NO. I BROMELIACEAE OF BRAZIL — SMITH 5 



needs comment, Mez's list of Bromeliaceae in Luetzelburg, Estudo 

 Botanico do Nordeste 3 : 104. 1923. This greatly increases our 

 knowledge of the distribution of species in arid northeastern Brazil, 

 yet curiously enough Mez seems to have forgotten it in his final mono- 

 graph in the Pflanzenreich. My other large source of data is Reitz, 

 who has kept me constantly informed as he adds to the flora of Santa 

 Catarina. 



In a few instances pictures, not specimens, must be cited as types 

 of species. This is particularly the case in Baker's species described 

 from the unpublished Icones of E. Morren. These are at Kew and in 

 some cases can be identified with specimens at Liege, but Baker saw 

 only the pictures. 



Since duplicate collections of Bromeliaceae are relatively rare and 

 since the important herbaria as regards the family have been covered, 

 there seems little point in listing exsiccatae. Instead, the numbers of 

 photographic negatives on file in the Chicago Natural History Mu- 

 seum, the Gray Herbarium, and the United States National Museum 

 are noted whenever they are likely to be helpful in establishing the 

 identity of a species. 



GEOGRAPHICAL CITATIONS 



The citation of localities in Brazil is extremely difficult for a num- 

 ber of reasons. First of all, many of the early collectors did not bother 

 to record localities or their data were lost, with the result that "Brazil" 

 is all we have left. Next, some of the older settlements have ceased 

 to exist and frequently places have changed their names. Who now 

 recognizes Sao Sebastianopolis as Rio de Janeiro or Desterro as 

 Florianopolis ? If it were not for Urban's great study of collectors in 

 the introductory volume of "Flora Brasiliensis," the case would be 

 hopeless. 



More recently there is the difficulty of changes in spelling and place 

 names by government action. Xapeco becomes Chapeco, retaining 

 the pronunciation but bewildering the reader. In its "Index to Map 

 of Hispanic America 1:1,000,000" (p. 402), the American Geo- 

 graphical Society explains the latest Brazilian orthography and its 

 listings cover the majority of localities involved in this paper. 



However, since the publication of this index in 1945, there has 

 been a new reform to eliminate duplication of names, not just within 

 a state but for the whole country. Thus, where Bom Jesus could 

 occur as the name for a dozen localities in one state, there must now 

 be but one place with that name in all Brazil. This has been accom- 



