The concentration of fishing effort depends on time, but if we 

 assume that for a given hour (that is to say^ in any L±±) the fishing 

 intensity is independent of time [We are effectively doing this when we 

 choose our cell width to be one hour in the complete census » ], and Li is 

 sufficiently large so that (on the average) a fisherman is equally likely 

 to fish at any location, then (8) becomes 



, area abde , E(t) ,. 



^ area acdf T+E(t)* ^^ ^ 



and it should be mentioned that for greatest reliability, E(t) must be 

 small compared to To (This should be considered when selecting the value 

 of T for the experimento ) These various assumptions are equivalent to 

 saying that the distribution of dots within acdf is random j ioe. independ- 

 ent of time and location o 



Introducing the approximation E(t) = Hj. = H/R, equation (9) becomes 

 H/R H 



T+H^i TR+H 



(10) 



Since (8) and (10) are equivalent, subject to our approximation, we can 

 equate them to obtain the estimate for fii, which is 



f - ^ii « TR-t-H 

 ii p H 



In terms of this expression, we can now rewrite equations (1), (2), 

 (3), (6), and (7) as 



Nh - ^ (^) ^^ (11) 



(12) 



. d ^ H ^ ^Fi H ^h (13; 



2 



32(N,) - ^ ^ (Si3)2 Z(^f Pd2 - ^] (15) 



These equations, (11), ooo, (1$), are our desired estimators. 



Ul. Interior — Duplicating Section 



Washington, D, C. 



^c 



