1851.] HOPKINS ON CHANGES OF CLIMATE. 69 



north of the barrier, on the contrary, the variation of temperature 

 would be more rapid than at present. 



15. The next supposed case is that in which the whole Atlantic, 

 from the equator northward to Greenland, Iceland, and the North 

 Cape, should be converted into land. In this case there would be 

 no reason why the isothermals should not preserve their parallelism, 

 with the exception of merely local deviations, from points near the 

 east coast of Asia to corresponding points near the west coast of North 

 America. Let us first consider the January isothermal of 32° F. in 

 the northern hemisphere. 



If we take this line as drawn independently of the disturbing in- 

 fluence of the Gulf-stream, but supposing the Atlantic still to exist, 

 it is characterized by a northward inflection as it approaches the At- 

 lantic, due to the influence of that ocean. In the present case there 

 is no reason for that inflection, and we may assume the isothermal to 

 pass nearly in a straight line from the Black Sea to the point where 

 the actual isothermal of 32° F. meets the coast of America. This 

 will render its course nearly a straight line between the opposite 

 coasts of the single continent into which the two existing continents 

 would be united in the case we are considering. Its course would be 

 very nearly, but not exactly, east and west, reaching a rather lower 

 latitude in Asia than in America, in consequence of the high table- 

 lands of south-eastern Asia. 



The three succeeding existing isothermals to the north on Dove's 

 map (those for 23°, 14°, and 5° F.) are very nearly parallel to that 

 of 32°, and equidistant from each other through about 80° of long. 

 in Asia, and through 30° of long, in N. America. In the interme- 

 diate space they are inflected to the north by the Atlantic. Their 

 positions in our present hypothetical case would doubtless be ap- 

 proximately determined by substituting straight lines across this in- 

 termediate space, for the actual inflected lines. (See Map, PI. II.) 



The present winter isothermals, north of those just mentioned, 

 become more irregular in their course in eastern Asia, as well as in 

 North America. In the former region the distance of successive 

 lines from each other is somewhat less than in the latter, owing to 

 the unsymmetrical position of the region of maximum cold, which, 

 instead of being symmetrically situated round the north pole, inclines 

 considerably towards north-eastern Asia. This position must be due, 

 in great measure, I conceive, to the influence of the North Sea, 

 warmed as it is by the Gulf- stream. In our hypothetical case this 

 cause would be removed, and we may conclude that the region of 

 maximum cold would be situated nearly symmetrically with reference 

 to our one great continent, but inclined from the pole towards that 

 continent, and from the great Pacific Ocean situated on the opposite 

 side of the pole. The probable longitude of its central point would 

 be nearly that of the present western coast of Europe. This would 

 bring the isothermals in our hypothetical case, lying on the north of 

 those already traced, into approximate parallelism with the equator 

 between the eastern coast of Asia and the western coast of America. 

 The distances between them (for such, at least, as should not be too 

 far to the north) would necessarily be much the same as between 



