was true. This is true because many of the fish caught here in the spring 

 must be enroute to northern waters where they had spent their previous 

 summers and where growth is more rapid. 



The above observations led to a hypothesis concerning the movements of 

 the fish during each year of their life: 



First to Second Autumn . Young (O-group) fish are distributed from Long 

 Island to North Carolina, but migrate to Virginia and North Carolina for 

 their first winter. In the following spring these fish (now I-group) move 

 inshore along North Carolina, most migrating to Virginia by mid-August. 



Second to Third Autumn . Yearlings move to warmer water off North 

 Carolina and Virginia to spend their second winter. In spring they return 

 inshore from North Carolina to New York, and in summer there is a general 

 northward movement. 



Third to Succeeding Autumns . The Il-group fish winter off Virginia and 

 North Carolina. Depending on their previous liistory, they migrate as fai' 

 north as New York or remain off Virginia and South Carolina. Movements are 

 repeated during each succeeding year of life. 



The hypothesis was tested by tagging experiments and analysis of scale 

 measurements : 



Weakfish of the O-group were tagged at Montauk, N. Y. in 1932. The 

 returns in 1933 consisted of l8 from southern waters and the remainder pre- 

 dominantly from Delaware Bay and southern New Jersey. In 193h most of the 

 returns were in northern New Jersey or New York. Later returns were mostly 

 from New York waters. 



A tagging experiment in lower Chesapeake Bay in 1931 indicated no 

 northward migration, while 30 percent of the returns of fish tagged off 

 Exmore, Virginia, in 1933 were from northern localities. 



Tagging experiments with I-group weakfish in Pamlico Sound, North 

 Carolina seem to indicate that the North Carolina sounds do not contribute 

 materially to northern stocks. The majority of recaptures were made in 

 Virginia and North Carolina waters. 



Declines in percentage recovery of tags from the various experiments 

 indicated that the stocks were declining at the rate of 50 to 7$ percent 

 per year. 



Frequency distributions of the mean spacing between the ten marginal 

 circuli of the lateral field of the first growth zone were constructed. 



A statistical analysis of the measurements indicated that year classes 

 should be treated separately. Comparison of the 193li O-group measurements 



78 



