104 Hydrographical Observations, North Atlantic, 1903 and 1904 



a similar feature occurs, although less well marked. A salinity of 

 35 - 3 per mille occurs in both cases at a depth of about 850 fathoms, 

 and this would seem to indicate the limit in depth of the surface 

 system of movements. The point is well worth further observation. 

 At depths greater than about 900 fathoms salinity seems to change 

 very little with depth — at least, clown to 2,000 fathoms ; the lowest 

 salinity in the 1904 Section I. was 35'21 in 1,600 fathoms. But north 

 of the Porcupine Bank (Section VI., 1903) the salinity falls suddenly 

 to 35 - 2 at about 800 fathoms, and then slowly to 35*08 in 1,500 fathoms. 



The observations of the Silver Belle in 1903 afford definite infor- 

 mation about the extension southward of water coming from the 

 Norwegian sea over the Wyville- Thomson ridge. The occurrence of 

 these southward movements may now be regarded as proved, and it 

 seems that they take place pretty frequently, if not every year. The 

 question of the cause of these changes is one that can best be dis- 

 cussed after the observations of the International Council have been 

 kept up for some time longer, but the work of the Silver Belle in 

 depths beyond 700 fathoms seems to support the hypothesis that we 

 are here dealing with stream currents set up by the banking of drift 

 water against a mass of land, in the same way as the Equatorial 

 currents give rise to the Gulf Stream. 



1904. — The outstanding features of Section I. — Ireland to the 

 Azores — are easily described. Temperature and salinity increase at 

 the surface from north to south ; vertically they diminish together till 

 the isothermal of 4° and the isohaline of 35 - 3 are reached at depths 

 varying from 850 to 1,000 fathoms. At these values the lines are 

 practically horizontal, and beyond 1,000 fathoms the fall of tempera- 

 ture and salinity is extremely slow. Whether the view that these 

 iso-surfaces represent the lower boundary of the surface circulation 

 be accepted or not, it is, I think, evident that they indicate a critical 

 level of some kind, and that they are practically horizontal along the 

 line under consideration. 



The distribution of temperature shows few irregularities, the chief 

 being a 'hump' of relatively cold water at Stations 6 and 7 between 



