The very small jack mackerel feed pri- 

 marily on minute crustaceans; the food particle 

 size ingested by mackerel of about 3 mnn. 

 length ranges from 0.04 to 0.18 mm. diameter 

 and by mackerel of about 9.5 mm. length from 

 0.10 to 0.56 mm. (Anonymous, 1953). 



The size of larvae (based on preserved 

 material which may shrink as much as 20 

 percent) at the end of the yolk-sac stage is 

 as follows (Ahlstrom and Ball, 1954): 



Farris (1959) gave the following data on larval 

 growth for live larvae that were not given any 

 nourishment other than that in the yolk; the 

 yolk sacs were absorbed by the sixth day. 



On the basis of the above and other data 

 on larvae Farris (1959, 1960, 1961) ob- 

 tained growth and survival curves for 

 larval jack mackerel. His survival curves 

 are based on the questionable growth curves 

 obtained fronn measurements of starving 

 larvae, and therefore any advantages of 

 his curve of survival with age over the 

 curve of survival with length are lost. 

 The length frequency distributions of larvae 

 taken in 1952 through 1957 (table 1) may be 

 used as length- survival curves if the numbers 

 of larvae in the 0.50 mm. intervals (2.00 to 

 5.00 mm.) are doubled. If the 6-year totals so 

 adjusted are plotted on semilog paper it is 

 apparent that the length intervals at 3.50 

 through 7.75 have an excellent straight-line 

 relationship to numbers of larvae. That is, 

 if Y_ = numbers of larvae, and X_ = length in 

 mm, log X ~ a-0.36428X or X = antilog a 

 (O.43224E.) or survival is 43 percent for each 

 millimeter of length increase from 3.50 to 

 7.75 mm. The value of "a" nnay be adjusted 

 so that a predetermined number of larvae at 

 hatching (2.00 mm.) nnay be used as a starting 

 point. In Figure 5 the observed numbers of 

 larvae were multiplied by 9.5427 to obtain a 

 computed value at 2.00 mm. of 1 million larvae. 

 The relation between length and numbers of 

 larvae is log Y 5.74889 - 0.364282^. The 

 fact that larvae smaller than 3.5 mm. fall 

 progressively farther below the extrapolated 

 line is explained by the incomplete retention 

 of very small larvae by the plankton nets. 

 This selection bias is discussed in detail by 

 Farris (1961). The data indicate that, unless 

 the mortality of these very small larvae is 

 different from that of the larger ones, the 

 nets retain about 15 percent of 2.0-mm.larvae, 

 47 percent of 2.5-mm. larvae and 86 percent 

 of 3. 0-mm. larvae. 



