252 DEOSERA ROTQNDIFOLIA. CHAP. X 



the blade more quickly in a longitudinal than in a 

 transverse direction ; though with time it can pass in 

 any direction. We know that the same stimulus 

 causes movement of the tentacles and aggregation of 

 the protoplasm, and that both influences originate in 

 and proceed from the glands within the same brief 

 apace of time. It seems therefore probable that the 

 motor impulse consists of the first commencement of 

 a molecular change in the protoplasm, which, when 

 well developed, is plainly visible, and has been desig- 

 nated aggregation ; but to this subject I shall return. 

 We further know that in the transmission of the aggre- 

 gating process the chief delay is caused by the passage 

 of the transverse cell-walls ; for as the aggregation 

 travels down the tentacles, the contents of each suc- 

 cessive cell seem almost to flash into a cloudy mass. 

 We may therefore infer that the motor impulse is in 

 like manner delayed chiefly by passing through the 

 cell-walls. 



The greater celerity with which the impulse is 

 transmitted down the long exterior tentacles than 

 across the disc may be largely attributed to its being 

 closely confined within the narrow pedicel, instead 

 of radiating forth on all sides as on the disc. But 

 besides this confinement, the exterior cells of the ten- 

 tacles are fully twice as long as those of the disc ; so 

 that only half the number of transverse partitions 

 have to be traversed in a given length of a tentacle, 

 compared with an equal space on the disc ; and there 

 would be in the same proportion less retardation of the 

 impulse. Moreover, in sections of the exterior ten- 

 tacles given by Dr. Warming,* the parenchymatous 



* ' Yidenskabelige Meddelolser de la Soc. d'Hist. nat. dc Copen 

 agiK',' Nos. 10-12, 1872, woodcuts iv. and v. 



