Abstract
Exudation was studied from the mouth-parts of the willow aphid Tuberolachnus salignus (Gmelin) inserted in the phloem of isolated 15 -cm. stem segments of Salix viminalis. By supplying water under pressure to the cut surface of the wood at one end of a segment and allowing it to run freely from the other, a pressure gradient could be set up in the xylem. By closing the free end, the pressure in the xylem could be raised uniformly. This last treatment caused the rate of exudation to increase. This effect of pressure could be balanced by an equivalent osmoticum introduced into the xylem. The increase was therefore due to the DPD in the xylem rather than any intrinsic effect of pressure. The effect of a pressure gradient (4 atm. applied) depended on the position of the mouth-parts in relation to the gradient. If they were near the low-pressure end, exudation from them increased. If they were near the high-pressure end, exudation decreased. In neither case was the total carbohydrate or total nitrogen concentration of the exudate changed. When longer segments were used (e.g. 41 cm.) the effect of a similar pressure (4 atm.) was greater than with 15-cm. segments although the point of application of the pressure was more remote from the mouth-parts. When the wood at the midpoint of a segment was severed leaving a narrow bridge of bark joining the two halves, raising the pressure in the xylem of one half led to an increased exudation from mouth-parts situated in the other half. Thus the effect of pressure was transmitted along the phloem since there was no rise in xylem pressure in the half segment bearing the mouth-parts. 10-4 M. dinitrophenol applied to a bark abrasion a few centimetres away from mouth-parts, reduced exudation. Applied to the cambial surface, exudation was reduced by 0.1 per cent, phloridzin and stopped completely by 10-4 M. dinitrophenol and 10-6 M. potassium cyanide. Inhibitors had no effect on the concentration of the exudate. No simple interpretation of the effects of pressure gradients is possible, but it is suggested they give qualified support to the Munch hypothesis.