PHYSICAL OCEANOGRAPHY OF THE GULF OF MAINE 
737 
from March, when the saltest water was definitely banked up against its right-hand 
wall (fig. 99), to April, when the data for stations 20107 and 20108 gave little evidence 
of this, though the salinity of the water over the slope of Georges Bank, had contin- 
ued almost unaltered. 
The course of events in the deeper strata of the gulf may then be reconstructed 
as follows for the period March to April of 1920: The presence of a much greater 
volume of water more saline than 34 per mille in April than in March proves an 
active pulse inward along the floor of the Eastern Channel, during the first part of 
the period. This indraft not only effected a considerable increase in the salinity 
of the bottom water of the basin of the gulf, but resulted in a wide expansion of 
the area occupied by water more saline than 34 per mille (cf. fig. 118 with fig. 
100), as well as raising its upper boundary closer to the surface. 
The state of the gulf in April, 1920, added to the data for the summer months, 
makes it almost certain that this 34 per mille water never overflows the coastal 
