222 FREDERICK G. E. PAUTARD 



specimens, the rhythmic contraction and relaxation might be the 

 result of those critical conditions necessary to keep the "contractile" 

 protein moving between the two extremes. The electrolyte concen- 

 tration is important for this effect and divalent ions may be essential 

 in determining how the system will behave (see, for example, Hoff- 

 mann-Berling, 1958). Oscillation seems to depend on conditions tend- 

 ing to produce relaxation, so that the local environment is impor- 

 tant. In the intact flagellum, this is best illustrated by the experiments 

 with enfeebled sperms, where the first condition essential to a revival 

 of activity is a swelling or dispersion of the milt, reflecting some 

 analogous change in the cells. 



The contraction of flagellar gels at low ATP concentrations in the 

 presence of KC1 and the extension of normal and contracted gels at 

 higher ATP concentrations in the absence of KC1, coupled with the 

 effects of pH in extending and contracting the gels from flagella and 

 muscle, suggest that these factors might be responsible for oscillatory 

 activity by cancellation of opposing halves of the cycle. By adjusting 

 the concentration of ATP (or possibly of electrolyte), for instance, a 

 local environment relaxing the system in circumstances where a local 

 enzyme will also contract the system, may be created. In this case 

 the system is a network of filaments decided by the geometry of 

 the participating proteins (and possibly lipids). This reactive network 

 is marshalled and extended, perhaps, by the overall nucleotide/ 

 electrolyte concentrations at increasing pH, but since this raises the 

 rate of nucleotide splitting (Tibbs, 1959, finds an optimum pH of 

 8.2 for Mg++ activation in whole perch sperm flagella), the concen- 

 tration of ATP and the attendant pH fall at tJie site of splittifig and 

 cause the network to contract. The loss of substrate and unfavorable 

 local pH will then tend to depress or halt enzyme activity so that the 

 local environment can once again take over and cause the network 

 to reextend to start the next cycle. During the contraction phase, the 

 network may behave as a pump, possibly even as a molecular pump, 

 altering the local hydraulic conditions, ejecting reaction products, 

 and changing the surrounding concentrations of salts, etc. A tentative 

 arrangement along these lines is illustrated in Fig. 13. 



In this scheme, it is not necessary to postulate specialized relaxing 

 mechanisms to reverse the contraction phase, as is the case in muscle. 

 Instead, as long as the system allows free entry and exit of participat- 



