MILLIPEDS — KEETON 55 



B. ligula seems to be near the main line leading from the early 

 indianae type to the more complex types in this group. In this species 

 there is a noticeable increase in the bulk of the gonopods and a 

 solenomerite has been developed. 



It seems quite probable that the line leading from ligula to the 

 more highly specialized forms in the group split into two main bran- 

 ches, one leading to vividicolens and eventually to the more complex 

 splendida, and the other leading to the forms with terminal plates or 

 homologous structures and t3^ically with toothed solenomerites. 

 Within the latter group, calceata, laminata, and kentutkiana all rather 

 closely resemble each other and it would be diflicidt to say with any 

 assurance just what their true relationships are. One possible group- 

 ing is shown in the diagram. 



B. hqffmam presents a perplexing problem. This species has a 

 terminal plate but no solenomerite. In addition, the postcingular por- 

 tion of the tclopodite is long, while it is very short in both laminata and 

 kerduckiana and but little longer in calceata. Placing hqffmani in any 

 scheme of relationships is thus quite difficult. It is without doubt a 

 highly specialized form and its true affinities have been obscm-ed. 

 For want of a better solution, hoffmani is shown in the phylogeny 

 diagram (fig. 11) as arising early from the calceata line. 



It is interesting to note that, with the exception of hqffmani, there 

 seems to be a general tendency in the entire ligula group towards re- 

 duction in the length of the postcingular portion of the telopodite. 



The tliree species glendalea, hubrichti, and falcifera, though easily 

 derived from the hansonia type, do not seem to fit well into either of 

 the above-mentioned major groups. Therefore, each is considered to 

 have arisen independently from the hansonia line. 



All the above speculations as to phylogeny were based primarily 

 on gonopod characters. On this basis the accompanying diagram 

 was drawn. After this had been done, two other characters not pre- 

 viously considered were selected and the evidence presented by them 

 was compared with the diagram. These two characters — the ridge 

 of the coUum and the shape of the midbody paranota — are not in any 

 way associated with the genitalia, yet they show much agreement 

 with the relationships shown. In general the species on the right side 

 of the chart (excluding the dentata-insolita group which will be men- 

 tioned later) have no coUum ridge and have rounded paranota, while 

 those on the left side have collum ridges and square or produced para- 

 nota. Exceptions in the case of collum ridges are plecta and some 

 specimens of eutypa eutypa. Exceptions in the case of the midbody 

 paranota are plecta, eutypa eutypa, turneri, and ligula. The first three 

 have square paranota but there is no indication of their being caudally 

 produced at the caudolateral corners. B. ligula is the only species 



