772 BULLETIN OF THE BUREAU OF FISHERIES 



the same dates in 1912 or in 1915. Probably the correct explanation is that 1916 

 was a tardy spring, when the effect of vernal freshening from the land continued 

 evident until later in the season than usual, and when the approach of water of 

 high salinity to the continental shelf was delayed until later in the season. As a 

 result of this retardation of the vernal cycle — associated, no doubt, with the severity 

 of the preceding winter and the lateness of the spring — the salinity of the s.urface 

 was very nearly uniform on July 24, 1916, right across the whole breadth of the 

 western end of Georges Bank, where a considerable north-south gradation is to be 

 expected at that season in more normal years (fig. 136). 



Contrasting with 1916 and with 1914, the summers of 1912 and 1913 may be 

 characterized as "salt" in the western side of the gulf, with surface values averaging 

 about 0.1 to 1 per mille higher at corresponding localities and dates than in 1914 — 

 August as well as in July — but with very little difference from summer to summer 

 in the eastern side. The surface values for 1915 paralleled those for 1914 except for 

 the closer approach of oceanic water to the continental shelf off Nova Scotia, men- 

 tioned above (p. 771). 



No wide annual fluctations in salinity have been recorded for any part of the 

 gulf at a given season, or are such to be expected. 



VERTICAL DISTRIBUTION 



The salinity of the deep strata of the gulf, like that of the surface, remains 

 more nearly constant during July and August than over any period of equal duration 

 earlier in the summer or in the spring. Two stations in the basin off Cape Cod, 

 four weeks apart in 1914 (stations 10214 and 10254, July 19 and August 22), exem- 

 plify this for the western side of the gulf, the values, depth for depth, being nearly 

 alike in spite of the time interval separating them, with the one station slightly the 

 more saline at some levels, the other at other levels. 



The graph (fig. 1 38) illustrates how little variation in salinity has been recorded 

 for the deeper levels in the western side of the basin at different dates in August of 

 different years, individual stations seldom differing by more than 0.2 to 0.4 per 

 mille in either direction from the mean values of 32.6 per mille at 50 meters, 33 per 

 mUle at 100 meters, 33.4 per mille at 150 meters, 33.9 per mUle at 200 meters, and 

 about 34.1 per mUle at 250 meters. 



Except in localities where the tide runs strong enough to keep the whole column 

 of water thoroughly mixed from top to bottom, the salinity of the gulf is invariably 

 lower at the surface in summer than on the bottom, as already stated for the spring 

 months. I should emphasize, also, that the increase in salinity with depth is con- 

 tinuous, or at most is interrupted by homogeneous strata; we have never found 

 fresher water imderlying salter in the gulf. Thus, the intermediate layer of low 

 temperature, characteristic of certain summers (p. 602), is not reproduced by the 

 salinity; but the vertical distribution varies widely from place to place in the gulf, 

 a convenient division in this respect being (1) into the coastal zone, (2) into the 

 basin, and (3) into the offshore rim. 



In the western section of the coastal zone, out to the 100-meter contour, the 

 vertical increase of salinity, with increasing depth, averages much more rapid in 



