750 BULLETIN OF THE BUREAU OF FISHERIES 



If the contrast between the salinities for the early spring of 1920 and for May, 

 1915, represents the succession normal for this time of year, a very considerable 

 freshening also takes place at greater depths in the eastern side of the basin from 

 March and April to May, the graphs (figs. 114 and 124) suggesting an average 

 decrease of about 0.6 to 0.8 per mille at 100 meters and deeper. Such a reduction 

 of the sahnity back to about the March values naturally would follow any slackening 

 of the inflowing bottom current, but would be less and less apparent the farther from 

 its source of supply. A regional relationship of this sort does, in fact, result from our 

 station data, which show the salinity of the bottom water of the western side of the 

 basin only shghtly lower in May and June, 1915, than in March or April, 1920 

 (fig. 112). 



The upwelling of water more saline than 33 per mille in the western side of the 

 basin, which follows or accompanies the incorporation of river water into the one side 

 of the gulf and of the Nova Scotian current into the other, causes a much more 

 abrupt transition in salinity between coastal belt and basin at 40 meters in May 

 (fig. 125) than in April (fig. 115) ; still wider than in March, and a regional distribu- 

 tion more nearly paralleling the surface (fig. 120). The gradation from 31.7 to 31.9 

 per mille next the land to 32.8 to 33 per mille in the west-central parts of the basin, 

 shown on this May chart, is probably typical for the month, though no doubt the 

 precise spread between inshore and ofl'shore values varies somewhat from year to 

 year and would probably have proved somewhat narrower in 1925, when the 40-meter 

 values for Massachusetts Bay in May averaged slightly higher (32 to 32.6 per mille) 

 than was the case in 1915 or in 1920. 



Up to May the decrease in salinity attributable to vernal freshening is 

 confined to even a narrower coastal belt at 40 meters than at the surface, 

 hardly any change being indicated more than 10 miles out from that contour 

 line in the western side of the gulf '^ or farther south than the offing of Cape Cod, 

 where the 40-meter values were somewhat higher on May 16 to 17, 1920 (32.3 to 

 32.5 per mille at stations 20125 and 20126), than they had been a month earlier 

 (32.1 to 32.2 per mille at stations 20116 and 20117 on April 18). The salin- 

 ity at this depth was also about the same in the southwest part of the basin and on 

 Georges Bank in that May (32.5 per mille) as it had been at the end of February. 

 In spite of this apparent agreement, however, the water less saline than 33 per 

 mille must actually have increased considerably in volume in the offing of Cape Cod 

 during the interval to account for its expansion out from the bank to the seaward 

 slope of the latter, where salinity decreased by about 1 per mille at 40 meters between 

 February 22 (station 20045, about 33.8 per mille) and May 17 (station 20129, 

 about 32.9 per mille). 



It is probable that the salinity of the 40-meter level falls below 32 per mille every 

 May over a considerable area out from the Nova Scotian shore of the gulf, where 

 the Nova Scotian current then holds sway; and if 1915 was a typical spring in these 

 waters (which I see no reason to doubt) the drift of this water of low salinity from 

 its more eastern source is directed more definitely westward toward the center of the 

 gulf at this depth than it is at the surface, with less evidence of any dispersion north- 

 ward toward the Bay of Fundy (p. 745). Reduced to terms of distance, the seasonal 



" Tbis follows an extremely irregular course. . 



