296 Annals of the Carnegie Museum. 



the bent pole straightens out, flips up, and often catches the fish 

 under its head. 



15. Angle-worms and Open Umbrella for Eels. — The natives of 

 Aregua, Paraguay, catch Symbranchus with a bunch of worms tied 

 to a string and an opened umbrella. When the eels seize the worms 

 they are jerked up with the upturned umbrella before they let go 

 of the food. 



16. Dead Animals and Imprisoned Fishes. — A freshly killed animal 

 or an imprisoned large fish left in the rivers over night are sometimes 

 full of Cetopsis and parasitic Pygidiidce on the following morning. 



17. Hollow Logs and Submerged Canoes. — Hollow logs and sunken 

 canoes frequently yield prized specimens, if quietly and quickly 

 removed from the water. In the case of such logs it sometimes is 

 possible to stop or plug the open end and then roll them out upon the 

 bank. 



18. Holes in Banks, Rocks, Stumps, and in Logs. — Rare specimens 

 were often taken by me in holes and crannies by the use of my naked 

 hands. 



19. Soft-wood Canoes. — I was told that the native caught a species 

 of Tachysurus during the spawning season with soft-wood canoes. 

 They are split and sunk quickly by means of weights. The erected 

 dorsal spines of the mass of assembled fishes penetrate the soft wood 

 and the fishes are held captives. This is possible from what I know 

 of this fish, but I did not see this mode of fishing practiced. 



Preserving and Transporting Specimens. 

 It is quite as difficult to save as to make collections while traveling 

 in Brazil. In the remoter districts formalin is the best preservative 

 on account of its small bulk. All medium-sized to large specimens 

 should be injected the same day on which they are taken with 95 

 per cent, alcohol or a little strong formalin. The native rum or ca- 

 chaca suffices to preserve specimens for only a few days. If kept 

 longer in this fluid they soften. In tropical climates it is best to kill 

 fishes by the use of strong alcohol, avoiding, however, a degree of 

 strength sufficient to cause distortion after death. As a rule 70 per 

 cent, alcohol is best, but there is no absolutely fixed formula for dif- 

 ferent species. It is almust impossible to transport fishes on mule- 

 back and in ox-carts, because they are shaken to pieces sooner or 

 later. 



