14 



is the fact that the pnlvinus is approximately equally illuminated 

 on all sides. 



The deflection of the terminal leaflet upwards is explainable 

 on the ground that the piUvinus is illuminated on one ,-ide greater 

 than on the opposite. 



In diffuse one-sided light, the leaflets endeavor in general to 

 assume that relation with the light that their palisade surfaces 

 would lie perpendicular to the rays and toward their origin. 

 "When the midrib is horizontal and the light is from above, 

 the expanded position results. It is of some iuterest to notice 

 the behavior of the leaflets when placed at various angles with 

 the direction of the diffuse ligth. An experiment was arranged 

 in which small vials, filled with water, were inserted in holes in 

 a horizontal wooden cylinder. The holes were bored in such a 

 way that the bottles would stand out at different angles. A 

 leaf in the expanded condition was placed between the grooved 

 split halves of a cork in each vial. Care was taken to remove 

 the leaves under water and use them at once, although experience 

 had already shoM^n that they remain turgid and sensitive to 

 light for several hours after abscision. If kept turgid by being 

 placed in water or in a moist chamber, their sensitiveness will 

 continue for over a day. Hence on account of their quickness 

 to respond to light, there was practically no danger of abnormal 

 results. Leaves were placed by an open window in the following 

 four relations to the light : 



90° to light, edgewise. 



90° to light, spongy side upward. 



Tip of terminal leaflet toward light. 



Horizontal (150° to light) pointing from light, with palisade 

 side up. 



In the first instance. 90° to light and edgewise, a 90° twist 

 on the part of the pulvinus of the terminal leaflet would bring 

 the blade into a proper light relation. "With the lateral leaflets, 

 a 90° deflection of each leaflet in the same direction would 

 produce the desired result. The results of the experiment did 

 not. however, fully coincide with those which one might think 

 ought to occur. The terminal did seemingly attempt to approach 

 the perpendicular-to-light relation, but a twist of about 45° 

 only was attained. The lower lateral leaflet responded slowly, 

 rising after a time to an oblique angle Avith the light. Only the 



