No C+ emission from the z=3.137 damped Lyman-alpha absorber towards PC1643+4631A

We describe a search for redshifted [C II] in a z=3.137 damped Ly-alpha absorption system that has a large neutral hydrogen column density and which was controversially reported to be a source of CO emission, indicative of rapid star-formation (Frayer, Brown & Vanden Bout 1994; Braine, Downes & Guilloteau 1996). There is no sign of [C II] emission in our spectrum, which was obtained during excellent observing conditions at the James Clerk Maxwell Telescope (JCMT) and covers 1890 km/s. The upper limit we have placed on the integrated line intensity (3 sigma(T_MB) < 5.9 K km/s for a profile akin to that of the CO lines) constrains the [C II]/CO(1-0) line-intensity ratio to 3 sigma < 8300, based on the line intensity reported by Frayer et al. (1994), or to 3 sigma < 58700 based on the data obtained by Braine et al. (1996). These limits are consistent with values measured in the Galactic plane and for nearby starburst nuclei; the former, however, is significantly lower than the ratio found in low-metallicity systems such as the Large Magellanic Cloud (which might be expected to have much in common with a damped Ly-alpha absorption system at high redshift). This can be taken as evidence against the reality of the CO line detections, with the proviso that a system significantly larger than present-day disk galaxies would NOT have been fully covered by our small beam whereas it WOULD have been properly sampled by the Frayer et al. observations. Finally, we demonstate (as did Ivison et al. 1996) that knitting together overlapping bands can generate erroneous results - specifically, an emission feature that has a width, profile and central velocity consistent with the controversial CO emission lines and which could have drawn us to entirely the wrong conclusions.