may have to be rejected. 
Changes from triplinerved to pinninerved leaves, and from domatia present to absent both 
are also suggested to have evolved independently in different occasions. At the same time, the 
presence of straight appressed hairs on lower surface of the leaves, which at some point was 
thought to define big sections within the neotropical species of Cinnamomum, apparently has 
arisen several times in the history of these species too. 
Unfortunately, none of these trends commented above can be assessed entirely since the 
information regarding the Old World species of Cinnamomum is missing in the analysis. First, of 
course, it has to be resolved whether all these species (neotropical and paleotropical) are indeed 
congeneric, for the support of a single clade holding all four morphological groups now included 
in Cinnamomum is rather weak. As mentioned early in the introduction, sections recognized in 
the Old World are separated on the basis of combination of characters that are not present in the 
American species. In addition to these differences it became apparent during this study, that 
domatia are absent in the species with strictly triplinerved leaves (section Cinnamomum), but 
present in those with pinninerved (to subtriplinerved) leaves (section Camphora). In the 
neotropics the reverse situation is found: species with domatia are those with mostly triplinerved 
leaves, while species with pinninerved leaves lack domatia. So far there is only one species in the 
Neotropics that connects the two groups: C. heterantherum, which has domatia and pinninerved 
leaves. 
Certainly for the moment it is advisable to keep the group of species studied here along 
with the Asian species within Cinnamomum, since their shared morphological characters fit well 
the concept of the genus. However, if one wants to represent phylogeny in the classification of 
244 
