We report an experimental study of peak and phase-space density of a two-stage magneto-optical trap (MOT) of 6 Li atoms, which exploits the narrower 2S 1/2 → 3P 3/2 ultraviolet (UV) transition at 323 nm following trapping and cooling on the more common D2 transition at 671 nm. The UV MOT is loaded from a red MOT and is compressed to give a high phase-space density up to 3 × 10 −4 . Temperatures as low as 33 μK are achieved on the UV transition. We study the density limiting factors and in particular find a value for the light-assisted collisional loss coefficient of 1.3 ± 0.4 × 10 −10 cm 3 /s for low repumping intensity.

We report the observation of a very high flux of ultra-cold bosonic chromium atoms in a magnetic guide. The beam is created by operating a magneto-optical trap/moving optical molasses within the magnetic field of the guide. A relative detuning between two pairs of the cooling lasers cools the atoms into a frame moving along the axes of the guide. When the atoms are cooled into a moving frame with a velocity of 6 m s−1 we observe a maximum of the flux of 6 × 109 atoms s−1. For these parameters the transversal temperature of the atoms after a 25 fold increase of the confining magnetic potential is about 1.2 mK. The longitudinal temperature is 400 µK.

We have realized a magnetic guide for ultracold chromium atoms by continuously loading atoms directly from a Zeeman slower into a horizontal guide. We observe an atomic flux of 2 · 10 7 atoms/s and are able to control the mean velocity of the guided atoms between 0 m/s and 3 m/s. We present our experimental results on loading and controlling the mean velocity of the guided atoms and discuss the experimental techniques that are used.

Unusually narrow spectral features are seen in fluorescence peaks from 85 Rb atoms under the action of two driving laser fields that are in a three-dimensional molasses configuration. One of the lasers, L1, is held at a fixed detuning from the "cooling" transition, while the other, L2, is scanned across the "repumping" transitions. The fluorescence peaks are split into symmetric pairs, with the separation within a pair increasing with the detuning of the L1 laser. For large detunings, additional small peaks are seen. A simple model is proposed to explain these experimental observations.

IN CONNECTION withstudies of certain gaseous reactions in this laboratoryneed was felt for a method for the accurate analysis of small amounts of hydrocarbon gases. Most of the existing methods