Occasional Papers of the Museum of Zoology y 



odonate life. Over other streams these flows have passed, 

 leaving no trace. For example, three or four kilometers back 

 from the Rio Magdalena at Puerto Berrio is a muddy forest 

 stream which we found very rich in odonate life, but we 

 found not a single Hetaerina there, though at Cristalina, dis- 

 tant about 25 kilometers, four species occur, one of them in 

 large numbers. A similar muddy stream without Hetaerinas 

 was found near Cumuto in Trinidad. In British Guiana are 

 species of Hetaerina which would very probably find the 

 Puerto Berrio and Trinidad streams congenial, but these spe- 

 cies have never flowed over the far distant similar habitats in 

 Colombia and Trinidad. The Hetaerina fauna of a tropical 

 stream is determined by its congeniality or lack of it for the 

 various species which, because of its geographical position, 

 are enabled to reach it. 



The character of a stream, its rate of flow, the temper- 

 ature and composition of its water, the geology of its bed, its 

 fauna and flora, are all subjects which may be studied and 

 determined. But the means or methods by which its plant 

 and animal inhabitants have reached it are not so readily 

 analyzed. As stated before, as to the dispersal of dragonfly 

 species we know as little as of their origin, and the problem 

 is difficult, if not impossible, of solution. Referring again to 

 the Rio San Esteban in Venezuela, we find three species of 

 Hetaerinas definitely distributed on the stream. These are 

 caja, macro pus, and capitalis. To the west across the moun- 

 tains, at the swift streams about Tachira, we find two of these, 

 macropiis and capitalis, and a few kilometers lower down at 

 La Fria, where the streams run out into sand, we find mucro- 

 pus and caja. But at both Tachira and La Fria occur also 

 the widely distributed miniata, which is not known east of 

 the Catatumbo River basin, in which both Tachira and La 



