478 PROCEEDINGS OF THE ACADEMY OF [Oct., 



have made their way from one separated area to another by their 

 own powers of flight, perhaps aided by favorable winds; that zones 

 III and IV may have been more continuous throughout the length of 

 Mexico and Central America in Tertiary times, when higher elevations 

 with consequent cooler temperatures prevailed over areas now low 

 and hot, as a result of previous greater volcanic activity or orogenic 

 revolution,^^ some of which latter causes have been invoked by bot- 

 anists^^ to account for the distribution of the plants of this district. 



That temperature is not the only factor in determining the dis- 

 tribution of the Odonata is shown by the fact that each of the zones 

 II-IV, even when continuous over large areas, contains species of 

 quite limited habitat within that zone. Such are Perilestes fragilis, 

 Argia wilsoni, gaumeri, popoluca, and cwpraurea, Argiallagma minu- 

 tum and species of Protoneura in zone II; Hetcerina rudis, Argia per- 

 cellulata, calida, harretti, rhoadsi and pocommia and Palcemnema desid- 

 erata in zone III; Hetcerina tolteca and Argia herberti in zone IV, 



Relations of the Odonate Fauna to Rainfall. 



The existence of pools, lakes, or watercourses of sufficient con- 

 stancy for the development of the aquatic larvse of the Odonata is 

 dependent on the water supply (ultimately traceable to the precipi- 

 tation) and on the factors which tend to prevent its loss by evapora- 

 tion or by sinking into the soil. These latter factors probably include 

 frequency of winds and of clouds, sheltering vegetation and the rela- 

 tive porosity of the soil and underlying rocks. The supply and con- 

 servation of the water of a given area have not only the direct effects 

 of furnishing the necessary living medium for the Odonate larvse, but 

 also, in a more indirect manner, must influence the supply of food for 

 both larvse and adults through the existence of vegetation and through 

 it of herbivorous insects and other animals. 



The influence of these environmental conditions on the Odonate 

 fauna has not yet been thoroughly investigated even in those parts of 

 the earth for which the physical data have been accumulated. For 

 Mexico and Central America, where the collection of these data has 

 been very limited (except at a few well-known localities), it is hardly 

 possible at this time to attempt to correlate the facts of Odonate 

 distribution even with those of precipitation. The existing measure- 



" Hill, Bull. Mus. Comp. Zool, XXXIV, pp. 205-207, etc., 1899. 

 '^Most recently by Braj^, Science for Nov. 9, 1900, pp. 709-716, and Botan. 

 Gazette, XXVI, pp. 121-152, 1898, vnth. citations from previous writers. 



