12 E. B. Williamson 



business at El Banco. For several hours before reaching the town we were 

 passing, on the left bank of the river, tlie finest and most extensive pasture 

 seen. El Banco is well situated on high ground at the juncture of the Rio 

 Cesar with the Magdalena. Above El Banco was extensive fishing in the 

 Cesar which was deep and swift with bare washed mud banks. The fish, 

 taken in nets, were dried on racks or on the ground. We followed a wide 

 road, the Camina Chimichague Y Chinguana, leading out from El Banco, 

 which at first passed through some slightly rolling country. The stream 

 beds here were all dry, with the banks and immediately adjacent country 

 wooded with bamboo and palms. About three miles from town two trails left 

 this road. The trail to the right apparently followed the Rio Cesar; the 

 other, to the left, w^as probably the Monipos road. These trails immediately 

 crossed a perfectly fiat, scantily grassed- sun-baked plain over which 

 termite nests were scattered with suiiicient regularity and with just suf- 

 ficient variation in size to give the impression of an old and very large ceme- 

 tery. That we were suffering from thirst and our heads fairly rocking with 

 the heat when this funeral plain came into view, detracted nothing from the 

 vividness of the impression. Both trails led to a scarcely perceptible de- 

 pression, marked by adjoining forest which traversed the plain like a broad 

 ribbon. The shallow creek bed in this forest was almost dry, but about the 

 long stagnant pools Psaironeura remissa flew by hundreds. One of these pools 

 was possibly one hundred yards long. All were shaded. Large numbers 

 of small brown monkeys took considerable interest in our work and showed 

 some offense at our presence. 



About half a mile out from town, on the right hand side of the main 

 road, are several marshes. We attempted to work these and then cut across 

 in an easterly direction to the Rio Cesar, but were unable to do this because 

 of low impenetrable jungle. The marshes themselves yielded but poor re- 

 sults. 



For about two miles which we explored above El Banco the right bank 

 of the Rio Cesar was largely pasture, the original forest having been almost 

 entirely destroyed. At this point a large affluent from the right comes into 

 the Rio Cesar. Like the Magdalena, the Cesar had lagoons, only on a 

 smaller scale, and in some of these, which were stagnant and muddy and 

 heavily shaded with brush, we took numbers of shade-frequenting coena- 

 grionines. Sometimes we secured these where the dense gloom made it al- 

 most impossible for us to distinguish the insects. It is rather remarkable 

 that the great forested areas of the north, at least of the Nearctic region, 

 have no such shade-frequenting dragonfiies as occur about both ponds and 

 streams in the tropics. Possibly our more northern species are more ancient 

 than the environment in which they now^ find themselves. 



We left El Banco the afternoon of January 26, bound up river for 

 Barranca Bermeja, 139^4 leagues above Barranquilla. Our collection now 

 totalled 113 species and 5,407 specimens. 



The days that followed differed but little from the days spent between 

 Barranquilla and El Banco. At night the boat was tied up to the bank. 

 During the day there were the same frequent stops to take on wood. The 



