22 
snout a. little more than 3 in length of body ; its base 
only 1 in length of liead; first ray not stiífened (*). 
The origine of the adipose is so indistinct that it 
hardly can be defined, but it is scarcely more than 1 
base of dorsal behind the end of the latter and is 
nearly of lialf the length of the body; it is the highest 
at its posterior end and is continons with the caudal 
fiu, the rays of which begin at 1 length of base of 
dorsal before the end of body. The middle rays of 
caudal are the longest (equal to length of liead); below 
tliey are confluent with the anal fin, the end of which 
is noticeable by the last rays being lengthened. All 
other rays of anal are of nearly equal length, but the 
first are shorter. The beginning of the anal fin coincides 
with that of the adipose, a little before middle of body. 
Base of ventrals somewhat behind middle of dorsal; 
the longest rays passing half-way to the anal. Anal 
aperture a little behind base of ventral. Distance between 
ventral and pectoral about 4 j/cj í 11 length of body. 
Pectoral fin abo ve the levei of the mouth-aperture, al- 
most at the margiu of the gill-openings ; their rays 
being equal in length to that of ventral. 
Colour may have grown pale by alcohol, but there 
is no sign of markings either on body, or on fins. 
Type: One specimen of a total length of 107 mm. 
frorn Alto da Serra in the State of S. Paulo. 
* 
* * 
I should not have ventured to describe this spe¬ 
cimen as a new species, if there were not the remar- 
(*) The hight of this fin camiot be indicated, as all the ravs are da- 
maged in our typical specimen, but it must have been shorter than length 
of base. 
