THE PHYLOGENY AND THE GEOGRAPHICAL DISTRIBUTION 

 OF THE GENUS ERYTHEMIS (ODONATA) 



By Clarence Hamilton Kennedy 



The following paper has been prepared at the suggestion of Mr. E. B. 

 Williamson, who, in the accompanying paper, has presented the systematics 

 of this group. The material has l>een arranged to parallel the arrangement 

 in the writer's article^ on the genus Libellula, and is one of a series of papers 

 which it is planned to present on the relationships of the American Odonata. 



Time is not available and material is not at hand for a definition of the 

 relationships of this genus to the other genera in the Libellulinae. For this 

 reason, the writer has followed the definition of groups presented by Ris,- 

 who followed Calvert.^ The penes of Lepthemis and of all the species of 

 Erythemis recognized by these authors have been examined. Ris includes 

 in this group only one other genus, Rhodothemis of the East Indies. As no 

 material of this species is in Mr. Williamson's or the writer's collections, it 

 has not been studied. However, Ris figures its wings and gives a detailed 

 description of its peculiarities. From this account it is evidently the least 

 specialized of the three genera in the armature of the legs, which have the 

 spines of the femur less developed ; and in the venation, which is "reduced" 

 as it is in credula only of the American genera. The location of Rhodo- 

 themis, the least specialized member of this series, in the East Indies sug- 

 gests an Oriental origin for the group as a whole and that the stock of the 

 American genera at some time in the past has spread into its present habitat. 

 As the majority of the species are tropical, the spreading into America may 

 have taken place when the Asia-Alaska land-bridge had a climate suitable for 

 such species, which must have been in Pre-Miocene times. The other possi- 

 bility is that they came across from Africa, but so far we have not recog- 

 nized any African relatives of the group. Credula, the most primitive of the 

 American species, has not been found north of Panama and St. Thomas, 

 while the two species that are found farthest north, hacnidtogastra* and sim- 

 plicicollis, are two of the most specialized, which distribution seems to con- 

 firm a southern or tropical entrance into America. 



The homologies of the lobes of the penis in the genera Libellula, Sym- 

 petrum, Orthetrum, and Orthemis were worked out by the writer in a pre- 

 vious article.^ All parts that were defined for these appear also in the series 



1 The Phylogeny and the Geographical Distribution of the Genus Libellula (Odo- 

 nata). Ent. Nezvs, XXXIII, pp. 65-71, 105-111, 1922. 



- "Libellulinen," Fasc. XIII, pp. 591-608 in Coil's Zool. du Selys, 1911. 



•' Nomenclature of Certain North American Odonata, Ent. Nezvs, XVII, pp. 30-31, 

 1906. 



* Williamson doubts the Georgia records of this species. 



5 The Morphology of the Penis in the Genus Libellula. Ent. Neivs, XXXIII, pp. 

 33-40, 1 92 1. 



