18 A JOURNEY IN BRAZIL. 
fins acting as sails rather than wings, and carrying them 
along on the wind. They skim over the water in this way 
to a great distance. Captain Bradbury told us he had 
followed one with his glass and lost sight of it at a consider- 
able distance, without seeing it dip into the water again. 
Mr. Agassiz has great delight in watching them.* Having 
never before sailed in tropical seas, he enjoys every day some 
new pleasure. 
April Wi. Yesterday Mr. Agassiz lectured upon the 
traces of glaciers as they exist in the northern hemisphere, 
and the signs of the same kind to be sought for in Brazil. 
After a sketch of what has been done in glacial investigation in 
Europe and the United States, showing the great extension of 
ice over these regions in ancient times, he continued as fol- 
lows : " When the polar half of both hemispheres was covered 
by such an ice shroud, the climate of the whole earth must 
have been different from what it is now. The limits of the 
ancient glaciers give us some estimate of this difference, 
though of course only an approximate one. A degree of 
temperature in the annual average of any given locality 
corresponds to a degree of latitude ; that is, a degree of 
temperature is lost for every degree of latitude as we travel 
northward, or gained for every degree of latitude as we travel 
southward. In our times, the line at which the average 
annual temperature is 32, that is, at which glaciers may be 
formed, is in latitude 60 or thereabouts, the latitude of 
Greenland ; while the height at which they may originate in 
latitude 45 is about 6,000 feet. If it appear that the ancient 
southern limit of glaciers is in latitude 36, we must admit 
that in those days the present climate of Greenland 
extended to that line. Such a change of climate with 
* See Appendix No. II. 
