PHYSIOLOGY OF THE SPERMATOZOON 107 



ment than those in the general suspension, where they 

 mutually impede one another's range of movement, so 

 that it seems impossible to explain the aggregation on 

 the assumption that it is due to CO2 paralysis. This 

 is a factor in the result, for the spermatozoa that gain 

 the center of the drop are paralyzed, but paralysis 

 explains neither the initial aggregation nor its growth. 



The behavior of the spermatozoa at the margin of 

 aggregations is curiously like the thigmotactic reaction 

 previously discussed ; they move very actively in circles 

 of increasingly short diameter before coming to rest. At 

 a certain point in the CO2 gradient the circus movements 

 predominate over those of translation; thus spermatozoa 

 reaching this point in the gradient are practically im- 

 prisoned, although active, and this without reference to 

 the way in which the point in the gradient is reached. 



If the clear zone of a fresh preparation be carefully 

 examined it can be seen that the spermatozoa are 

 moving across it in streams directly toward the center; 

 none move in the opposite direction. For some minutes 

 this steady centripetal migration of the spermatozoa 

 across the clear zone may continue, but by degrees it 

 ceases, though the rotary movement of the spermatozoa 

 at the margin of the aggregation continues for a long time. 

 The persistence of the aggregation is thus due to a 

 behavior change which is the same as that given in 

 response to contact. But the cause of the aggregation 

 lies in directive movement up the CO2 gradient (chemo- 

 taxis). This interpretation supplements the account 

 given in my fifth study of fertilization (19 13). 



The sperm of Nereis exhibits similar behavior with 

 reference to other acids, thus demonstrating that the 



