IJo 



L. B. SLOBODKIN 



scatter (Fig. 4). If we assume that the proportion of the food provided that 

 is consumed by a population is given by 



Pjr observed 



Pjr calculated from equation (i)' 



then the scatter in the relation of yield per unit food to jF is greatly reduced 

 (Fig. 5). 



YIELD 



( Animals+ eggs/corrected food level /4 day 



ADULT REMOVAL 



Fig. 5. The relation between yield per unit food consumed per four days and F when adult 

 animals are selectively taken, on the assumption that deviation from equation (i) is due 

 to failure to consume all of the food provided (D. pukx). 



(b)Hydrids 



Thirty-two populations o£ Hydra oligactis were maintained, sixteen by each of 

 two technicians. Each block of sixteen included one population at each of 

 four food levels and four fishing levels. The food measuring procedure 

 apparently differed somewhat between technicians so that there was a 

 consistently higher population and yield per unit food from one set of 

 populations. 



When the mean of the eight Pjr estimates at each level of F is plotted 

 against F (Fig. 6), equation (i) is found to hold. 



