138 GOEDONIA. 



1916. G. penangensis, Eidley, Penang and Singapore. It was 

 misnamed G. excelsa by King in 1890. 



1916. G. hirtella, Eidley, mountains of the Malay Peninsula from 

 Perak to Negri Sembilan. 



To these 21 species I have four to add, being G. taipingensis 

 from the hills near Taiping, G. concentricicatrix from Selangor, 

 G. Havilandii from Sarawak and G. lanceifolia also from Sarawak, 

 making 28 in all : but as I consider the Asiatic species of Haemo- 

 charis to be Gorclonias, the genus is really in my view of well over 

 30 species : but all of them want more study. 



Attempt to divide Gordonia in conformity with its 

 discontinuous distribution. 



Szyszylowicz (in Engler's Pflanzenfamilien, iii. part 6, 1893 

 p. 185) divided the genus Gordonia into two sections, the American 

 species forming the first, and the Asiatic species the second. 

 He. gave no sectional names, and did not define the two. 

 Korthals and Pitard have separately gone further in at- 

 tempts to cut the Asiatic species from the American. Korthals'' 

 attempt was made long ago (Yerliandel. over de Natuurlijke Ges- 

 chiedenis der Nederland overzeeische Bezittingen, Kruidkundig, 

 1839-1842). He made two new genera to hold the Malayan 

 species regardless of the existence of Polyspora; and on p. 127 

 he wrote " The plants on which the genera Antheeiscliima and 

 Closascliima (his new genera) have been established, were formerly 

 written up and described under the genus Gordonia; but a closer 

 comparison with material of the American types, leads me to ob- 

 serve that this union of Indian and American plants can be severed, 

 and leads me on into presenting two new genera for the former, 

 cut out of Gordonia." This shows him dominated by the idea that 

 plants from opposite hemispheres should not be in the same genus. 

 And he continues defining the first genus Antheeiscliima upon 

 doubtful interpretations of its bracts and calyx adding the last 

 sepal (which is seen in fig. 1. to the top and right) to the petals and 

 another sepal to the bracts so as to reduce the number to 3 ; and 

 on " the stamens being free in Gordonia or united into five bundles ; 

 but in Antheeiscliima grown together with the petals into a ring." 

 He next admits this genus as rather indistinctly defined but 

 " Closascliima differs by reason of its more shrubbv growth, lesser 

 development of leaves, and flowers and in some floral characters; 

 it has five or apparently by the enlargement of a bract six sepals, 

 five petals spirally arranged and five stigmas almost without 

 styles." 



The last character is that of TIaemocliaris into which his genus 

 has been placed in later years, while the single species making the 

 genus Antheeischima has been returned to Gordonia as Blume's G. 

 excelsa. Yet Korthals continues under Closascliima to describe the 

 C. marginata as having a style. 



Jour. Straits Branch 



