EXPLORATIONS, WESTERN ATLANTIC, STEAMER BACHE, 1914. 51 
the 100-meter contour being only about 2° (p. 60), and that the 
temperature rose, passing offshore, from about 6°—7° near the land to 
10°-12° over the continental slope, just as in 1914 (fig. 50); but the 
coast water as a whole was 1°-2° warmer at corresponding localities 
and depths in 1916 than in 1914.¢ Unfortunately the Roosevelt lines 
did not run offshore far enough to meet the warm “‘ Gulf Stream’”’ water. 
The salinities for the two years likewise agree, in so far as they rise 
from the land seaward (fig. 50), and in the flooding of the surface 
next to the land with water fresher than 30°/,,.. But in 1916 the water 
over the shelf between the 20 and 100 meter contours was prac- 
tically uniform from surface to bottom, and the coast water as a 
whole was slightly salter than in 1914. 
A difference far more important, if anything more than apparent, 
is that the profiles for 1916 (fig. 51, 52) do not show anything com- 
parable to the downpour outside the slope, so unmistakable in 1914; 
but it is possible that something of the sort would have appeared, 
had the profiles run far enough offshore to reach the warm ocean 
water, for the curves for 35°/,. and 35.2°/,. salinity strongly suggest 
the corresponding values for 1914 (fig. 49), so far as they go. 
Assuming the density of the ocean water to have been about the same 
in 1916 as in 1914, which was probably the case, there would have 
been the same dynamic tendency for the water over the slope to sink, 
in 1916 as in 1914, because the density was practically the same, at 
corresponding locations on the slope, for the Roosevelt as for the Bache 
stations (p. 59, 60). There is nothing in temperature to forbid it; on 
the contrary, the fact that water colder than 10° projected seaward 
from the shelf into the warmer water offshore in 1916 (fig. 51) dis- 
tinctly indicates a seaward flow at about the 50-meter level; and the 
temperature curves over the slope for the two years are readily 
reconciled with each other on the assumption that the seaward flow 
over the outer part of the shelf was localized in the upper 30 meters 
in 1914, as indeed salinity demands, whereas in 1916 it was rather 
deeper. In 1916 the slope, at 150-250 meters, was washed by water 
of 12°, a typical warm belt of the sort we are familiar with further 
north in summer (Bigelow, 1915), whereas in 1914 there was no bottom 
water warmer than 10° along this line. But as winter cooling seems to 
have progressed further by the end of January in 1914 than in 1916, 
this difference is, to all intents and purposes, a seasonal one. 
The salinity of the downward flowing tongue of January, 1914 
(34-34.5°/,.), together with its comparatively low temperature, 
identifies it as the mixed water resulting from the contact of ocean 
with coast water. This contact, as is well known, takes place all 
along the continental] slope as far north as the Grand Banks of New 
aThe minimum temperature was lower in 1916 (station 8451, 5.8°) than in 1913 (station 10157, 6.2°); 
but this difference may be due to different geographic locations. 
