to the left of the slope water boundary at the 

 surface. 



The section shown in figure 10 was taken 

 about 5 hours later, and one more section of 

 the sequence was made 20 hours later (fig. 11). 

 Both sections represent a dissipation stage 'of 

 the interaction system which has been formed 

 by the process shown in figure 8. 



The thermal structure in figure 10 is some- 

 what simpler than in the previous section (fig. 

 9), but the two cold-water stripes, one near 

 the Gulf Stream boundary and the other be- 

 tween the two shallow warm-water cores, are 

 still very well marked on the surface tempera- 

 ture trace. The cold-water stripes are almost 

 eliminated in the next section (fig. 11). The 

 surface temperature is relatively smooth be- 

 tween the two steep temperature gradient 

 zones at the boundaries on both sides of the 

 intermediate zone. 



A number of the sections indicate that the 

 Gulf Stream boundary moved to the left during 

 the absence of an interaction system. To illus- 

 trate this movement, the upper 60 m. of five of 

 the sections from transect I were replotted 

 (fig. 12), and corresponding geographic points 

 were placed in approximate vertical align- 

 ment. (Full plots of these sections appear in 

 figures 3, 4, 5, 8, and 9.) These sections start 

 during one interaction cycle and end during the 

 succeeding cycle, 41/2 days later. Because the 

 first and last of these sections represent about 

 the same stage in the successive cycles, the time 

 of a cycle from one strong and well-developed 

 interaction system to another was about 41/2 

 days (108 hours). 



When the intermediate system decayed, the 

 Gulf Stream boundary moved to the left and 

 just one boundary remained. It became the 

 slope water boundary when the new Gulf 

 Stream boundary developed near lat. 36° N. The 

 slope water boundary is almost exactly at the 

 same location as it was in the previous cycle 

 (fig. 3), but the new Gulf Stream boundary is 

 about 6 miles (11.1 km.) farther to southeast 

 and the intermediate zone is wider. 



The next section across the Gulf Stream, 

 along transect I (fig. 13), was made 10 days 

 later. Oceanographic stations and BT's were 

 taken in this and all the following sections. 



There was no intermediate zone, and the left 

 boundary of the Gulf Stream corresponds to 

 the situation in section 3 (fig. .5) ; however, this 

 time the single boundary is about 24 miles 

 (44.4 km.) to the southeast. Time interval be- 

 tween sections 3 and 10 was about 278 hours. 

 If two cycles occurred during this time, the 

 average period would be 139 hours. 



Over half the cycle of a small interaction 

 system was also recorded in three sections on 

 transect I (figs. 13, 14, and 15). These figures 

 suggest that about half the cycle was com- 

 pleted in 30 hours. This timing is in reasonable 

 agreement with the 108-hour period estimated 

 above for the cycle of a larger interaction sys- 

 tem. Presumably the time of a cycle may vary 

 considerably. The life span of an interaction 

 system probably depends less on volume and 

 shape than on the energy it contains after the 

 separation stage. 



The sections along transect II (figs. 16, 17, 

 and 18) show stages in the maximum develop- 

 ment of another strong interaction system. The 

 three sections record such a small part of a 

 complete cycle, however, that the duration of 

 that cycle cannot be estimated. 



The appearance of a rather weak interaction 

 system can be detected in the sections along 

 transect III (figs. 19, 20, and 21). 



Altogether, six different interaction systems 

 were observed during the Geronimo cruise. 

 The mean cross-sectional area of the Gulf 

 Stream water occluded into them is estimated 

 to have been 1.69 x 10'' m.' 



DIRECTION AND VELOCITY OF FLOW 



Beginning at the stage of development shown 

 in figure 8, the interaction system is cut off 

 from the body of the Gulf Stream and acts in- 

 dependently. It presumably flows in the same 

 general direction as the Gulf Stream, but the 

 cold stripe that separates the interaction sys- 

 tem from the Gulf Stream may have some tend- 

 ency toward an opposite flow. This tendency is 

 probably only occasionally strong enough to 

 overcome opposing flow from both sides, in- 

 duced by the horizontal shearing stress. Verti- 

 cally, the cold stripes and the whole interaction 

 system sometimes seem to extend to depths 

 greater than 300 m. but many sections indicate 



GULF STREAM OFF CHESAPEAKE BAY 



413 



