76 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [DeC. 17, 



and of land in the northern hemisphere, it would appear that the 

 mean annual temperature of a sea station on the parallel of 50° N. 

 latitude, would not exceed that of a similar sea station in equal S. 

 latitude, unless the temperatures should be materially influenced by- 

 local causes. Hence the lower temperature of the Falkland Islands, 

 as compared with that of the corresponding northern parallel, must 

 be due either to local influences, or to too high an estimate of the 

 mean annual temperature in western Europe in the hypothetical case 

 of the absence of the Gulf-stream. I am disposed to attribute it to 

 the latter reason. The temperature of the Falkland Islands may be 

 depressed by a southern continent not very distant from them, and 

 by consequent accumulations of ice ; but it would seem that the 

 region of western Europe about the parallel of 50° would, in the 

 absence of the Gulf-stream, be liable to an equal similar influence 

 from the Scandinavian region. I am disposed, therefore, to think 

 that the isothermal lines for this case, represented on the map, ought 

 to meet the coast of western Europe at points rather more southerly, 

 so as to indicate temperatures for each locality about 2° or 3° lower 

 than those now indicated. I might have made this correction on the 

 map, but, as the determination of the positions assigned to these 

 isothermals was entirely independent of any comparison with places 

 in the southern hemisphere, I have thought it better to allow them 

 to remain, as a proof of the approximate accordance of results arrived 

 at by independent considerations. These corrected positions of the 

 isothermals would assign to Snowdon, in the absence of the Gulf- 

 stream, a mean annual temperature of 39° or 40°. 



19. In more northern latitudes than that of Snowdon, our fore- 

 going reasoning would lead us to conclude that the mean annual tem- 

 perature would be increased by the submergence of Europe, but only 

 in a comparatively small degree for insular stations and those situated 

 immediately on the shores of the Atlantic. The temperatures, there- 

 fore, of the northern extremity of Scotland and of Iceland, under our 

 present hypothesis, may be taken somewhat greater than those given 

 in the table of art. 13, p. 68, for the case in which the absence of the 

 Gulf-stream was assumed. The temperature of the Alps would pro- 

 bably differ little from that given in the same table. The correction 

 mentioned in the preceding paragraph, if adopted, must, of course, 

 be applied also to these temperatures, 



§ II. On the Height of the Snow-line and Descent of Glaciers below 

 it, (1.) at the present time; (2.) in the previous hypothetical cases. 



(1.) On the Height of the Snow-line at the present time. 



20. Knowing the mean annual temperature at any place on the 

 earth's surface, we can calculate for that place the height at which 

 the mean annual temperature of the atmosphere will be that of 

 freezing, provided we know the rate at which the mean temperature 

 decreases, in ascending from the lower into higher regions of the 

 atmosphere. This rate has been determined with sufficient accuracy 



