108 Observations made in 1905 



in 1904 (compare Stations *2 and 3, 1905, with Stations 56 and 58, 

 1904). Again, the salinity of Section I. indicates that the Mediterranean 

 water, represented by the layer of maximum salinity at a depth of 

 about 600 fathoms, had moved mostly northwards, forming a stratum 

 of over 36 per mille about 500 fathoms thick at Station 5, and over 

 400 fathoms thick at Station 3, but thinning to about 100 fathoms at 

 Station 4, off the coast of Spain. The thinning at Station 4 may be 

 due to downward movement from the surface of water drifted against 

 the land. 



The supposition that the movement of Mediterranean water was 

 more northerly during or just before the second series of observations 

 than in the first, is confirmed by a comparison of the sections east- 

 ward from Madeira to Gibraltar. Taking the isohaline of 36 per 

 mille again, we find its western extension at Station 34 in 1904 and 

 Station 6 in 1905, a difference of about 140 miles ; the layer of water 

 over 36 is thinner, and the layer under 36 being correspondingly 

 thicker. 



We may, therefore, conclude that at different dates the extension 

 of water from the Mediterranean varies very considerably northwards 

 and westwards, and possibly also southwards, though as to this we are 

 as yet without information. Whether the variations are periodic or 

 irregular, or (as seems likely) both, we do not know. The causes of 

 these variations must necessarily be extremely complex, as they are 

 probably due partly to far-reaching changes in the movements of the 

 Atlantic waters, and partly also, though in a less degree, to fluctuations 

 in the volume and salinity of the water pouring out of the Mediter- 

 ranean. It seems certain that the northward movement of the 

 Mediterranean water may extend at least as far as the entrance to the 

 English Channel, and it is obviously of the first importance to 

 determine how often, and under what conditions, this takes place. 

 The conditions which give rise to the formation of Rennell's Current 

 may induce something of the nature of upwelling of waters west and 

 south-west of Ushant, and such upwelling might bring Mediterranean 

 water to the surface and into the Channel. 



