ] < –1.04 and [ O/H ] = –1.52 for this absorber .
We suggest that the discrepancy is due to the low resolution of the Vogel & Reimers data ( FWHM \approx 300 km s ^ { -1 } ) which introduces serious blending and saturation problems .
The photoionized model with [ M/H ] = –0.45 has a particle density n _ { H } \approx 0.02 cm ^ { -3 } , a size of a few hundred pc , and a mass of roughly 1 \times 10 ^ { 5 } M _ { \odot } , assuming the absorber is spherical .

We detect unsaturated \ion C4 and rather strong \ion N5 “ associated ” absorption at z _ { abs } = 2.7125 .
The apparent column density of the weak \ion N5 1242.8 Å line is greater than the apparent column density of the stronger \ion N5 1238.8 Å line in this absorber , which indicates that the \ion N5 profile is affected by unresolved saturation or that the \ion N5 absorbing gas does not completely cover the QSO emission source .
If the latter interpretation is correct , then the associated absorbing gas must be close to the QSO .
We have used the observed spectral energy distribution of the QSO , corrected for intervening Lyman limit absorption , for photoionization modeling of the associated absorber , and we derive [ N/H ] \geq –0.65 and [ C/H ] \geq –0.82 .
Other corrections ( e.g. , to account for dust in intervening absorbers or ‘ Lyman valley ’ attenuation ) will make the spectral energy distribution harder and increase the metallicity estimates .
However , the absorption profiles suggest that the constant density single slab model is too simplistic : we obtain b ( \ion N5 ) = 25.2 \pm 1.3 km s ^ { -1 } and b ( \ion C4 ) = 11.4 \pm 1.1 km s ^ { -1 } from profile fitting , and therefore the gas which produces the \ion N5 absorption does not have the same temperature and non-thermal motions as the \ion C4 gas .

Finally , we briefly examine the number of \ion Mg2 systems detected per unit redshift , and we tentatively conclude that dN / dz is dominated by weak \ion Mg2 lines with W _ { r } < 0.3 Å .