PHYSICAL OCEANOGRAPHY OP THE GULF OP MAINE 
561 
1920 (station 20075); but this seems to have constituted its western boundary dur- 
ing the next four weeks, because the whole column warmed by about 1° on Ger- 
man Bank and near the Cape between March 23 and April 15 (stations 20085 and 
20103, 20084 and 20104), instead of chilling, or at least remaining stationary in tem- 
perature, as would have happened with any considerable flow of 0° to 1° water from 
the east. Nor did any extension of icy water develop to the southwestward along 
the offshore banks or continental slope during the interval. 
Fig. 30. — Mean air temperature (solid curve) and water temperature (broken curve) in Boothbay Harbor, Me., for 
10-day intervals from July 1, 1919, to June 30, 1920 
The greatest inflow of this cold water into the gulf may therefore be expected 
between the last week of March and the middle of April in “early” years, but not 
until the last of April or first part of May in “late” years. In spite of this annual 
variation in date, the close agreement between the late April-early May tempera- 
tures of 1915 and 1919 in the region most affected by it, and the uniformity in tem- 
perature in the eastern side of the gulf summer after summer, enlarged on below 
