In the next years a number of telescopes with aperture diameter much larger than the existing HST will see their first-light (Mountain et al. 1994). In this framework the Adaptive Optics (AO) technique (Beckers 1993) can provide a substantial improvement in the resolution of optical imaging in the next decades, provided that a reliable way to obtain routinely AO compensated images in the visible becomes available. While it has been already pointed out that using nearby Natural Guide Stars (NGSs) as reference does not allow for a full sky coverage, several techniques have been proposed in order to overcome the problem. Mesospheric Sodium Laser Guide Stars (LGSs), Foy & Labeyrie (1985), has been proven to be effective to close an AO control loop (Olivier et al. 1995). The first scientific results with Rayleigh LGSs have been published in the astronomical literature (Drummond et al. 1995; McCullough et al. 1995). The problems of conical anisoplanatism are well studied (Fried & Belsher 1994) and the solutions to the problems arose have been selected (Tallon & Foy 1990). LGS technology is, in practice, still in its infancy, and it is realistic to expect that in a decade these systems will be routinely available for most of the modern telescopes. The tilt indetermination problem (Pilkington 1987; Olivier et al. 1993), however, still requires a convincing solution: the Double Adaptive Optics (Rigaut & Gendron 1992) increases significantly the sky coverage, but the full sky is not obtained; the tristatic configuration (Belen'kii 1995; Ragazzoni 1996a) requires huge modifications on the telescope structure; the auxiliary telescopes technique (Ragazzoni et al. 1995) needs a large available space in the neighborouds of the observatory (Marchetti & Ragazzoni 1996), the propagation delay technique imposes to share the telescope aperture with the laser projector (Ragazzoni 1996b) and the multicolour LGS (Foy et al. 1992, 1995) requires a large laser power.
While one can hope that in the near future other techniques, or a combination of existing ones, should lead to a reliable, routinely usable, full AO compensation of any point in the sky, a careful analysis of the existing techniques should be carried out. In the following a detailed analysis of an uncovered aspect of the multicolour LGS technique is carried out.