transect along lat. 27°N. This value was then 

 multiplied by the length of the transect to give a 

 net quarterly meridional transport through the 

 transect. The hypothetical transect extends 

 eastward from the Bahama Islands, 668 km, to the 

 same longitude as the eastern end of Standard 

 Section A-7 (about long. 70°12'W). Thus it crosses 

 the same portion of the Antilles Current as that 

 cut by Standard Section A-7, but about 180 km 

 upstream of it. Therefore, meridional Ekman 

 transports computed for the transect along lat. 

 27°N can be compared with measured geostrophic 

 transports through A-7. Although the effects of 

 lateral boundaries were not considered, the piling 

 up of water against the Bahama Banks would 

 result in a southeastward geostrophic flow, 

 further substantiating the result of this report. 



The results of these computations, for this 

 hypothetical transect, show a large range of net 

 quarterly meridional Ekman transport values, 

 from 60xl0 3 m 3 /s northward to 20xl0 3 m 3 /s 

 southward with an overall mean of net transports, 

 over 28 yr, of 15±2xl0 3 m 3 /s northward (the 

 range gives the limits of the 95% confidence level) 

 and an SD of 11 x 10 3 m 3 /s. When the 28 yr of net 

 meridional transports were averaged by quarters, 

 there was the appearance of distinct seasonality, 

 with the lowest average value in the first quarter 

 (January- March) amounting to 7±4xl0 3 m 3 /s 

 northward with an SD of 12xl0 3 m 3 /s. The 

 transport increased in the second (April-June) and 

 third (July-September) quarters to 15±3 and 

 17 ±2 x 10 3 m 3 /s northward with respective SD's of 

 9 and 6xl0 3 m 3 /s. The fourth quarter (October- 

 December) had the highest value of 23±4xl0 3 

 m 3 /s northward, with an SD of 12 x 10 3 m 3 /s. These 

 values for the Ekman transport are three orders of 

 magnitude too small to account for the transport 

 increase in the Gulf Stream. Thus locally induced 

 Ekman drift can be ruled out as a significant 

 contributor. 



There still is a possibility that an Antilles 

 Current could account for the observed increase in 

 transport of the Gulf Stream. If a strong, narrow 

 band of the current hugged the eastern edge of the 

 Bahama Banks and joined the Gulf Stream before 

 it crossed Standard Section A-7 (Figure 1), it 

 would have escaped detection in Ingham's (1975) 

 analysis. The existence of such an intense current 

 would contradict Knauss' (1969) observation that 

 the transport increase in the Gulf Stream takes 

 place gradually from the Florida Straits to Cape 

 Hatteras, with no large increase in transport 



(>2xl0 6 m 3 /s) south of lat. 32°N and the sugges- 

 tion by Worthington (in press) and Sturges (1968) 

 that the increase in transport of the Gulf Stream 

 takes place over its entire length and at all levels. 

 Nevertheless a study in preparation by R. Yager 

 (pers. commun.) using direct transport measure- 

 ments appears to show a narrow (80 km), intense 

 (12xl0 6 m 3 /s) current to the northwest hugging 

 the east side of the Bahama Banks. 



A measure of the significance of Ekman 

 transport in moving the larvae of pelagic fishes 

 northward to the Gulf Stream can be obtained by 

 deriving a rough estimate of the average speed of 

 neutrally buoyant objects in the wind-driven 

 layer. For this the average northward transport is 

 divided by the area of the cross-section through 

 which the flow is occurring (depth of layer x 

 length of section). Using the familiar empirical 

 relationship, 



D = 



7.6W 

 Vsin<£> 



(Defant 1961 Vol. 1:422), 



where D is the depth of the wind-influenced layer, 

 W is the wind speed (here the median wind speed, 

 5.5 m/s shown for lat. 25°-30°N, long. 70°-75°W in 

 the U.S. Naval Oceanographic Office atlas 1963), 

 and is the latitude, we obtain an estimate of the 

 average depth of the wind-influenced layer to be 

 about 60 m. From the depth (60 m), the length of 

 the section (668 km), and the net transport 

 computed earlier (15±2xl0 3 m 3 /s), we obtain an 

 estimate of the average northward velocity of 

 larvae to be 0.04 cm/s. It is apparent that this 

 velocity, which translates to 0.03 km/day, is 

 considerably smaller than the geostrophic veloci- 

 ties through lat. 28°35'N reported by Ingham 

 (1975) which generally ranged from 5 to 40 cm/s 

 either northward or southward. 



The vertical distribution of ichthyoplankton 

 could have a considerable effect on their transport 

 by wind-driven currents; however, their vertical 

 distribution is not well known. If, in order to ob- 

 tain a maximum possible velocity, we assume that 

 the larvae remain in the upper meter or so of the 

 wind-driven layer instead of spending time at 

 various depths throughout it, then their wind- 

 driven drift speed would be considerably greater 

 than the 0.04 cm/s average. Using the relationship 



V n 



Vsinc^ 



(Defant 1961 Vol. 1:418), 



224 



