352 



Fishery Bulletin 102(2) 



of a simulation. Then one school of an appropriate magni- 

 tude, M, was added such that the M +^F ll/ , v A„.,,, was exactly 

 equal to S. Note that under this construction the school 

 sizes in the distribution do not vary with S (except for the 

 one school of size M), whereas the frequency of schools by 

 size do. An example of the initial distribution with the use 

 of this construction is given in Table 2. 



Analytical models 



Analytical models of aggregation and recruitment are 

 presented, where the models are developed from first prin- 

 ciples and the parameters have an interpretation in the 

 physics and biology of the recruitment process. Hopefully, 

 the nature of the parameters can guide model selection, 



