916 BULLETIN OF THE BUREAU OF FISHERIES 



tendency of this water to hug the coast, not to fan out over the basin. At certain 

 points along the coast local fishermen have long been aware that this results in a 

 considerable southwesterly drift parallel with the coast, so much so that it is locally 

 named the "spring current." The progressive development of a coastwise band of 

 low salinity, which results from this event, is well illustrated by the successive surface 

 charts for March and April, 1920 (figs. 91 and 101). Such distribution as appears 

 on the latter and on the corresponding chart for May (fig. 120) could persist only 

 with the water of the coastwise belt setting parallel to the general trend of the 

 northern and western coast lines of the gulf, as already explained (p. 910). In the 

 same way the expansion of water less saline than 32 per mille southward from the 

 northern margin of the gulf, along its western shore, in a narrow band past Cape 

 Ann to Massachusetts Bay, from March to April, and so out past Cape Cod toward 

 Georges Bank by May (fig. 120), is unmistakable evidence of a general set of the 

 surface water around the coast line along this same route. 



The evidence afforded by salinity is therefore clear to the effect that when the 

 outpouring of land water is at its maximum in spring it parallels the land, with a 

 dominant flow alongshore from east and northeast to southwest and south, instead 

 of spreading seaward, as happens off river mouths in many parts of the world. In 

 other words, when the velocity of the left-hand side of the Gulf of Maine eddy is 

 greatest it hugs the shore closest. The abrupt transition from surface salinity lower 

 than 30 per mille to higher than 32 per mille, recorded 15 to 20 miles out from the 

 western sector of the coast line between Cape Elizabeth and Cape Ann in May, 1915, 

 gives a rough indication of the breadth of the zone along which the combined dis- 

 charges from the Kennebec, Saco, and Merrimac Kivers are carried when the latter 

 are in flood; and some indication that the main axis of this "spring current" is directed 

 southward across the mouth of Massachusetts Bay, with some tendency to veer 

 westward aroimd the coast line of the latter after it passes Cape Ann, is traceable 

 on the surface charts for April and May, 1925 (figs. 102 and 119), as noted above 

 (p. 743). In this respect salinities and the drifts of bottles set out in Ipswich Bay 

 (p. 890) prove mutually corroborative. 



The charts of surface salinity for late summer for the several years, combined 

 with the bottle drifts, suggest that the northern and western sides of the dominant 

 eddy may be expected to trend more out from the land as the summer advances; 

 but the isohalines point to considerable differences in this respect in different years, 

 as just described (p. 770). 



The chief line of dispersal for the discharge from the St. John River is located as 

 tending toward the southwest past the eastern side of Grand Manan, by the sudden 

 freshening of the surface recorded by Mavor (1923) at Prince station 3 from April 

 to May in 1917 (p. 808; fig. 165), agreeing, again, with the routes probably followed 

 by the bottles that drifted out of the bay in 1919 (p. 870); but the increase in 

 salinity that takes place at this location from May to June and July is evidence 

 equally positive that the velocity of this drift is at its maximum for only a few 

 weeks (perhaps only a few days), though some movement of the surface water 

 probably takes place in this direction throughout the year (p. 917). 



