ZINGIBERACEAE 



163 



he had not seen the flowers of most of the species, and hence 

 did not know of the peculiar distinguishing character of the 

 inflorescence. This may also explain why t. 68 was placed under 

 Zerumbed, rather than under Curcuma, where it really belongs. 



Of the six species enumerated in chapter seventeen, under 

 Zerumbed, the first and second are discussed below; the third 

 (album), fourth (giring), and sixth (manga), all cultivated in Java, 

 are still undescribed under the binomial system; while the fifth 

 (frigidum) , as regards the name ties, but not the description, 

 belongs to Curcuma viridiflora Roxb. Fl. Ind. 1 (1820) 34, a 

 species of the section Mesantha. Roxburgh himself reduced 

 the fourth species to Curcuma viridiflora Roxb., but that was 

 evidently an error on his part. 



CURCUMA LONG A (non Linn, quae est prob. C. aromatica Salisb., vide 

 Hermann Hort. Acat. Lugd. Bot. Descr. (1687) 208, t. 209) Koenig 

 in Retz. Obs. 3 (1783) 72, non alio rum. 

 Curcuma domestica minor (et major?) Rumph. Herb. Amb. 5: 164, 

 t. 67. 



Rumphius's description agrees exactly with our commonly 

 cultivated curry plant, as do the Malay names cited by him, 

 and the form that he described and figured is apparently 

 identical with that described by Koenig as Curcuma longa. It 

 is characterized by its central spike, white bracts, and orange 

 rhizomes. In Java a tall form and a short form occur, but I 

 cannot detect any notable differences between them. Curcuma 

 longa Linn., Sp. PI. (1753) 2, was based wholly on Curcuma 

 radice longa Herm., which probably is the same as Curcuma 

 aromatica Salisb. Linnaeus, however, reduced the Rumphian 

 species, as figured, to Curcuma longa Linn., in Stickman Herb. 

 Amb. (1754) 20, Amoen. Acad. 4 (1759) 129, Syst. ed. 10 (1759) 

 843; while in the second edition of the Species Plantarum (1762) 

 3, he erroneously referred it to Curcuma rotunda Linn., which is 

 supposed to be the same as Kaempferia pandurata Roxb. 



CURCUMA PETIOLATA Roxb. Fl. Ind. 1 (1820) 37, sensu latissimo K. 

 Schum. in Engl. Pflanzenreich 20 (1904) 102. 

 Curcuma agrestis Rumph. Herb. Amb. 5: 164. 



This species, in the wide sense that it was interpreted by K. 

 Schumann, was until recently known only from Pegu and 

 Moulmein, but has been detected growing also in Java. Rumph- 

 ius's description agrees so closely with that of Roxburgh, 

 that I do not hesitate in reducing Curcuma agrestis Rumph. to 

 C. petiolata Roxb. 



