to each other, and compare their attributes, than when assessing homology in flowers. Floral 
characters are harder to evaluate for it is not easy to visualize which are the comparable 
morphological units. Since gynoecium and perianth vary little in the Lauraceae, but the 
androecium varies widely, the condition of this floral structure always has been used as an 
important character to recognize subordinate taxa within the family. 
Total number of stamens and the number of sporangia per anther are often used as the 
characters to consider from the androecium, but the fact that individual whorls vary 
independently supports the idea that each staminal whorl can be taken as a separate 
morphological unit. In doing this, the number of sporangia becomes an attribute for each whorl 
of stamens, as well as it does the transformation of stamens into staminodes or their elimination. 
Even more, sporangia themselves can be considered characters, and their position on the anther 
and dehiscence be the attributes. This is a better way to evaluate possible phylogenetic lines 
within the family, because comparable units are not mixed. 
Floral characters in Hortonia are in general easy to homologize and compare with those 
of the Lauraceae, except for their phyllotaxy. In Hortonia flowers, there are up to 12 stamens 
arranged in a spiral way. The same occurs in the Atherospermateae and Laurelieae (with few to 
12 stamens), and in the other groups of the Monimiaceae with paired glands on the stamens 
(Monimieae and Peumeae, which have numerous stamens). There are no identifiable whorls to 
compare with the Lauraceae. Therefore, the change from spiral to whorled phyllotaxy has to be 
assumed to have happened in the ancestor of the Lauraceae. 
Evaluation of trends of variation in fruit characters is not difficult in terms of 
understanding what the morphological units are, but rather in determining the primitive (original) 
pb Al 
