162 MOVEMENT OF CILIA AND FLAGELLA 



because only in the effective stroke can cilia assist each other, 

 while in the longer recovery stroke the ciliary movements will be 

 slowed by the increased viscous resistance. In fact, the bending 

 cycle takes about the same time in viscous media in spite of the 

 fact that the amplitude is reduced. 



In the bending cycle of a symmetrically beating flagellum, a 

 fibrillar contraction follows the return to normal length after 

 stretching of the fibril, i.e. it immediately follows shortening, and 

 it is easy to visualise the way in which shortening could excite 

 contraction. Much the same may happen in the recovery phase 

 of ciliary beating, for, as Harris wrote, " distal to each element 

 of the advancing contraction on the effective side will be a region 

 that is bent into a concave curve " ; this bending could excite the 

 fibrils by their shortening on that side. Fibrils on the opposite 

 side may be excited as in flagella. The bent region is more or 

 less limited to the basal region of the cilium for most of the 

 effective phase, and the bending wave only starts to move up the 

 ciliary shaft towards the end of this period. In cilia that are 

 short (or compound cilia of greater length) the effective stroke 

 is quickly completed and the bending wave soon spreads up the 

 cilium, but in longer cilia the effective stroke may take longer, 

 so that some time elapses before the bend at the base is complete 

 and contraction spreads up the cilium. The shortening theory 

 of propagation of the contraction explains this adequately, for 

 only towards the end of the effective phase will the tension in 

 fibrils on the effective side of the cilium be released sufficiently 

 for the fibrils to shorten and propagation of the contraction to 

 take place. 



If such a simple mechanical explanation is correct, it is surprising 

 that glycerine extracted models do not show propagation of the 

 contraction, although it is possible that in this case the contractions 

 are not localized, but extend for the whole length of the cilium. 

 Some further observations on these models may be very valuable 

 in this respect. 



The propagation of the contraction is evidently closely 

 associated with the contraction itself, for environmental changes 

 which affect the frequency of beat also seem to affect the rate 

 of propagation. For example, in viscous media the rate of 

 propagation is slowed by about the same amount as the frequency, 



