In the lee of the islands there was a striking resemblance between many 

 of the features of the flow pattern and the "east wind case" described by Long 

 (1952) in his discussion of the results of model studies of flow past a barrier in 

 a rotating spherical shell. This was particularly noticeable in the results of 

 cruise 1, which had the stations most closely spaced and a series close inshore. 

 Cn the equatorward side Lon^; found that the fluid in the immediate vicinity of the 

 barrier curved northward along the obstacle for about a quadrant (90°) while the 

 bulk of the fluid continued in a general westerly direction. This phenomenon could 

 account for the strong northerly flow along Hawaii during cruise 1 and the greater 

 dynamic height at station 8 during both cruises IC and 12. ^ong the boundary 

 between the bulk of the fluid that was continuing westward and the wake of the 

 barrier he found a series of horizontal clockwise eddies having their axes directed 

 easterly. Ke attached no significance to their direction and doubted that they 

 could exist in a stratified fluid such as the ocean. However, it is felt that the 

 persistence of the easterly flow down to the 300-decibar surface on cruise 1 is a 

 manifestation of this phenomenon. 



On the poleward side he found that the flow separated from the barrier 

 and moved zonally to form a series of cyclonic eddies. This would suggest that the 

 eddies found in the survey area are merely the first of a series that occur down- 

 stream from the islands. However, it must be remembered that only part of the In- 

 cident flow passes through the channels, so that a complete analogy to the single 

 barrier case is not valid. 



It has been demonstrated that a similar series of eddies v;ould result 

 from the flow through the channels. Rossby (1936) showed in his Vvake Jtresun Theory 

 that as a jet or wake, which has been discharged into a fluid at rest, ^^asses down- 

 stream it acts as a series of sinks with respect to the surrounding fluid. Large 

 quantities of fluid are removed from the surroundings, and this removal is associ- 

 ated with the creation of a countercurrent along the left side. The mass transport 

 increases duvjnstream, but is intermittently reduced through the discharge of eddies 

 along the boundary of the currents. On the right side a compensation is set up 

 which decreases in intensity and becomes negligible. Thus, the eddy west of Kawaii 

 can be attributed to the flow through Alenuihaha Channel and the one west of Oahu 

 to the flovv tl_rough Kauai Channel, ^^gain aiinur discrepancies between the selected 

 model and prototype are evident. In the first case it has already been shovvn that 

 durir.g cruise 1 part of the flow was the result of the clock-wise flow around the 

 island of Kavjaii. In the second case the magnitude and strength of the eddy, par- 

 ticularly during cruise 10, seemed iLirger than could be expected from such a 

 limited source as Kaiwi Channel. 



The variations in strength, magnitude, and position of the eddies during 

 the three cruises reflect the varying direction and strength of the incident cur- 

 rents which were inferred from the monthly wind and current data. However, the 

 stability of the eddy south of Oahu, as shown by GiUi observations made as much as 

 1 week apart and after a 90° shift in the incident winds, indicates that these 

 changes must occur very gradually. 



When the sigma-t cress sections are compared to the inorganic phosphate 

 cross sections of cruise 12 (fig. 32) to determine whether enrichment of the sur- 

 face layers by lateral mixing (along sigma-t surfaces) occurred in the eddies or 

 in any other areas, the results are insignificant compared to those reported for 

 equatorial waters by Cromwell (1951) and Austin (1954). The greatest depth from 

 which nutrients could have reached the surface by lateral mixing during any of the 

 cruises was about 100 m., i.e., at stations 12 and I4 during cruise 1 (fig. 6, 

 panels C and D) or at stations 1 and 25 during cruise 10 (fig. 16, panels A and D) . 

 Such mixing could not have produced much enrichment, because the phosphate-rich 

 layer was below 150-200 m. during cruise 12. 



The great depth of the phosphate-poor layer in the Hawaiian Island area 

 can be explained by the circulation which produced the high-salinity water that 

 lies at its lower limit. This layer is the result of subsidence and spreading out 

 of water which has had its density increased by evaporation and cooling as it moved 



17 



