THE MECHANISM OF BEAT 159 



in such a way that the bending to one side is nearly maximal, but 

 continues for a smaller part of the complete cycle, while the 

 bending to the other side is much less and continues for a longer 

 time. All gradations are possible, therefore, from a symmetrical 

 flagellar beat to an extremely asymmetrical ciliary beat, merely on 

 the basis of a change in phase difference between the contraction 

 of neighbouring fibrils of the peripheral ring. 



It may be possible one day to test this idea, for many factors 

 may vary the phase difference, either by their effect on the 

 contraction or excitation processes, but we do not yet know 

 enough about the normal beat in most cases to be able to detect 

 anything but a major change in the form of beat. It would be 

 most instructive to consider those factors which destroy the 

 symmetry of beat of a sperm tail; perhaps someone will soon 

 study this, for at present we only know that the beat tends to 

 become more asymmetrical with age (Gray, 1955), a result that 

 might be expected if the contraction was slowed while the 

 excitation between fibrils remained at the same rate. 



The actual time required for the proposed fibrillar excitations 

 seems reasonable for both sea urchin sperm tails and Stentor 

 membranelles. If the phase difference is about one -sixth in 

 symmetrically beating sperm tails, and each bending cycle takes 

 about 30 msec, the excitation interval between neighbouring 

 fibrils is about 5 msec. In Stentor, if the phase difference is about 

 one-twenty-fourth (as in Fig. 44c), and the time for one beating 

 cycle is 36 msec, then the excitation interval would be 1 '5 msec. 

 A parallel case in nerve physiology might be the time delay during 

 excitation at a nerve synapse, where an interval of 1 to a few 

 msec is usual. Few cilia will show both a faster beat and a more 

 asymmetrical beat than Stentor, so that we should not expect 

 excitation intervals of less than about 1 msec, although those of 

 some slowly beating cilia may be much longer. 



The mechanism of propagation of the contraction wave along 

 the fibrils is less amenable to this sort of approach, but some 

 suggestions made by Harris (1961) make a good start. The rate 

 of propagation of contraction under normal conditions seems to 

 lie in the range between 100 and 1000 /x/sec (see Table 12). 

 As Harris pointed out, such a speed is far too low to be caused 

 by the purely mechanical means concerned in the propagation 



