A study has been made of the transfer of heat from 8 heated wall to air flowing through packed beds in a tube of 1 in. diameter. The data are correlated by equations: (a) for Dto-< 1600; a! < 900: P hDt _ _ = o.134 ( yy-@"( D;, ),J' k Dt G (b) for 1600 < ,-< 3500: P The thermal conductivity of the packing has no influence in region (a) but affects the transfer in region (b). The flow rates used were lower than those reported by LEVA et olii ([I], [2], [3]). but it appears likely that the present data trend to agreement with the functions proposed by LEVA as the flow rate is increased. The most significant features of these results are that in the flow ranges studied the length of the packed tube is an important variable, but the packing diameter has little effect on the heat transfer. R&um6-Etude experimentale de la transmission de la chaleur entre Is paroi chaude et I'air qui parcourt un tube & garnissege de 25 mm de diametre. Les equations ci-dessus (a et b) representent les r&&tats experimentaux. La conductibilite du garnissage eat sans influence dans le regime a, mais affecte la transmission dans le regime b. Les debits utilids sont plus faibles que ceux de LEVA et toll. [l], [2], [3]. 11 semble bien que les resultats actuels tendent vers ies formules proposees par LEVA, qmmd les debits croissent. Conclusion la plus significative 8, tirer des mesures: dans la gannne des debits BtudiBs, la longueur du tube garni est une variable importante, tandis que le diametre du garnissege n'a qu'un faible effet sur la chaleur transmise.