6 E. B. Williamson 



separated earlier distinct continental land masses, have proved efficient bar- 

 riers for many species. These valleys may be mechanical l)arriers or they 

 may merely mark separate regions in each of which, through long adjustment, 

 conditions have resulted which prevent the occupation of one region by many 

 species from another region. Barriers, in any case, exist only in connection 

 with specialized habits of dragonflies. For a high stream species a long 

 mountain range might offer a continuous highway, which would be an insur- 

 mountable barrier for a low pond species. A long, broad valley, forming a 

 great stretch of suitable habitat for the pond species, might prove an impas- 

 sable barrier for the high stream species. In either case, if the river valleys 

 are barriers, Erythemis has occupied and passed them, and if the higher 

 regions themselves are exclusive to species from other regions, Erythemis, 

 because of certain characteristics discussed and doubtless others, has success- 

 fully forced its way, at least to some extent, into the new environment. As 

 a result, the present distribution of species offers, I believe, an unsolvable 

 problem for the student who would determine the place of origin and trace 

 the routes of distribution of the species of this genus. For example. Funda- 

 cion. Colombia, lies in the relatively isolated Santa Marta region, and at this 

 one station five species of Erythemis occur, and these five occur in Central 

 America, the West Indies, and to the south to Paraguay. At Tucacas, Vene- 

 zuela, four species have been recorded, and I have no doubt more careful col- 

 lecting there would add at least two species. At El Banco, in the lower 

 Magdalena, six species have been taken. 



When we speak of the distribution of dragonflies or anything else, we 

 imply, of course, the known distribution. The relation of the known distri- 

 bution to the real distribution depends on the thoroughness with which col- 

 lections are made. And this thoroughness depends on the number and distri- 

 bution of collectors and the degree of success of each collector in obtaining 

 everything in his specialty in each locality he works. And the collector's 

 degree of success, other things being equal, depends on the relative conspicu- 

 ousness of the things he collects. The imagoes of species of Erythemis are 

 conspicuous insects, and the records of their capture are correspondingly 

 numerous. Though collectors and localities visited in the past have been 

 few in numbers with what we may expect in the future, probably our 

 knowledge of the distribution of species of Erythemis corresponds at least 

 in a broad way with the real distribution of these insects. Compare the status 

 of our knowldge in this genus with that of some of the gomphines, for 

 example. To cite one case, we know that the genus Erpetogomphus occurs 

 in South America because of the capture, which might be described as acci- 

 dental, of two specimens in Colombia and three in Venezuela — a total of five 

 specimens for the continent. 



Definition of thf Genus 



The following definition of Erythemis is compiled almost entirely from 

 other authors. Head moderate or small ; eyes joined for only a short dis- 

 tance; frons moderately prominent, rounded, without anterior ridge; vertex 

 moderately high, apex slightly emarginate. Hind lobe of prothorax erect, 



