polymorphic characters is not incorporated into the analysis. However, I considered that the 
record of polymorphic characters within genera can be the result of a taxonomy based on general 
resemblance, and characters of unusual species in several genera may not represent evolutionary 
lines within those genera but offshoots from other genera that come to resemble other taxa. Then, 
the insertion of these data would produce false variation for character states within genera in the 
analysis. That is why initially a comparison of core character syndromes for the genera was 
preferred. In any case, the data matrix (figure 19) devised for this study can be used to assess in 
the future other combinations of characters for the genera, or subdivide the genera into as many 
different "natural groups" as can be recognized within them. A summary of character status is 
presented in figure 20. 
Results. 
Effectiveness of cladograms may be diminished since the analyses were stopped 
prematurely. As can be seen in the data matrix (figure 19), the amount of characters is low 
compared to the number of taxa, a condition which diminishes the posssibility of getting an easy 
solution to the cladograms. It was expected that the characters with several states would 
compensate in some way for this lack of characters in the analyses, yet this did not happen. In the 
case of the analysis without restrictions on character states changes, the memory allocated in the 
program was insufficient to hold all the equally parsimonious trees produced. For the analyses 
with restrictions on character changes time necessary to compute all equally parsimonious trees 
was extremely long, and the search was stopped after several days of computer work on each 
analysis. Strict consensus trees were computed for all the analyses, but terminal taxa mostly were 
231 
