1872.] 



' Shearwater ' Scientific Researches. 



621 



surface-temperature of the Gulf of Mexico is usually lower than that of the 

 Caribbean Sea, and during several months in the year it is rather below 

 that of the Gulf-stream in the Florida Channel (§1 24). Hence the surface- 

 layer, at least, of the Gulf-stream would appear to be derived from that 

 warmer current which, after flowing westward along the south of Cuba, 

 turns abruptly round Cape Antonio, and then flows eastward towards the 

 Florida Channel. — But we might expect that the Gulf of Mexico, even 

 more than the Caribbean Sea, would have its sub -surface temperature ele- 

 vated by the limitation to the admission of Polar water ; and such an 

 elevation we find to be presented in the section between the Tortugas and 

 Havana, which may probably be considered, as regards this stratum, to 

 represent the Gulf of Mexico generally. Accordingly, on comparing its 

 thermal condition (§ 129) with that of the corresponding stratum of the 

 Caribbean Sea, as shown in the temperature-soundings of the ' Mercury' 

 (§ 127), we see that the temperatures between 100 and 400 fathoms are 

 a few degrees higher in the Tortugas section than in the Caribbean Sea ; 

 the temperature of 62°, which extends in the former to 300 fathoms, 

 ranging in the latter only to 200. But even this downward range contrasts 

 strongly with the limitation of that temperature to the uppermost 100 

 fathoms in the open Ocean, and the rapid fall in the next 100, as shown in 

 the 'Mercury ' soundings to the east of Barbadoes (§ 127). This down- 

 ward extension of surplus heat is retained through all the early part of the 

 Gulf-stream ; as has been already shown in the contrast between the 

 temperatures of the Polar Current and of the Gulf-stream off Sandy Hook 

 (§ 128). 



137. It can scarcely be doubted that the extraordinary thickness of the 

 superheated stratum in the Gulf-stream has much to do both with its own 

 retention of a high surface-temperature in its northward course and with 

 its power of imparting heat to the Air above it ; for it is obvious that the 

 thinner the superheated stratum, the more speedily will its temperature be 

 brought down by that of a colder superincumbent atmosphere, But it is 

 also clear that its loss of heat may be really much greater than the reduc- 

 tion of its surface-temperature would indicate ; for as fast as its superficial 

 stratum is cooled down, its increase of density will cause it to sink until it 

 meets with water as cold as itself, the warmer sub-surface stratum rising 

 into its place. And since this process will continue to take place so long as 

 there is a subjacent stratum of higher temperature than the superincum- 

 bent Atmosphere, the swr/ace-temperature of the current may be maintained 

 with comparatively little reduction, though the temperature of its subjacent 

 strata may have been reduced almost to that of the ordinary Oceanic water 

 which it overlies. But in the very same proportion that it thus loses heat, 

 will it impart heat to the Atmosphere above ; and may thus, by warming a 

 vast body of air which is rapidly borne towards North-western Europe, 

 furnish an important element in the amelioration of its climate. — The de- 

 pendence of the retention of a high surface-temperature on the thickness of 



