196 CENTAUREA. [CH. VIII 



side of the lamellae produces no effect. Oliver^ has shown 

 that if one lamella is prevented from moving, a touch on 

 it still provokes movement in the other lamella. In our 

 experiments we fixed one lobe by cementing it to the 

 corolla, using mastic dissolved in ether for the purpose. 

 When the cement is dry it is well to push a strip of wood 

 between the style and the corolla: by a wedge of this 

 sort the fixed lamella is forced to be more or less horizontal 

 and is perfectly free from contact, even at the base, with 

 the opposite lobe. 



(245) Stamens of Centaurea cyanus. 



We use the garden form of this species for demon- 

 strating the fact that the stamens are irritable ^ 



Select a floret which has expanded but has not 

 extruded its style, placg it on a piece of wet filter-paper 

 and under a simple lens split the floret at the swollen 

 part of the corolla-tube, which can then be opened wide 

 enough to give a view of the filaments. Cover the 

 preparation with an inverted watch-glass and leave it for 

 15 minutes, — to recover from the effects of the operation. 

 If the filaments are now gently touched, a writhing 

 contraction is plainly seen. The anther tube may in this 

 way be made to swing over first to one side then to 



' Berichte d. deutschen Botan. GeselUch. v. 1887, p. 167. Transmission 

 of stimulus has only been observed in Martynia lutea, M. proboscidea 

 and Mimulus cardinalii. There is said to be no transmission in Mimulus 

 luteus. 



2 Pfeffer employed C.jacea and Cynara scolymus. See hie Physio- 

 logische Untersuchungen, 1873, p. 80. 



