EFFECT OF A CHANGE IN THE GULF-STREAM. 401 



be mentioned the remarkable fact recently observed by Dr. 

 Giinther, that out of 173 tropical marine fish, no less than 57, 

 or 30 per cent., occur on both sides of the isthmus in both 

 the Atlantic and the Pacific.* 



Mr. Croll, however, has pointed out that at present the 

 " S.E. trade winds of the Atlantic blow with greater force than 

 the N.E. trades, and the consequence is that the S.E. trades 

 sometimes extend to 10 or 15 N. lat., whereas the N.E. trades 

 seldom blow south of the equator. But during the glacial 

 epoch the very reverse must have occurred. Hence the great 

 equatorial current of the Atlantic must during that period 

 have been driven considerably south of its present position." -f* 

 Even at present, while the greater part of the water enters the 

 Gulf of Mexico, one portion is deflected southwards, which in 

 the case mentioned above would happen to the greater portion, 

 if not the whole. 



Under existing circumstances, however, the southern divi- 

 sion is comparatively small ; by far the larger portion of the 

 great equatorial current turns northwards, and warms the 

 Northern hemisphere, so that the comparatively high tempe- 

 rature of the Northern Atlantic is in some measure due to heat 

 derived from the Southern hemisphere. In a recent memoir,* 

 Mr. Croll has shown the great effect produced by the Gulf- 

 Stream on the present climate of Europe. He calculates that 

 it conveys as much heat as is received from the sun by 

 3,121,870 square miles of the Equator: nearly as much as is 

 received from the sun by the entire Arctic regions, the pro- 

 portions being as 15 to 18. Our present climate is 12 higher 

 than the normal due to its latitude, but Mr. Croll points out 

 that this is by no means to be considered as measuring the 



effect of the Gulf-Stream. The temperature of the whole 



/ 



* Trans. Zool. Soc, vol. vi. p. 397. 

 f Croll, Philosophical Magazine, Aug. 1864. 

 I 1. c. Feb. and Oct. 1870. 



9 n 



_i LJ 



