96 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 
the cold wall on the American side of the Atlantic to be able to trace 
with security the origin of its waters.” Quite a different explanation 
for the cold wall is proposed by Tizard, (1907, p. 343) who believes 
that the chief factor in forming the cold coast water is the discharge 
of fresh water from the rivers along the American coast, by which 
means large quantities of cold fresh water and fresh ice are emptied 
over the coastal slope. And he argues that neither upwelling of 
oceanic bottom water, nor the Labrador Current, has anything to do 
with the formation of the cold wall. 
The partial isolation of the Gulf of Maine from oceanic waters by 
the sill formed by George’s and Brown’s Banks, makes it possible that 
its cold waters need a different explanation from those of the “cold 
wall” west of Cape Cod; and the discussion of the latter is best post- 
poned until we have a better knowledge of their salinity. But so far 
as the Gulf is concerned, we can safely say that the low salinities in 
July and August certainly show that its waters are not predominantly 
Atlantic abyssal water welling up over the continental slope, because 
the salinity of the bottom water over most of the North Atlantic is 
about 34.9 (Murray and Hort, 1912). 
The same index, salinity, shows that Tizard has suggested a factor 
of real importance, for besides the fresh water emptied into the Gulf 
of Maine annually by its rivers (p. 90) there is also the annual rainfall 
of about 40 inches, a total annual increment of fresh water, which 
would make a layer more than a fathom thick over the entire Gulf. 
To offset this, there is the annual evaporation; and while this is not 
exactly known for any off-shore station in the Gulf, conditions on the 
neighboring coasts indicate that it is probably less than the rainfall. 
Rainfall and inflow from rivers combined are likewise considerably 
in excess of the annual evaporation all along the coast of Nova Scotia 
where the salinity, according to both Dickson, (1901) and Schott, 
(1902) is 32%p or less. 
The Gulf of St. Lawrence, has, of course, been mentioned by pre- 
vious authors as a source of fresh water, but its importance must be 
greater than has been usually recognized, because of the enormous 
extent of its watershed, including the St. Maurice, Saguenay, Humber, 
and other large rivers, besides the St. Lawrence itself. Its rainfall, 
too, exceeds evaporation. The little that is known about the currents 
in its two mouths (Dawson, 1910) shows that its main outlet must be 
through Cabot Straits, as Schott represents it in his chart of ocean 
currents, (1902, pl. 39) not through the Straits of Belle Isle. The 
comparatively fresh St. Lawrence water is continuous with the water 
