F o rm .— Still more oval and elongated than the last species, the profile falling in a gentle curve from
the dorsal to the end of the snout. D epth exactly one-third of the length, excluding caudal :
head one-fifth of the entire length, caudal included. Maxillary rather dilated towards the bottom,
and quite straight, not curving backwards as in both the last species. T eeth rather larger,
and more conspicuous. Nostrils larger. Only twelve scales in the depth, the lateral line
occupying the seventh row from the top : thirty-seven or thirty-eight in the lateral line. Scales
of a rather different form ; the basal margin more sinuous, the free margin not so regularly
curved, the coloured striæ hardly obvious. Dorsal more oblong than triangular, higher in relation
to the depth, of which it equals two-thirds, commencing at a point anterior to the middle
o f the length, excluding caudal. T he pectorals reach a little beyond tho ventrals, which are
attached a little in advance of the dorsal. T he anal has all the rays longer, and more nearly
equal, the posterior part of the fin not being so much sloped off : there are three spines at the
commencement, the first two minute, the third not quite half the length o f the soft rays : this
fin is furthermore distinguished from that of the two former species by the rays being set with
asperities, which communicate a scabrous harsh feel to the touch, when the finger is passed
along them from the base upwards.
C o l o u r .— Not noticed in the recent state. In spirits it appears more silvery than either of the two
last species : the back and upper part of the sides being brownish. A humeral dusky spot, and
the remains of what was probably a bright silver band along the middle of the side from the gill
to the caudal. A t the base of the caudal is a dusky spot, which is prolonged in a line along
the central rays to the commencement of the fork. T he other fins faintly edged with dusky,
but otherwise pale.
Habitat, Rio de Janeiro.
The more oval and elongated form, straight maxillary, and scabrous anal fin,
at once distinguish this species from either of the two last. It was taken by Mr.
Darwin in fresh water, at Rio de Janeiro, in June.
4 . T e t r a g o n o p t e r u s t æ n ia t u s . Jen.
T . corpore ovali, valde compresso, alliludine tertiam partem longitxidinis, hâc ad hasin
furcoe caudalis mensa, cequante: osse maxillari margine posteriore recto: pinnâ
dorsali suprà ventrales accurath exorienti; anali loevi, altitudine retro cito
decrescente; radiis plicis membranaceis nullis : squamis in lineâ laterali 4 0, in lined
inter pinnas ventrales et dorsalem transversâ 14.
D. l/IO ; A. 3/22, &c.
L o n g . unc. 2 . lin . 2.
F o rm .—D epth and general form similar to those of the T . rutilus; also the same num ber of rows
of scales, the lateral line occupying the eighth from the top. Anal smooth, and similar to that
of the T . Tuiilus in form, but in the number of the spines and soft rays agreeing with the
T . scabripinnis. T he maxillary straight, but hardly so much dilated as in the last-named
species, being o f nearly equal breadth throughout. The teeth are more numerous than in
either, amounting in the lower jaw to fourteen or more. The ventrals are in an exact line with
the commencement of the dorsal.
C o lo u r.—M uch as in the T . scabripinnis. T he same siivcr band, only more b riiiiant: aiso the
same humerai spot, and the spot at the base of the caudal extending along the middle rays.
Habitat, Rio de Janeiro.
The above description is that of two small specimens, similar to eacli other,
obtained by Mr. Darwin in a running brook at Socego, in the province of Rio de
Janeiro. Tliey are probably not full grown ; but so evidently distinct from either
of the last two species, the characters of which are in some measure combined in
them, that I have not hesitated to give them a separate place. The silver band is
more or less obvious in all the species of this genus brought home by Mr. Darwin,
but it is much brighter in this than in any of the others.
5 . T e t r a g o n o p t e r u s i n t e u r u p t u s . Jen.
P l a t e X X I I L f i g . 4 .
T . corpore ovali, valde compressa: altitudine tertiam partem longitudinis totius
æquantc : ore parvo ; osse maxillari brevissimo, dilatato, margine posteriore recto :
dentibus minutis, multicuspidatis : dorsali subtriangulari, pone ventrales exorienti ;
anali (in uno sexu?) scabrâ, altitudine retrd subito decrescente; his pinnis plicis
membranaceis nullis : squamis in linea longitiulinali 3-5, in lineâ inter pinnas
ventrales et dorsalem transversâ 10 .■ lineâ laterali interrupt^, haud finem pinna
pecloralis attingente.
D. ] 10 ; A. 2/18 ; C. 18, & c ; P . 11 ; V. 7
L o n g . u n c . 2 . lin . 8.
Form.— Oval, and not very dissimilar in general form to the T . rutilus, but rather more elevated
above the shoulder. D epth exactly one-third of the entire length: head one-fourth o f tbe
same, caudal excluded. Profile not exactly straight, but very slightly hollowed out at the
crown, then as slightly protuberant above the eyes, whence it falls more rapidly in front, giving
the snout a short and blunt appearance. M outh much smaller than in any of the preceding
species, owing to the extreme shortness of the maxillary, which is broad, with the anterior margin
curving outwards, but the posterior one straight. Teeth very small ; the points on the
cutting edges numerous (five or six on each tooth) and nearly equal; apparently only one row
on the intermaxillary, and the same in the lower jaw ; none on the maxillary. Eyes and nostrils
as in the other species, b u t the anterior orifice of the latter larger. Subopercle much
larger, forming a greater portion of the gill-flap. Scales relatively larger ; only ten in the
depth, and thirty-five in the length. They have no deep striæ on the basal portion, and consequently
no fivn ; but they are very regularly marked with the usual finer striæ concentrically
arranged, except on the free portion where they become indistinct.
The lateral line occupies the sixth row of scales from the top, but is very soon interrupted,
coming to an end before it has reached the length of the pectoral, and not carried over more
than eight or nine scales in the lengtli. Dorsal subtriangular, commencing exactly at the
middle of the length, caudal excluded. Anal shorter than in the other species, and not reaching
so near the caudal ; two spines, but the first extremely minute. Caudal forked for half its