THE AMERICAN TETRAGONOPTERINAE. 133 
Seales thin, not conspicuously regularly imbricate; each scale with several 
nearly parallel horizontal striae; anal sheath composed of a single row of scales 
along the front of the fin; caudal lobes scaled for about one eighth of their length. 
Origin of ventrals a little nearer tip of last anal ray than snout, slightly 
in advance of the dorsal; penultimate dorsal ray more than half the length of 
the longest which is three and three fourths in the length. Anal emarginate, 
ventrals reaching anal; pectorals not to ventrals. 
In life, base of upper caudal lobe red, base of lower caudal lobe yellow, 
some yellow on under side of caudal peduncle and in front of anal. A circular 
spot about as large as eye on base of middle caudal rays. A dark line in front 
of dorsal, a series of spots behind it. A well-defined humeral spot on and over 
the second and third scales of the lateral line. 
HemiIGRAMMUS GILL AND HypHEssSOBRYCON DurRBIN.! 
Compressed, rarely subeylindrical. Greatest depth on or near the vertical 
from the first dorsal ray. Postventral region narrow. Maxillary not slipping 
under the preorbitals. Lower jaw included, when the mouth is closed, but 
usually projecting beyond the snout when the mouth is open. Premaxillary 
teeth in two rows, the outer row more or less incomplete. All teeth conical or 
flat or fan-shaped, with 3-7 cusps, usually graduated from the enlarged median 
cusp. 
Seales eycloid, regularly imbricate, striae few and variable in number. 
No interpolated scales or rows of scales. Anal sheath short, of a single, usu- 
ally incomplete series of small scales. Lateral line somewhat decurved to nearly 
straight, the series of scales just below it parallel to it. Pores developed on the 
first three to eighteen scales. 
The genus Hemigrammus was placed in the synonymy of Astyanax (Tetra- 
gonopterus) by Ulrey because Liitken had shown that in some supposed speci- 
mens of 7’. fasciatus the lateral line is complete, in others not. Two other 
species, 7. theringii and T. oerstedvi are given by Ulrey as possessing these inter- 
mediate characters. But in the description of 7. oerstedii it is specifically 
stated that the lateral line is complete, not interrupted. Also all the speci- 
mens of 7’. iheringii so far secured have a complete lateral line. This left 7. 
1 The account of these genera has been worked up from my original manuscript by Mrs. Marion 
Durbin Ellis (Miss Marion Lee Durbin). She has devoted so much labor and care to these genera 
that she alone should be credited with the authorship. 
