812 BULLETIN OF THE BUBEAU OF FISHEEIES 



fluctuations in the salinity of the surface water at the mouth of Massachusetts Bay, 

 and so may confuse the seasonal picture. 



In 1925 the surface salinity remained close to the annual minimum at this local- 

 ity for several weeks (perhaps this is always the case) . A considerable increase was 

 then recorded (to about 32.3 per mUle; p. 756) ; but if this is an annual event (which 

 is by no means certain) it is followed by a second freshening, for the surface records 

 for this region for July and August, in the several years of record, have averaged 

 only about 31.5 per mille (or 31.3 per mille, if one station for July, 1916, be included), 

 with 32.09 per mille as the maximum. The salinity then increases slowly through 

 the autumn and early winter, as just described (p. 799). Differences in circulation 

 may bring the surface to its saltest there as early as the last of December, as seems 

 to have happened in 1920 (p. 805), or not until well into March, as in 1913 (p. 808). 

 Comparison between the graphs for the Bay of Fundy (fig. 165) and for the mouth 

 of Massachusetts Bay (fig. 166) brings out the interesting difference that while the 

 surface salinity of the former continues comparatively constant throughout the year, 

 except for the period of 4 or 5 months that covers the vernal freshening and its 

 eclipse, the salinity rises and falls over a period of 8 or 9 months off Massachusetts 

 Bay, with only the winter describable as comparatively static. 



The differences in salinity from season to season at the surface are so much wider 

 than the differences at any given season from year to year that inclusion of the lat- 

 ter does not rob the composite graph (fig. 166) of its illustrative value. Annual 

 fluctuations, however, introduce a more and more serious source of error at greater 

 and greater depths, as the effects of vernal freshening from above become less and 

 less apparent, until the former may nearly, if not quite, equal the seasonal fluctua- 

 tions at depths no greater than 40 meters. Consequently, a combination of the data 

 for different years gives a less trustworthy picture of the seasonal progression for the 

 deep water; and monthly data for any one year, which would yield such a picture, 

 are yet to be obtained. 



Nevertheless, when such data as are available are combined, by seasons, for the 

 40-meter leveP^a rather definite progression does appear, with values averaging 32.8 

 to 33.1 per mille for the cold half of the year (November through March), decreasing 

 to 32.6 per mille in April, 32.5 per mille in May, 32.3 per mille for July to October, 

 and increasing again through the early winter. While the 40-meter value was as 

 high there on June 16 and 17, 1925 (33.17 per mOle), as any recorded for February 

 or March, this is the only record for the period July to October that has been higher 

 than the mean for the year (approximately 32.6 to 32.7 per mille). On the other 

 hand, only 1 of the 10 records for the period January to March has fallen appreci- 

 ably below the annual mean. 



The salinity of the 40-meter level, therefore, may be expected to vary by about 

 0.7 per mille at the mouth of Massachusetts Bay during average years, being most 

 saline at about the same season that the surface is at its maximum (late winter) , but 

 not at its freshest until two or three months after the salinity at the surface has 

 passed its minimum (in May) and begun to increase once more. However, the 

 unusually saline state of the water in this region in June, 1925, is sufEcient evidence 



1' November to January, 6 stations; February to March, 5 stations; April to May, i stations; June, Fish Hawk cruise 14 in 

 1925; July to August, 6 stations; September to October, 2 stations for the several years. 



