CLOSED SEAS 63 



British Columbia, Chili, Australia, and the Cape). 

 In addition to these upwellings alongshore, it has been 

 suggested that they also occur at certain points in 

 the open ocean. There is also some evidence for the 

 existence of horizontal mid-water currents, in a direc- 

 tion the reverse of that of the currents immediately 

 above them. Observations on this question are 

 sorely needed. 



Before leaving the subject, a word must be said on 

 closed seas. The classic instance of these is the 

 Mediterranean ; this includes two deep basins of more 

 than 1,000 fathoms, but its entrance at the Straits 

 of Gibraltar is barred by a ridge of 60 to 130 fathoms 

 in depth. Consequently, the cold bottom water of 

 polar origin never penetrates into it, and the difference 

 in temperature between the surface and the bottom is 

 comparatively slight. Mediterranean water is through- 

 out warm and of high salinity. A surface current 

 from Atlantic to Mediterranean through the Straits 

 is compensated by an undercurrent in the reverse 

 direction ; the water which thus escapes into the open 

 Atlantic spreads, and sinks by reason of its high salinity, 

 carrying its high temperature to quite unusual depths. 

 An instance of a cold closed sea is furnished by the 

 Faeroe Channel, shut off from the main Atlantic by 

 the Wyville Thomson Ridge. 



While exact observations on currents, and especially 

 on mid-water currents, are much needed, the usual 

 methods for their determination (by buoys, log-ships, 

 etc.) are so rough and inexact that they cannot 

 usefully be employed by any but skilled hands. Of 

 more elaborate machines the Ekman current meter 

 (p- 55) nas proved very useful in shallow waters, 

 and a more heavily built pattern could no doubt be 

 made to do deeper work. But for accurate measure- 



