The accumulation of river water between Cape Cod and C hesapeake Ba> 355 



averaged out of the data, it is worth noting that the smaller cocllicicnis are found 

 at the time of year when vertical stability is developing to maximum values. 



The value of horizontal mixing coefficients obtained for this area are in substantial 



agreement with previous determinations for coastal areas. SvbRDRUF' and Fli;ming 



(1941) obtained a coefficient for the California coastal region of 2 IO*cm' sec; 



and BowDEN (1950) computed coefiicicnts for the Irish Sea ranging from 0-36 to 9-0 



X lO'cm^sec. 



It has been mentioned above that inclusion of the excess spring river flow of two 

 successive years could account for the excess accumulation of river water during the 

 summer. Such an increase in river flow would increase the spring-summer coeflkients 

 of eddy diff'usion by only about 25%, and would not make them as great as those 

 observed for the summer-winter change. It must be concluded, therefore, that the 

 excess accumulation of river water in the summer is a joint efi'cct of the increased 

 river flow and the decreased turbulence. 



DISCUSSION 



The flushing times estimated from the accumulation of fresh water, and some 

 general aspects of the circulation, are approximately confirmed by the drift bottle 

 studies described by Miller (1952). Practically all of the bottles were recovered 

 south of the point of release, confirming the net drift southward parallel to the coast. 

 The proportion of returns decreased greatly with distance from shore, however, 

 and very few bottles released more than 15-20 miles from shore were returned. 

 This confirms, for the surface waters at least, the active offshore transport as well as 

 the narrow band of current parallel to the coast. 



The rate of drift of the bottles released near shore varied over wide limits. Several 

 gave velocities greater than 10 miles per day, but all of these were released either 

 just south of Delaware Bay or off Chesapeake Bay, where the net drifts appear to 

 occupy a very narrow coastal strip. Most of the bottles drifted at rates of 3-5 miles 

 per day. A bottle travelling at this rate, which did not beach en route, could traverse 

 the 350 mile stretch of coast in 70-117 days. This corresponds approximately to 

 our flushing time within the 20 fathom contour of 108-125 days. Since very few 

 bottles were recovered from greater distances offshore, it is impossible to check our 

 flushing times for the deeper areas in this way. 



The flushing times are based upon the assumption that no substantial volume of 

 fresh water is added to the area except as river flow. Miller (1952) obtained eighteen 

 returns of bottles released on or near the 20 fathom contour in our area A. Their 

 average drift was about 3-5 miles per day. The drift bottles, of course, do not give 

 an evaluation of the movement of the deeper waters, but it seems probable that the 

 surface drifts are greater than the average. Thus a maximum estimate of the advection 

 of fresh water into the area can be made assuming that the water at all depths within 

 the twenty fathom contour is passing from area A to area B at this velocity. The 

 appropriate cross-section area is 3-52 X 10" ft^ so that the total volume transport 

 would be 7-5 x 10'° ft^day. Since the fresh water content ranges from 7-98-S-40°<„ 

 the corresponding transport of fresh water would be 5-6-6-3 x 10» ft'/day. This 

 includes the river water discharged directly into area A from the rivers emptying 

 into Long Island Sound (Table II, Group A) which accounts for about one-fifth of 

 the total gaged drainage, or about 2-9 :■: 10« ft« per day. Including the remainder. 



