Phylogenetic Systematics of I guanine Lizards 1 23 



Under the other two alternative interpretations, the presence of the first derived state in 

 Dipsosaurus is considered to be convergent, as in Figure 47E and F. For this reason, I 

 have chosen to leave the relationships among Brachylophiis, Dipsosaurus, and the new 

 ingroup (node 3) unresolved in the assessment of polarities for the lower-level analysis, 



I used the same basic methodology for determining polarities in the lower-level analysis 

 (Appendix III) that I used in the preliminary analysis, where the relationships of the 

 outgroups to the ingroup are uncertain (Appendix II). For reasons presented in Appendix 

 III, I considered polarity to be determinable only when both Brachylophus and 

 Dipsosaurus exhibit the same character state. 



Using Brachylophus and Dipsosaurus as additional outgroups for analysis at a lower 

 hierarchical level, I was able to determine polarities for 13 of the 20 characters whose 

 polarities could not initially be determined (Table 8). The number of premaxillary teeth 

 turns out to be two characters (hence the numbering in the character list as characters 43- 

 44) representing transformations in opposite directions from the ancestral condition, a 

 mode of seven premaxillary teeth. Although character 84 (superciliary scales) differs in 

 Brachylophus and Dipsosaurus, it seems reasonable to conclude that state A is derived, 

 since it is found in neither Brachylophus nor Dipsosaurus and represents one end of a 

 continuum that has the conditions seen in these two taxa at the other end. Both 

 Brachylophus and Dipsosaurus exhibit the same state for character 61, but this character is 

 irrelevant to an analysis of relationships at the level in question because it does not vary 

 within the new ingroup. Characters 56 and 80 also do not vary within the ingroup, but 

 their polarities are undeterminable because Brachylophus and Dipsosaurus exhibit different 

 conditions. 



The use oi Brachylophus and Dipsosaurus as additional outgroups for an analysis of 

 relationships at a lower hierarchical level necessitates a reevaluation of the polarities of 

 those characters whose polarities had already been determined using more remote 

 outgroups. The reasoning behind polarity reevaluation is similar to that behind polarity 

 assessment and is presented in Appendix IV. Under this reasoning, the only characters 

 whose polarity assessments needed to be changed after reevaluation were character 18 

 (polarity reversed) and character 46 (changed to undeterminable). Character 46 is a four- 

 state character, and what becomes undeterminable is whether state A or state B is ancestral. 

 Therefore, I have lumped states A and B as state and consider states C and D to be 

 successively more derived conditions (i.e., C = 1, D = 2). 



Eight of the characters used in the preliminary analysis of relationships among all 

 iguanines cannot be used in the analysis of relationships of all iguanines other than 

 Brachylophus and Dipsosaurus, either because they must be interpreted as synapomorphies 

 of a basic taxon that are convergent with a condition found in Brachylophus or Dipsosaurus 

 (characters 13, 51, and 62) or because they do not vary within the new ingroup (characters 

 25, 48, 66, and 77). These characters were removed from consideration, and the 

 remaining characters were combined with those whose polarities were newly determined, 

 using Brachylophus and Dipsosaurus as outgroups, and whose derived states characterized 



