308 MERRILL. 



iiiKler the name of F. ruiuiiKjidiin, and also a second J^]iili[)i)ine species, 

 F. sphaerocephala, in the Botany of tlie \'()yaui' of the Bonite, Atlas, 

 1843. In 1883, Naves in the Xovissima Ap[)i'ndix to the third edition 

 of Blanco's Flora de Filipinas, 385, 286, enumerates four species, which 

 are all, witli the possible exception of /'". hizonensis, admitted on 

 erroneous identifications, and can be ignored. He reduced Tillandda 

 pseud o-anana& Blanco to Freijcinetia insignis Blume, l)ut this is a manifest 

 error, as an examination of Blanco's description shows conclusively that 

 Tillandsia pscndo-uminas can not be a Frei/cinctia, but is probably a 

 Pandanus, and possibly the same as P. copclandii Merr. Blanco did not 

 consider any species of Freycineiia in his Flora de Filipinas. 



Having recently had an opportunity to examine the types or authentic 

 material of all the Philip])iiie s})ecies considered by Warl)uvg, in the 

 herbaria at Kcw and Berlin, it became evident that a certain nunilicr of 

 recently descrilu'd forms were invalid, three of the species (lescril)ed by 

 Mr. Elmer, and one by myself. Tn justice to Mr. Elmer, however, it 

 is manifest that the determination of two of his species as new, F. 

 hichanensis and F. confusa. was due to errors in Warburg's monograph, 

 the former being identical with F. ferox Warb., the leaves of the t}'pe of 

 which are about 1 m long l)ut described as .30 cm long, the latter being 

 the same as F. ridaHi Hemsl. The affinity of the latter was recognized 

 by Mr. Elmer, l)ut Hemslcy's sj)ecies was placed by Warburg in the 

 wrong section of the genus, the type lieing a very iimnature specimen. 



In view of the fact that a recent paper has been published on Philip- 

 pine Pandanus, it has been thought advisable to prepare a list of the 

 known species of the other large genus in the family, Freycineiia, giving 

 also a provisional key to the species. Twenty-four species are recognized, 

 all of which are endemic in the Pbi]i|)pines, so far as is known, giving 

 the Archipelago a far gi'calci- known number of species than any other 

 geogra])hical region in which the genus is found. T.iUzon alone has 

 eighteen species, while the region about !Mount Banajao, Province of 

 Tayabas, Luzon, is remarkable in ha\ing no b'ss than eleven sj)ecies of 

 the genus, more than are known from iuiy single island in the Malayan 

 region; New (luinea and the Alalay Peninsula coming lirsl with but 

 eight species, Cch-bcs next wilh se\cn. .lava with six. lloi-iieo ami New 

 (y'aledonia with four each, Sumatra wilh ihree. and various otbei- islaiuls 

 with one or two species each. The above disl libution list is based largely 

 on W'arbui'g's inonograph. and the niurdier of species actually known 

 from some of the above islands jnay be iargei- than the figures given, 

 while undoubtedlv a great many undescribed I'oiins remain to be col- 

 lected. 



