300 The Botayiical Gazette. [November, 



downward at the end by the weight of the carpels, which 

 hang opposite to each other when in their normal position. 

 At this stage in the process of ripening, the carpels have been 

 detached from the carpophore for the greater part of their 

 length, only a small part of it adhering at the extremity. To 

 be completely detached it needs to be stripped up, for the 

 loosening of the carpel takes place from below upward. But 

 the adhesion at the top may be strong enough to turn the 

 fruit over while released under the force of a sudden blow. 

 So it will be slung from its support and propelled the lower 

 end foremost. This end, being less tapering and rather 

 heavier, will facilitate the process. 



To illustrate my meaning by a pair of carpels, we will con- 

 sider the blows as coming from the right hand. By the dis- 

 placement of the stem to the left under this impulse, and its 

 sudden stoppage when the limit of motion is reached, the 

 energy imparted to the right hand carpel will cause it to turn 

 over in order more easily to tear itself away and it will be 

 thrown to the left of the plant. In some cases it may be 

 jerked from its support and thrown somewhat to one side, be- 

 ing diverted from the line of direction of the blow by stronger 

 adhesion to the carpophore and propelled with a diminished 

 intensity. The left hand carpel may also be detached, par- 

 ticularly if the blow be violent, and thrown forward with the 

 right, since more fruit is sent towards the point to which a 

 blow tends than in the opposite direction. But it often re- 

 mains attached, the arched and slender carpophore seeming 

 to act like a spring to weaken the shock it would have re- 

 ceived, and the carpel is carried forward with the stem on 

 which it is supported, to return with it to the position of rest, 

 or beyond, should the reaction be sufficient. The sudden stop- 

 ping of the stem when the limit of motion is reached on the 

 return will have an effect on the carpel like that in the former 



but in the opposite way, sending it to the right, or to- 



wards the point from which the blow tends. 



The clue to this explanation was obtained by experiment- 

 ing with plants held in one hand and snapping the branches 

 bearing the umbels with the fingers of the other. While 

 working in this way, one of the carpels flew from the stem 

 held in front and a little to one side, and shooting by fell on 

 a table back of my chair. It was heard to fail on some paper 

 lying on the table, and was readily found. The distance from 



