1908.] NATURAL SCIENCES OF PHILADELPHIA. 469 



Arranging the subfamilies in the order of their percentages of endemic 

 species, the series obtained is Cordulegasterinse {two species onl}^ 

 100%, Gomphinffi 85.7%, Agrionina? 72.2%, Calopteryginffi 47.8%, 

 Lestinse 42.8%, Libellulina? 16%o, iEshninffi 13%. The Cordulegas- 

 terinse, many Gomphinse, most Libellulinse and iEshninae have well- 

 developed powers of flight. Perhaps the great majority of the other 

 three subfamilies are feebly-flying insects, yet some of their species ap- 

 pearing in the present faunal district are very widely distributed, e.g., 

 Enallagma civile, Ischnura ramburi, Anomalagrion hastatum. It is 

 consequently impossible to account for the relative endemicity of the 

 subfamilies by such general considerations. 



If the relative endemicity of these groups is not always inversely 

 proportional to the powers of flight, as these figures seem to indicate, 

 and if nearly 40% of the West Indian Odonata are not to be found 

 in Mexico and Central America in spite of favoring winds, the expla- 

 nation of the present distribution of this group of insects may perhaps 

 be found in the past distribution of land and water^^ in these regions. 



Distribution of the Odonata within Limited Portions of 

 Mexico and Central A:\ierica. 



Table 2 and the remarks on the fauna of the Mexican plateau 

 (page 463) have already illustrated this topic to some degree. Table 

 5 gives the number of endemic species and of those common to the 

 three countries whose Odonate fauna is best known. 



Accepting the areas of Mexico (exclusive of Campeche, Yucatan and 

 Baja California), of Guatemala and of Costa Rica as approximately 

 655,000," 63,000,15 ^nd 21,000 square miles (1,700,000, 164,000 and 

 54,000 square kilometres) respectively, it follows that, in proportion 

 to its area, Costa Rica is much the richest coimtry of the three, both 

 in its total number of species and its number of endemic species. 



Tables 6-8 give the number of species and the number of locahties 

 at which they were collected in each of the States or Departments of 



'^ Compare the geological data embodied in the sketch maps of Gadow (Proc. 

 Zool. Soc. London, 1905, II, pp. 235-6); also the discussions in the papers of 

 Chapman {Bull. Ajner. Mus. Nat. Hist., IV, pp. 318, 326-9, 1892) on birds, 

 Simpson {Proc. U. S. Nat. Mus., XVII, pp. 428, 438, 447, 1894) on land and 

 fresh-water mollusks, and Ortman {Proc. Ainer. Philos. Soc, XLI, pp. 309, 341, 

 347) on fresh- water decapods, of the West Indies. 



'* Romero, Geographical and Statistical Notes on Mexico, p. 91, New York, 

 1898. 



1^ Century Dictionary, Vol. IX, New York, 1906. Dr. Sapper gives the approxi- 

 mate area of Guatemala as only 110,000 square kilometres, Mittelamerikanische 

 Reisen u. Studien, p. 424. 



