cladograms of two groups of genera (except few taxa) that have been suggested to conform 
different subfamilies in the Lauraceae. The Laureae, including Actinodaphne, Dodecadenia, 
Neolitsea, Litsea, Laurus, Lindera, Sassafras, Parasassafras, and Umbellularia; and the Perseae, 
including the rest of the genera. This point is more apparent in figure 23, where the ancestor was 
included in the analysis. It is interesting to note in this respect the closeness of the genera 
Chlorocardium and Cinnadenia to the genera in the Laureae as well. These genera have been left 
as of uncertain affinity in Rohwer's taxonomic scheme of the family. Another point to stress is 
the exclusion of Parasassafras, Sasassafras, and Umbellularia in all cladograms from the 
Laureae. This is due to the lack of glands in stamen whorls I, I], and IV, and the abscence of 
supernumerary whorls of stamens, all these characters considered sy phies for the 
Perseae in the analysis. 
A second general aspect to comment is the support, at least partially, for some of the 
subgroups mentioned by Rohwer. Particularly, genera in the Aniba, Beilschmiedia, Cryptocarya, 
and Mezilaurus subgroups come together or close in the cladograms. This is more evident in the 
cladogram resulting from the analysis without restrictions (figure 21), and especially the one 
resulting from incorporating the ancestor (figure 23). It is interesting that some genera from 
different subgroups result in the same clades suggesting a closer relationship among them, as has 
been envisaged by Rohwer; e.g. the clade Anaueria - Ravensara in the cladogram of figure 22, or 
the clade Aniba - Williamodendron in figure 21. 
On the other hand, the scattered distribution of the genera in the Ocotea and Persea 
subgroups in the cladograms was something unexpected, for their morphological similarities 
indicated they are well defined groups. Certainly some pieces of these subgroups are maintained 
PE 
