NO C-ALPHA ABSORBER TOWARDS PC 1643+4631A( EMISSION FROM THE Z=3.137 DAMPED LYMAN)

We describe a search for redshifted [C II] in a z = 3.137 damped Ly al pha 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] emissio n in our spectrum, which was obtained during excellent observing condi tions at the James Clerk Maxwell Telescope (JCMT) and covers 1890 km s (-1). The upper limit we have placed on the integrated line intensity (3 sigma(T-MB) < 5.9 K km s(-1) for a profile akin to that of the CO l ines) 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 pla ne and for nearby starburst nuclei; the former, however, is significan tly 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, wi th the proviso that a system significantly larger than present-day dis k 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 tog ether overlapping bands can generate erroneous results - specifically, an emission feature ne that has a width, profile and central velocity consistent with the controversial CO emission lines and which could h ave drawn us to entirely the wrong conclusions.