462 LORANDE LOSS WOODRUFF. 



Consequently it is highly probable that a slight initial shift of 

 the definitive onset of endomixis in the various races is the 

 explanation of the nearly simultaneous appearance of the process 

 in all the races under all the conditions. 



Although the 'time-periodicity' characteristic of the species 

 has been shown by the present experiments to be practically 

 unmodifiable under the general environmental changes which 

 were employed, it has been found that the 'generation-periodic- 

 ity '- or the number of cell divisions between one occurrence of 

 endomixis and the next may be modified to a considerable 

 degree by the culture conditions which lower the division rate. 

 In other words, the rhythm appears to be more susceptible of 

 modification in regard to generations than time. As has been 

 previously noted, this is a surprising result, since a profound 

 reorganization process such as endomixis must be closely related 

 to the general metabolism of the cell and this is expressed to a 

 large extent in growth and reproduction. Further work on this 

 problem is in progress. 



Finally, the cessation of endomixis in these experiments was 

 .invariably followed, usually within a rhythm or two, by the 

 death of the culture involved. This indicates strongly, if it does 

 not prove, that a periodic occurrence of the definitive endo- 

 mictic phenomena is a sine qua non for the continued life of the 

 race a conclusion which is concordant with all previous data 

 in regard to endomixis. 



