York, James, Albemarle Sound, and Santee- 

 Cooper) the slope of the regression line fell 

 very close to the abscissa. This was true 

 whether the samples were just 0, I, II or any 

 combination of these. The same held true if the 

 counts were from the total number of gill rakers 

 on the first left branchial arch or just the num- 

 ber of gill rakers from the upper arm of the 

 same branchial arch. From observation of 

 table 1 it seems that the (plus) or (minus) slopes 

 are just random variations and if a large num- 

 ber of samples were taken the total population 

 would have a slope of zero. 



The above results indicate that there is 

 no change in the number of gill rakers from O 

 to I. We do not have adequate collections of III 

 and older age groups to determine if the number 

 of gill rakers change with these ages . 



Variation in number of gill rakers with 

 sex. --A study was made of striped bass, that 

 were collected from the Delaware River in 1952, 

 to determine if the number of gill rakers varied 

 with sex. Specimens were immature and ranged 

 in standard length from 72 to 198 mm . As 

 there is no known way of determining the sex 

 from external characters, a dissection was 

 made . A section of the gonad was removed and 

 a smear was examined using a compound micro- 

 scope. 



At- test was used to compare the num- 

 ber of gill rakers in the sexes (table 2) and 

 proved not to be significantly different. There- 

 fore, it is concluded thai there is no sexual 

 difference as far as gill rakers are concerned. 



Populations . 



The Chesapeake Bay . --The 1955 year 

 class and the age group are available from the 

 majority of the river systems in the Chesapeake 

 Bay region. An analysis of variance was run on 

 this group using the total number of gill rakers 

 (table 3) and the F - value was highly significant. 

 This same high significance also occurred when 

 the number of gill rakers on the upper arm of 

 the first branchial arch was used as the test 

 criterion (table 4). For these data, the analysis 

 of variance shows that more than one population 

 exists in the Chesapeake Bay. 



In order to determine where similarities 

 and differences occurred, meaningful comparisons 

 were carried out based on geographical locations. 

 Therefore, an analysis of variance using the total 

 number of gill rakers was performed on samples 

 from the Chesapeake area, excluding those from 

 the Rappahannock, York and James River Systems. 

 The resulting F - value was significant at the one 

 percent level but with a much lower magnitude 

 than when these river systems were included 

 (table 5A). 



As the mean of the Patuxent River sample 

 was the lowest value of any in the Upper Bay area 

 it was deleted in the next test. The resulting F - 

 value proved to be significant at the 5 percent 

 level (table 5B). Still another analysis of vari- 

 ance was carried out with the Wicomico River 

 sample dropped from the above grouping. The 

 mean of the latter was the highest except for that 

 of the Choptank, but due to its more southern 

 location on the eastern shore it was more mean- 

 ingful to omit it. The F - value from this test 

 was not significant (table 5C). 



A similar series of analyses of variances 

 was worked out using the number of gill rakers 

 on the upper arm of the gill arch . The F - value 

 was highly significant when the whole bay area 

 was included (table 4). When samples from the 

 Rappahannock, York and James were excluded 

 the resulting F was highly significant, but the 

 magnitude was less (table 6A). Omitting the 

 Patuxent sample from the Upper Chesapeake Bay 

 area, F became non- significant (table 68). 



T - tests were executed for the 1955 year 

 class and age based en total numbers of gill 

 rakers with the following results: A test between 

 the Rappahannock and York River System samples 

 was not significant (table 7A) . However, when 

 those of the York and Rappahannock Systems 

 were tested against the James River sample, there 

 was significance at the one percent level (table 7B) . 

 Also there was found to be no significant differ- 

 ence between upstream and downstream samples 

 from the James River (table 7C). When the 

 samples from the west shore of Chesapeake Bay 

 (excluding Rappahannock, York, and James) were 

 lumped and compared with the combined east 

 shore samples, the result was just significant at 

 the 5 percent level (table 8A) . A similar com- 



