GLACIAL PHENOxMENA. 001 



the temperature as to cause a lovrering of the snow-line 

 upon the Ancles, and the formation of local glaciers among 

 tlie serras of Brazil. "N^liat are required, in order that we 

 may he able to read correctly the geologic history of South 

 America, are more careful research and greater caution in 

 making hasty generalizations from insufficient data. If 

 we may rely vipon the facts gathered, the question is one 

 not respecting the existence of an " intertroj^ical cold 

 epoch," but simply of the Ibnitation of glacial action, or 

 ice-phenomena, under the tropics.* 



* Since the above was written and placed in the hands of our pub- 

 lishers, a new volume has appeared from the pen of Prof. Hart, of Cor- 

 nell University, who is at present (November, 1870) in Brazil, having 

 charge of an expedition from the university above named. This work, 

 " Geology and Physical Geography of Brazil," is the most accurate and 

 exhaustive account of the geologic and topographical features of that 

 country that has ever been given to the public. Every part shows the 

 presence of the careful and discriminating observer. Upon an examination 

 of the work, we were pleased to find that what we have said above, rela- 

 tive to glacial phenomena, is fully corroborated, and all fairly placed, by 

 the researches of this eminent geologist, beyond conjectural ground. To 

 the evidences of glacial action in equatorial America, gathered by Agas- 

 siz, and which we have given above. Prof. Hart adds a long series of phe- 

 nomena, as the results of his own explorations during two journeys in 

 Brazil. Upon page 29 he says : " I desire to record here the fact that I began 

 my studies of the Brazilian drift with a conviction that Prof. Agassiz was 

 wrong, and I feel much gratified that my independent observations have 

 so fully confirmed the results of his own." After speaking of the 

 phenomenon, so observable in Brazil, of the decomposition of gnciis 

 rock in silii, and the marks that distinguish the material thus resulting 

 from true drift, he goes on to give, in different portions of his work, evi- 

 dences of drift found in the coast-provinces of Brazil, " from the Bay of 

 Rio Janeiro to the Amazons." Prof. Hart regards the sandstone and 

 clays of the Amazonian Valley as tertiary deposits, thus limiting glacial 

 action, as we have supposed, in our remarks above, facts demand in 

 South America; yet, in exact correspondence with the views we have ven- 

 tured, he says : " My conclusions, after all, do not affect his " (Agassiz's) 

 " theory of the former existence of glaciers under the tropics, down to tlic 

 prostnt level of the sea, a theory which I hold as firmly as he." But, for 



