704 BULLETIN OF THE BUREAU OF FISHERIES 



shores of Nova Scotia (< 32.2 per mille) is thus shown to be of local origin — i. e., merely 

 a part of the generally low salinitj'' of the coastwise belt, resulting from the drainage of 

 fresh water from the sundry streams that empty along that sector of the coast line. 



At the time of our spring cruise in 1920 the surface water over the eastern half 

 of Georges Bank and in the southeastern part of the basin of the gulf was more saUne 

 than 32.5 per mille, this area of high salinity indenting Y-hke into the inner parts of 

 the gulf, with its one arm extending northward along the eastern side of the basin to 

 the mouth of the Bay of Fundy and the other westward toward Cape Cod in a man- 

 ner better shown on the chart (fig. 91) than verbally. It is probable that this contrast 

 in sahnity between the western and eastern ends of Georges Bank is characteristic of 

 this season of the year. 



The distribution of saUnity on Georges and Browns Banks also makes it proba- 

 ble that the sal test sirrface water in the Eastern Channel and in the neighboring part 

 of the basin of the gulf then took the form of an isolated pool entirely cut off from 

 the still more saUne surface water (>33 per mille) of the Atlantic basin outside the 

 edge of the continent, reflecting some local stirring or upweUing of the water. 



Apparently it would not have been necessary to run out more than about 25 to 

 30 miles from the continental edge of Georges Bank in February and March to have 

 encountered surface salinities of 33 per mille and upward; but the low value (32.16 

 per mille) at our outermost station on the Shelburne profile (station 20077) suggests 

 that the isohaline for 33 per mille then departed farther and farther from the conti- 

 nental slope, passing eastward from Georges Bank, to leave a widening wedge of less 

 saline water next the edge of the continent. 



The most spectacular event in the yearly cycle of salinity of the Gulf of Maine 

 is the sudden freshening of the surface near its shores, which follows the spring 

 freshets of its rivers, an event happening earlier or later, according to the date when 

 the snow that blankets New England, New Brunswick, and Nova Scotia melts and 

 the ice in the lakes and streams goes out. In this respect the spring of 1920 was 

 late, following a severe winter. The effect of this outpouring of land water makes 

 itself evident, by lowered salinity at the surface, earlier off some parts of the coast 

 than off others. However, this regional variation does not correspond directly to 

 the latitude of the rivers concerned, because the effect of the Kennebec was made 

 evident in 1920 by surface salinity nearly 1 per mille lower close in to its mouth 

 (station 20058) than either to the westward or to the eastward of it as early as 

 March 4 (fig. 91); but any effect that the discharge from the Merrimac may have 

 had on the preexisting salinity up to that date must have been confined to the 

 immediate vicinity of its mouth, because the surface was then about the same for 

 the general sector between Cape Elizabeth and Cape Ann as for the offing of the 

 river (32.2 to 32.3 per mille). 



In 1925 (an earlier spring on land as well as in the sea) fresh water from the 

 Merrimac had developed a streak of low surface salinity (30.7 per mille) for about 6 

 miles out from the mouth of the river by March 12, with slightly higher surface 

 values (31 to 32 per mille) to the north and south {FisJi Hawk stations 20 and 28, 

 cruise 9, pp. 1009, 1010). While higher values in Massachusetts Bay (32.4 to 32.9 per 

 mille; Fish Hawk cruise 8, March 10, stations 2 to ISA; p. 1004) prove that low salin- 

 ities from this source had not yet spread southward. past Cape Ann, the freshets from 



