TABLE 3. — ^An example, using data on head length from the present study, showing 

 how the Quenouille T-test for the separation of regression lines of similar 

 slope can be employed for racial studies in fishery investigations 



Parameter, F (formula l) T (formula 2) T (formula 3) 



location and between between between 



date variances coefficients lines 



Gulf of Maine 



September 1955 1.386 NS 0.0166 NS 26.^62** 



New Jersey 



October 1955 



Gulf of Maine 



April 1956 1.052 NS 0.0061 NS 32.-495** 



New Jersey 



April 1956 



Formulae: (^xy) 



F (larger) _ ^y^ - x^ . ,^ ^ 



(1) r where V = sy . x'' 



V /"smaller^ n — 2 



(2) T -b^-b^ where (sdy . x^) / N^ + /Vj - 4 



"1.2 ^1.2 



,V N 



' 1 2 



(3) r =- 1 2 w 2' 



V^l ^^ ^ 1x^1 Ixhj 



where b 



sy . x^ = 



I.x^l + 2 x^2 



(Srtj.yj +X2-y2)^ 



Symbols used are: ** - significant at the 1 percent level, and 

 NS - not significant. 



do not differ significantly, but when and New Jersey ) were found to be highly 



compared with the Gulf of Maine c die c- significant (table 5). 



tion the results were highly significant 

 (table 4). 



Length of First Ventral Fin 



The T-test was also applied to this 



paranneter and the comparisons with the Comparisons were made with silver 



Gulf of Maine with the total southern hake collected in the northern area 



collection (Nantucket Shoals, New York, (Gulf of Maine) and those from Rhode 



