176 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



Table 6. — Deviations of individual cruise frequencies of lengths of larvae and postlarvae from the average 

 frequency ' of the 9 cruises of the season of 1932 



' Deviations were taken from the theoretical rather than observed values. The theoretical values were derived from the ob- 

 served values by fitting straight lines to the points resulting from the plot of logarithm of numbers against logarithm of lengths in 

 Sg. 6. 



i From 3 to 12 mm., inclusive, the average was of the first 7 cruises; from 13 to 51 mm., inclusive, it was of 9 cruises. 



' 10 was added to the logarithm of each number in order to simplify notation in the case of decimal numbers. 



There is, in addition, external evidence that the chosen series of homologies is 

 correct and the alternate series incorrect. 



The geographic distribution of successive stages needed to fit the alternate 

 series would not be hi hannony with &ny possible system of drifts. The 3- and 4-mm. 

 larvae of cruise IV were off Long Island and the 6- to 8-mm. larvae of cruise V were 

 mainly in the offing of the southern coast of New Jersey by the next cruise. To 

 assume that these were homologous would require drifting at an average rate of 25 

 miles per day, which is far too fast for non-tidal currents in this area, comparing 

 rather to such swift ocean currents as the Gulf Stream (Iselin, 1936, p. 43). On the 

 other hand, the system of homologies indicated by the letters in figure 7 requires no 

 fantastic assumptions as to drift. In fact, it will be shown below (p. 183) that the 

 movements of larvae designated by this system of homologies follow a pattern closely 

 and definitely related to wind-impelled drifts. 



Furthermore, the growth rate of the larvae that would be indicated by the 

 alternate series is not consistent with the lengths of the smallest post-planktonic 

 stages. The range in size and the modal lengths of small post-planktonic mackerel 

 taken in July and August of certain years have been indicated in figure 8. Unfor- 

 tunately, the earliest available sample of such material in the 1932 measurements was 

 drawn August 30, nearly 50 days after the latest tow net material. It lies close to 

 the projected S-S and T-T lines of the chosen homologies and far from the projected 

 line that would result from the alternative homologies. That this does not result by 

 coincidence from altered growth rates intervening between cruise material and post- 

 planktonic material is shown by the range and modal sizes from earlier dates in 1926 



