indefinitely by the survivors of the tagging ordeal, and on the further 

 assumption that the probability'' "^f finding the tags is about the same in 

 each of several years after tagging_, the rate of decline in the numbers 

 of tags returned should correspond to the rate of decline in the numbers 

 of fish in the porjulation represented by the tagging sample. 



Figure 11 represents the logarithms of the relative numbers of tags 

 returned in each year for each experiment, compared with lines correspond- 

 ing to ^0% and 7^% rates of decline. It will be noted that most of the 

 points representing the southern experiments fall within these lines, while 

 only those from the Montauk experiment lie above the line corresponding to 

 a S0% mortality rate. kTiether this difference is due to lower fishing in- 

 tensity or to lower natural mortality rates in the northern area is not 

 known. 



Evidence from the Scales 



Scales maybe used to distinguish races or to trace migrations if the 

 portions of the scales formed while fish are present in any locality are 

 sufficiently different from the corresponding portions of scales formed in 

 other localities to be recognized subsequently. 



Gilbert, 1919? demonstrated the existence of distinct races anong the 

 sockeye salmon ( Oncorhynchus nerka ) spawning in the Fraser River system by 

 means of differences in the stream growth portion of the -scales of salmon 

 spawning in various tributaries. Lea, 1919, obs'-'rved that the annuli ai'e 

 more sharply defined and the growth zones narrower on the scales of young 

 herring from northern Norway than on those of herring from the southern 

 part of the West coast, Runnstrom, 1936, applied these observations to 

 detailed studies of the subsequent migrations of .young herring originating 

 in these localities. From the rriarked differences in numbers of circuli in 

 the first growth zone of the scales of cod living north of Cape Cod, 

 Schroeder, 1930, concluded that he could distinguish the stocks living in 

 these localities. 



In the present study^ it was desired to extimate the proportions in 

 which weakfish originating in various localities are mixed in the adult 

 stocks. This was done by comparing frequency distributions of measure- 

 ments of the mean spacing b etween a selected group of circuli of the 

 first growth zone of the scales of adult weakfish with the corresponding 

 distributions from the scales of 0-group weakfish from three areas; New 

 York-New Jersey (hereinafter designated as "Northern") , Virginia and North 

 Carolina. 



Ten marginal circuli of the lateral field (Figures 12 and 13) were 

 chosen for measurement. The scales were examined directly with a binocu- 

 lar dissecting microj^cope and measurements were made with a comparator 

 in units of l/2liOO inch. Marginal circuli were chosen b ecause the spacing 



6h 



