mediate between H. guianensis and H. lutea, and for this and 
comparable reasons he and Schultes have recently made /utea 
a variety of H. guianensis. . . . One might with almost as much 
reason render the genus unispecific’’. In another context, he 
wrote that, while he preferred to recognize *‘nine — or fewer — 
species. ..., “‘in nature and in various localities entities so 
intergrade that if one wishes . . . he could. . . reduce the genus 
to one species and consider it in terms of trinomials with many 
forms appended’’. 
Nor did Seibert discuss in great detail Mueller’s two Sec- 
tions Euhevea and Bisiphonia, except to state that he had 
“arrived at the conclusion that exact number of anthers is of 
little taxonomic significance within the limits of certain tenden- 
cles’’, pointing out that the number of anthers ‘‘may vary 
within the species and between flowers on the same tree.’ 
Ever since 1945, when, jointly with Ducke, I reduced Hevea 
lutea to varietal rank under H. guianensis, 1 have considered 
the classical infragenetic grouping of species to be both un- 
workable and unnatural. 
Il. HISTORICAL NOTES ON HEVEA MICROPHYLLA 
A natural classification of species into infrageneric groupings 
should rest, whenever possible, preferably on several different 
characters — for example: both floral and fruit characters — in 
which little or no integradation is discernable. Extensive field 
work on wild Hevea in the Amazon and examination of 
thousands of specimens in the major herbaria have convinced 
me that such differences exist and that they may be used as the 
basis of an infrageneric classification which, in my opinion, is 
natural, showing two rather widely divergent trends in evolu- 
tionary development of the genus. 
In 1905, Ule, who had spent a long period studying Hevea in 
numerous areas of the Amazon, described a most interesting 
species: Hevea microphylla, which he had collected in fruit in 
1902 on the Ilha Xibaru, slightly downstream from the mouth of 
the Rio Branco on the Rio Negro in Brazil. Later exploration 
has shown that this species is endemic to the Rio Negro, from 
the middle to the upper course of the river. 
248 
