6 species of Phoebe in the Neotropics, Meissner (1864) recorded 14 species, and by the end of 
the nineteenth century Mez (1889) accepted 45 species. Kostermans accepted 87 species, which 
he transferred to Cinnamomum, and about 15 more specific names have since been published 
(e.g., Vattimo, 1962, 1977, 1984; van der Werff, 1987b, 1988, 1991b). As several species were 
first described without information about the fruit, and floral characters had not been critically 
revised in the whole group, the little more than one hundred names currently in use were 
undoubtedly an overestimation of the actual number of species. Since presence of particular 
features is no guarantee for the correct placement of a taxon in the classification system, the 
group of American species formerly recognized in Phoebe needed to be thoroughly surveyed in 
order to define which of them really go under the genus Cinnamomum, and to find out if Phoebe 
is indeed absent in the Neotropics. Therefore, a critical survey of herbarium specimens, along 
with selected and suitable observations and collecting in the field, would result in a better 
estimate of the number of taxa, possibly fewer than 60 (van der Werff, 1991a). 
It is pertinent to mention that some botanists currently consider in Cinnamomum only 
those species (formerly in Phoebe) that have triplinerved leaves (Vattimo, 1962). At the same 
time, some American species of Phoebe with pinninerved leaves whose floral features do not fit 
Cinnamomum characters have been already removed to other genera (Rohwer, 1986, 1991). 
Another problem I will examine here is the relationship of the American species to the 
two sections recognized for the genus in the Old World tropics. While in Phoebe, the species 
studied here were placed in two subgenera, Heteranthera and Euphoebe (Mez, 1889), based on 
the number of cells found in the stamens of the third whorl, however, as mentioned before, the 
sections Cinnamomum and Camphora are based on different (combinations of) characters 
7 
