480 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



coordinates of the specimens larger than 95 mm. 

 fall along the extension of the calculated regres- 

 sion line, indicating that the head growth may 

 continue at the same proportional rate. 



Eye . — The regression of eye diameter on stand- 

 ard length is shown in figure 58 and table 11. A 

 line fitted to this regression for specimens from 

 6.0 to 83.5 mm. standard length shows a propor- 

 tional rate of increase for the two variates from 

 about 12 to 83.5 mm. (0.095-mm. increase in eye 

 diameter per 1.0-mm. increase in standard length). 

 The alinement of the two smallest specimens sug- 

 gests that a faster eye growth rate may prevail 

 for specimens below approximately 12 mm. The 

 coordinates of the specimens larger than 83.5 mm. 

 fall below the extension of the calculated regres- 

 sion line, indicating a decrease in eye growth rate 

 above this size. 



Snout. — The regression of snout length on 

 standard length is shown in figure 58 and table 

 11. A line fitted to this regression for specimens 

 from 6.0 to 95.0 mm. standard length shows a 

 proportional rate of increase for the two variates 

 within this size range (0.11-mm. increase in snout 

 length per 1.0-mm. increase in standard length). 

 The alinement of coordinates of specimens larger 

 than 95 mm. above the extension of the calcu- 

 lated regression line suggests that an increase 

 occurs in snout growth rate at about that size. 



Gill rakers. — Lower limb, 18 to 21 ; upper limb, 

 6 to 9; total, 25 to 29 (table 13). About 80 per- 



Table 13. — Caranx bartholomaei: Correlation of numbers 

 of lower-limb to upper-limb gill rakers of 84 specimens 



(The upper number in each block is the count obtained for that combination, 

 and the number in parentheses below is the approximate percentage of 

 that count in the total sample] 



OS 



w 



< 



a 



5 

 m 



« 



04 



o. 

 P 



LOWER-LIMB GILL RAKERS 

 18 19 20 21 



cent of the specimens examined above 20 mm. 

 had one or more rudimentary gill rakers at the 

 origin of the lower limb; one was present on 

 specimens up to 131 mm., 2 at 151 mm. and 176 

 mm., 4 at 182 mm., and 5 at 243 mm. and 315 



mm. About 75 percent of the specimens above 

 20 mm. had one or more shorter or rudimentary 

 rakers at the origin of the upper limb; one was 

 present on specimens up to 151 mm., 2 at 176 

 mm. and 182 mm., 3 at 243 mm., and 6 at 315 

 mm. Ginsburg (1952: 96) reported 3 to 4 upper- 

 limb and 1 to 2 lower-limb rudiments in large 

 specimens (up to about 278 mm., converted). 

 Apparently, there is an association between in- 

 crease in body size and increase in number of gill 

 rakers which become smaller or rudimentary. 



Scutes. — Range of the mean number of scutes 

 above 100 mm. standard length: about 22 to 28 

 (fig. 59). Ginsburg (1952: 96), working with 

 specimens about 26.5 to 278 mm. (converted), 

 gave a scute range of 20 to 31. Meek and Hilde- 

 brand (1925: 352), with specimens up to about 

 120 mm. (converted), gave a range of "about 22 

 to 35." The developing scutes appear between 

 12.7 mm. and 14.3 mm. (figs. 50 and 51). The 

 first scutes to complete their development are 

 present between 22 and 23 mm. All scutes of 

 fish above 80 to 100 mm. have completed, or 

 nearly completed, their individual development 

 (fig. 59). 



Lateral line. — Range of mean lateral-line ratio 

 above. 60 mm. standard length: about 1.0 to 1.4 

 (fig. 60). The lateral-line ratio-standard length 

 relation is variable, but the value of the ratio 

 tends to increase with respect to standard length 

 up to about 50 mm., and it may decrease gradually 

 above this size. Nichols (1939: 9), with speci- 

 mens 16 to 51 mm. standard length, declared the 

 curved lateral line to be about equal to the 

 straight lateral line. This generalization is ac- 

 commodated by the perimeter of figure 60, but 

 the variation and change of this ratio with growth 

 are much greater than Nichols' statement indi- 

 cates. 



Preopercular spines. — There were too few small 

 specimens to evidence the early growth of the 

 preopercular angle spine. This spine apparently 

 begins to decrease in length at some size smaller 

 than 12 mm., and is completely covered by the 

 preopercular margin between about 32.5 and 39 

 mm. (fig. 61). The number of preopercular 

 upper- and lower-limb spines varies considerably 

 with size, but a trend of decrease in number of 

 spines with increase in body size is apparent 

 (table 1). 



Pigmentation . — There are two rows of elongated 



