to be at the base of the evolutionary line of the Monimiaceae. More recently that condition has 
been corroborated by other authors (Thorne, 1974; Endress, 1980). Indeed, the selection of 
Hortonia as an outgroup for the Lauraceae would leave as a possibility the existence of an 
ancestor which had the "predisposition of developing valvate dehiscence" (to use Endress & 
Hufford (1989) words), a feature that was lost in the line leading to Hortonia but retained and 
developed in some Monimiaceae and the Lauraceae. Thus, instead of taking another genus or a 
functional outgroup of several taxa of the Monimiaceae that present paired glands on the stamens 
or valvate anther dehiscence, Hortonia was prefered as the outgroup to polarize the characters in 
the Lauraceae, for that genus is the most conservative in characters among the Monimiaceae, and 
to reduce the chances of having many equivocal designations for ancestral character states. 
Instead of including the outgroup in the analyses to root the cladograms, two different ways of 
"rooting" were accomplished: 1) insertion of a hypothetical ancestral taxon ("Archaeodaphne") 
elaborated by outgroup methodology, and 2) coding characters with a priori premises regarding 
sequence of transformation series. 
Character argumentation. 
Character evaluation to assess polarity should be based on homologous characters. The 
Lauraceae are commonly said to have a high level of homoplasy, but this may be the result of 
comparing characters that are not wholly homologous. Vegetative features are not hard to 
compare (homologize) between taxa, e.g. presence of leaves, phyllotaxy, venation pattern, 
stomate type, and vestiture. In other words, there is a narrower margin of making a mistake in 
deciding what a leaf is, and where it is placed on the stem, or to consider two leaves homologous 
220 
