OF SEA-WATER IN THE DIFFERENT PARTS OF THE OCEAN. 
241 
and St. Thomas, whose mean salinity is 35*7 per 1000 ; while two degrees more to the 
north the mean of two observations is 36*7, which seems to be the normal salinity of 
the West Indian Sea. In the Caribbean Sea, where the Magdalene river gives a new 
quantity of fresh water, the sea contains on the surface, according to one observation, 
36*104 per 1000 salt. I have unfortunately no observation from the Mexican Gulf, 
nor from the beginning of the Gulf-stream, where it leaves the Mexican Gulf, but to 
the north of the Bermudas it contains only 35*883 per 1000 salt, about the same quan- 
tity which the equatorial current contains between 20° and 30° W. longitude. From 
that place the salt of the Gulf-stream increases constantly during its course towards the 
north-east, viz. 36*105 per 1000, 36*283 per 1000. In 43° 26' N.lat. and 44° 19' W. long., 
about 16° of longitude to the east of the southern mouth of the St. Lawrence, between 
Nova Scotia and Newfoundland, it sinks suddenly to 33*854 per 1000, and rises from 
thence slowly in its course towards the east to 34*102 and 35*597, until, midway between 
Newfoundland and the south-western cape of Great Britain, it has risen to 35*896 per 
1000, a quantity of salt which diminishes very little in the whole North Atlantic Ocean 
between Scotland and Iceland. During this whole long course, from the Bay of Benin 
to Spitzbergen, this remarkable current shows a constant oscillation between the diluting 
influence of the large rivers and the evaporation occasioned by the high temperature of 
the current. 
Now we shall try to trace its further progress. I have always thought that the East 
Greenland current was of polar origin, and that it carried the waters from the large 
opening between Spitzbergen and the northernmost coast of Greenland into Davis’s 
Straits, where it turns and mixes its waters with the polar current that comes from the 
North American polar sea through Lancaster Sound, and the numerous other sounds 
that connect Baffin's Bay with the American polar sea, but I never had an opportunity 
of making comparative analyses of the water from that but seldom visited part of the 
ocean. Colonel Schaffner had the kindness on his voyage between the eastern part of 
Iceland and the south part of Greenland to take a number of samples, which I have 
analyzed, and the result of which will be found in my fourth region, the East Green- 
land current. The mean of twelve observations of water, taken for the greatest part by 
Colonel Schaffjster (three by Captain Gram), is 35*278 per 1000 salt, where one analysis 
of water taken in the ice-pack is left out, being no fair sample of sea-water from that 
region. In comparing this mean number with that of the North Atlantic Ocean (35*391), 
there will hardly be found any difference in the quantity of salt the two contain ; while 
there is a great difference between these and the real polar current of Baffin’s Bay, 
which is 33*281 per 1000, or of the Patagonian polar current (33*966). I think we may 
infer from this fact, that the East Greenland current is a returning branch of the Gulf- 
stream, and that the east coast of Greenland proportionally gives very few icebergs and 
very little glacial water to the sea. For comparison’s sake I shall mention here that the 
sea about midway between Norway and Spitzbergen contains 35*222 per 1000. I found 
the water taken on the south side of that island to contain 35*416 per 1000, while that 
