4 OSBORN, Staniinal Mechanism of Passifiora Caerulea. 



take to mean extrorse) in a line with the thick stiff 

 filaments," when the flower first opens. The filaments 

 have already begun to move outwards, forming an angle 

 of about 30° with the axis of the gynophore, having by 

 now completed their full growth. The end is blunt, and 

 there is a tendency for the tissue on each side of the con- 

 nective to expand, forming, as it were, two shoulders. 

 The tissue of these has, also, thick-walled rounded cells. 

 The anthers now hang in the position shown in Fig. II., 

 the knobs on their backs impinging on the shoulders of 

 the filament. This state of equilibrium is unstable, there 

 being a tendency to swing laterally, the four projections 

 serving as bearings. In the great majority of cases this is 



i^ 



Fig. III. Side view of anther showing the way in which 

 the knobs of tissue on its b:ick are clasped by the distal 

 end of the filament. The distance between the anther 

 and the filament is exaggerated to show the connective. 

 X4. 



what happens, and the second movement described by 

 Warnstorf occurs, viz., the anthers come into a tangential 

 plane by moving through 90'' about the connective. In 

 this position the knobs interlock, as shown in Fig. III., 

 the end of the filament curling laterally, and clasping the 

 projections on the anthers. A very elementary knowledge 

 of mechanics will show that the anther would not 

 remain in this position but for the formation of the joint 

 that directs it there. That this direction does actually 

 occur can easily be proved by twisting an anther back into 

 a line with the filament, when after a little while it will of 

 itself resume the crosswise position. The swing of the 

 anther may be in either direction, and is apparently 



