Sl'MM.^RYXylfm reco\'CTy from cavitation-induced embolism was studied in 1-yr-old twigs of Laurus nobilis L. Cavitation was induced by applying pre-established pressure differentials (AP,,,) across the pit membranes of xylem conduits. AP,,, were M 3, 1'75 and 2'26 MPa, corresponding to about 50, 77 and 100"t, of tbe measured leaf water potential at the turgor loss point. AP,,, were obtained eitber by increasing xylem tensions or by applying positive pressures from outside, or by a combination ofthe two. The percentage loss of bydraulic conductivity (PLC) did not change, regardless of how the AP,,, were obtained. This confirmed that xylem ca\ itation was nucleated by microbubbles coming fTom outside the vessels. Positn e pressures, htnve\er, amplified (up to 75'^'o) and sped up the .\\'lem refilling (20 min) in comparison with tbat measured in unpressurized twigs (c.SO'^'o in 15 b). Twigs girdled proximally to tbeir pressurized segment 1 min after tbe desired pressure value bad been reacbed, did not recover from embolism. Tbe later tbe twigs were girdled vvith respect to wben they were tested for PLC, the higber was tbeir recovery from embolism, suggesting tbat some messenger was transported in tbe phloem wbicb stimulated xylem refilling. Indol-3-acetic acid (lA.-X) applied to the exposed cortex of both pressurized and unpressurized twigs, induced an almost complete reeo\ery from PLC. We hypotbesize that tbe refilling of capitated xylem might be a result of an auxin-induced increase in the pbloem loading with solutes. Tbis would cause radial transport of solutes to ca\'itated xylem conduits via tbe rays, thus decreasing their osmotic potential and making tbem refill. No positive xylem pressure potentials were measured during x>lem recovery from PLC.
