THE MECHANISM OF BEAT 



157 



It is assumed in all cases that the fibrillar contraction lasts about 

 the length of the full cycle, so that a second contraction takes 

 place when the first has completely died away. Even fairly large 

 errors in this assumption will not affect the patterns produced 

 very much. 



The model used represents a short length of the flagellum, and 

 a complete flagellum may be built up from a large number of 

 these models placed end-to-end. Contractions in the peripheral 

 fibrils will be propagated up the length of the flagellum from base 



RIGHT 



»- 



z 

 u 



z 



i 



o 



z 



5 



z 

 u 



(S 



z 

 < 

 I- 



M 



UJ 



LEFT 



Fig. 44. Graphs showing the calculated sequence of resultant 



bending moments (about the line X-Y in Fig. 41b) in cilia 



where the phase difference between adjacent fibrils is (a) ^, 



(b) y^ ^'^d (c) -J^ cycle (see text). 



to tip, activating each unit in turn, so that each unit will be 

 slightly out of phase with neighbouring units. If the bending 

 cycle of each unit is symmetrical, the bending cycle of the whole 

 flagellum will be symmetrical, while if the bending cycle of the 

 component units is asymmetrical, the bending cycle of the whole 

 flagellum will be asymmetrical and that of a cilium, for we have 

 seen that the beat of a Stentor membranelle is like a very asym- 

 metrical flagellar beat. The same contractile machinery, acting 



