W 
served by the GILLISS in the intrusion were 11.2°C and 33.2°/oo, re- 
spectively. A plot of surface salinity (figure 8) shows excellent 
agreement with the temperature pattern derived from the ART flight of 
12 May in that the 33.5°/oo isohaline is nearly coincident with the 
13°C isotherm delineating the cold intrusion. An STD drop (point A 
on figure 8) near the cold filament shows a surface layer having a 
thickness of approximately 50 meters (figure 9). A second STD drop 
(point B on figure 8), taken in the Gulf Stream approximately 37 kilo- 
meters south of point A, is provided for comparative purposes. Note 
that below 250 meters the observed Slope Water is indistinguishable 
from Gulf Stream water below 440 meters on the T-S diagram. 
A cold intrusion (such as that observed on 12 May) with parabolic 
cross section, surface width of 4 kilometers, and depth of 50 meters 
would require a drift of approximately 10 cm/sec to be within the cri- 
teria set by Stommel for maintenance of the cold filament. Aithough 
no current measurements were made during the present survey, measure- 
ments by Howe (1962) show that drift of 10 cm/sec is well within 
reason. Failure to observe the intrusion on three of the twelve flights 
comprising the survey is in agreement with the discontinuous nature of 
the cold filament. Nonobservance of either the intrusion or the fila- 
ment by ART at the surface does not preclude Ene presence of either or 
both immediately below the surface. 
Northern Edge Displacement 
Considerable variation in the position and orientation of the Gulf 
Stream northern edge was observed throughout the survey. Variations 
appeared to be of two types: (1) relatively small, short-term dis- 
placement of the northern edge about the mean position and within the 
two-standard-deviation envelope of the northern edge computed by Pickett 
(1968) and (2) displacement of the northern edge outside Pickett's enve- 
lope. Maximum displacement of the observed features is unknown, since 
it is unlikely that the observations coincided with the time of maxi- 
mum displacement. 
Two relatively small-scale displacements are worth noting. The 
first of these occurred during the 3-day period between the flights of 
17 and 20 December, when a 180-kilometer section of the northern edge 
moved as much as 40 kilometers to the northwest (figure 10). It is 
also of interest to note that downstream the northern edge on 20 De- 
cember was east of its position on 17 December, thus suggesting 
clockwise rotation of this gradient. 
The second relatively small-scale displacement shows a change in 
angular orientation of the northern edge downstream of a warm eddy lo- 
cated at approximately 36°00'N,74°30'W (figure 11). On 12 Mav 1969, the 
orientation of the northern edge east of 74°00'W was 079°T. Four davs 
later, on 16 May, the orientation had changed to 048°T. The northern 
edge was displaced nearly 30 kilometers from its previous position at 
6 
