FLORA OF THE BRAZILIAN PLATEAU. 311 



by me for the first time ; but the species were nearly 

 all different from those of the coast region, and the 

 general aspect of the flora still more markedly 

 different. There was no trace of that luxuriance 

 which w^e commonly expect in tropical vegetation ; 

 monocotyledonous plants, except grasses, were very 

 few, and, in place of the large ferns that abounded at 

 Santos, I found but a single Gleichenia, allied to a 

 species that I had gathered in the Straits of Magellan. 



Although a fair number of plants were still in 

 flower, I soon came to the conclusion that night frosts 

 must be not unfrequent at this season, and that a 

 considerable proportion of the vegetation must be 

 annually renewed. I found several groups of small 

 trees, chiefly of the laurel family, and for the first 

 time saw the A raiicaria brasiliensis, possibly in a wild 

 state ; but none of the trees attained considerable 

 height, and I doubt whether in a state of nature this 

 plateau has ever been a forest region. I was rejoiced 

 to see again, growing in some abundance, the splendid 

 Bignonia veniista, and was led to doubt whether its 

 real home may not be in the interior, and its appear- 

 ance at Santos due to introduction by man. 



We possess a fair amount of information as to the 

 climate of the Brazilian coasts, but our knowledge of 

 the meteorology of the interior provinces is miserably 

 scanty. I was led to conjecture that, although the 

 district surrounding San Paulo is not divided by a 

 mountain range from the neighbouring coast region, 

 the climate must be very much drier, and that the 

 rainfall is mainly limited to the summer season. 



In the course of my walk, I unexpectedly ap- 



