632 
BULLETIN OF THE BUEEAU OF FISHERIES 
close to the western shores. The salinity of the Nova Scotia current being consid- 
erably lower than that of the water it meets in the Gulf of Maine (p. 727), its presence 
causes low salinity as well as low temperature such, indeed, as prevailed at our few 
gulf stations for 1916. Salinity, however, is not a safe criterion for northern water 
in the western side of the gulf, because it is also dependent on the amount of run-off 
from the rivers, which was greater during the spring of 1916 (p. 837) than usual. 
No serial observations were taken in the open gulf during the summers of 1917 
to 1919, but Mavor’s (1923) data for the Bay of Fundy classify 1917 and 1919 as 
normal seasons. Brooks (1920), however, points out that 1920 continued a “cold" 
year in the gulf through the summer, by the testimony of bathers along New Eng- 
Temperature, Centigrade 
Fig. 68.— Vertical distribution of temperature off Cape Elizabeth on August 15, 1913 (A, station 10104), 
and on August 7, 1923 ( B, latitude 43° 18', longitude 69° 44') 
land beaches. This was followed by a summer of at least average warmth in 
Massachusetts Bay, and probably over the gulf as a whole, in 1922 (p. 995). By 
contrast the summer of 1923, like that of 1916, was unusually cold in the deeper 
waters following a severe winter, with unusually heavy snowfall, and a tardy spring. 
Surface readings would not have suggested this more than a mile or two out from 
the land anywhere in the western side of the gulf. In fact, the coast sector between 
Cape Ann and Penobscot Bay was actually a degree or two warmer on the surface 
in 1923 than in 1912 at the end of the first week of August, as illustrated by the 
curves for 16° and 18° temperature on the charts for the two years (fig. 47), with 
readings of 16° and upwards right into the land off Cape Elizabeth in 1923, where 
we have usually found the coast skirted by a belt 1° to 3° cooler (p. 588). 
