EDWIN CHAPIN STARKS 49 rows of scales, while those from the lake are almost entirely colorless in alcohol.
165. Gerres olisthostomus Goode and Bean.
Specimens were taken at Natal and Lake Papary. Those from Natal are darker, the young have 4 or 5 narrow vertical cross bands on the middle of the side, and in addition, lighter longitudinal bands following the rows of scales. No trace of cross bands is on the Papary specimens, though the others are evident.
Famity POLYNEMID-.
166. Polydactylus virginicus (Linnzus).
This fish was found in abundance about Natal.
FAMILY SCIZ NIDA.
167. Cynoscion microlepidotus (Cuvier and Valenciennes). This species is apparently common at Para, as several specimens were collected in the market.
They agree very well with the description published by Steindachner, but show the following differences. ‘The interorbital space (bone only) is contained from 434 to 5 times in the head. The maxillary usually scarcely reaches to the posterior margin of the orbit, but in the largest specimens, a foot in length, it reaches to the margin. The vertical fins are covered with small, thin, inconspicuous scales but they can scarcely be said to be thickly covered, not, at any rate, as compared with Cynoscion squamipinnis. 168. Cynoscion leiarchus (Cuvier and Valenciennes). Two specimens were collected at Para. This species resembles Cynoscion phoxocephalus most closely. The head is more compressed, the caudal is weakly double concave rather than simply concave, the eye is larger, the postorbital part of the head not so long, and the posterior outline of the anal is not so oblique.
169. Cynoscion phoxocephalus Jordan and Gilbert.
A single large specimen of this species was taken at Natal. It has been compared with specimens from Panama and no important differences appeared. The body and head are a trifle deeper, and it has 23 dorsal rays rather than 21. This form has heretofore been known only from Panama Bay.
50 FISHES FROM BRAZIL 170. Sagenichthys ancylodon (Bloch and Schneider).
Sagenichthys mordax Gilbert and Starks (Mem. Cal. Acad. Sci. IV., 1904).
Three specimens from 7 to 8 inches in length were collected at Para, which enable me to compare directly, for the first time, this species with Sagenichthys mordax from the Pacific.
When Sagenichthys mordax was described it was compared with current descriptions of S. ancylodon and from these it appeared to differ in’ having larger scales in the lateral line, smaller scales on the rest of the body, and shorter gill-rakers. All of these characters have been inaccurately described for S. ancylodon and a direct comparison of the specimens from both coasts fails to show any specific differences. There are about 100 oblique series of scales above the lateral line in S. ancylodon (not 85) and half as many scales on the lateral line (not 75). The gill-rakers are one-third of the diameter of the eye (not two-thirds). The specimens from Para have a somewhat larger eye than those from Panama, but the difference is not too great to be accounted for by the difference in size of the specimens. 171. Nebris microps Cuvier and Valenciennes.
Four large specimens of this species were taken at Para, each of them a little over a foot in length. This makes a direct comparison between this species and its representative in the Pacific, Nebris occidentalis Vaillant, for the first time possible.
Comparing them with some specimens from Panama of similar size the mandible is strikingly weaker and does not protrude nearly so much, lacking the large, sharp process at the symphysis. The lips are thinner and the mouth is less oblique. There are 95 series of scales above the lateral line, ‘counting the nearly vertical series, and 11 scales between the lateral line and the front of the soft dorsal. These counts in Webris occidentalis are respectively 115 and 20. The scales on the belly are very much larger in Nebris microps. The pectoral is shorter, or 1% in the head, while in the other species it is almost equal to the head. ‘The radiating striations on the preopercle are not nearly so coarse and do not end in such coarse pectinate processes. The body is marked with 6 or 7 wide, conspicuous, dark cross bars, which are not evident in Webris occidentalis and have not been described. The following is the color of fresh specimens. The back is dusky drab, with wide, nearly black, cross bars, anteriorly not so wide as the interspaces, posteriorly more crowded and equal to them. ‘There is a trace of one over the opercle, one under the spinous dorsal, four under the soft dorsal, the first
EDWIN CHAPIN STARKS 5 of which is under the 8th to the 11th rays, and the last under the last rays, and one across the caudal peduncle. ‘The lower parts are very bright chrome yellow, or sometimes deep orange, being brightest on the head. ‘The ventrals, pectorals and anal are bright yellow or orange, the first two growing coal black toward their points, and the anal dusky with dots. The dorsal rays grow dark toward the edge of the fin, and the caudal is slightly yellow, obscured by dark points. 172. Plagioscion squamosissimus (Heckel). Several specimens were collected at Para, the largest 9 inches in entire length. This species may readily be separated from Plagioscion auratus and Plagioscion surinamensis by the short, slender, second spine of the anal, which is contained from 3% to 4% times in the head, and is about as thick as the distance from the front of one anal ray to the front of the next, or scarcely wider than the ventral spine. In the other two species it is twice or more times wider than this. In the mandible there is a row of small, fine, sharp teeth, and just inside of them is a row of from 7 to 10 widely spaced canines, many times larger than the outer row. ‘The two rows are so closely approximated that they appear like a single row with a tooth enlarged at intervals. In the premaxillary there are two rows similar in size to the lower teeth, but the outer row is the large row in this case, instead of the inner, and the rows are more distinctly separated. The differences in size between the teeth in this species is much greater than in the other two species here considered, or in other words, the enlarged teeth are much more enlarged. TABLE OF MEASUREMENTS EXPRESSED IN HUNDREDTHS OF LENGTH. Length in mm. to caudal base teacdmnblOOthSior lengthy. ..s.5.0+- oa Wena chy ome mia xt lary iyi rps:
ohs)e <-s orci 2 942124 WON CAGHAMeLEHOMEVC eye ais iat) < siete ere lene Wadthvorimterorbital Space. .2.-.2...-2. 4. Tu 25 (aaiel VE 21000) bP eee eR ee Wen oGhnotpectOnale aise cM en's) 6 ies ajsrsic leh encthvotiventrallaayai).cisierfeelsliwtaeeho WMenethionsecondanallspine:...4.5.4.5...) Wadthvorisecondianalspine).f)50.)..3./y.5.,.« Length of fourth dorsal spine............. Wenowmpon caudal paises a2) ls sh, ae siolerncchs Numberior/dorsal rays... 2) as 2 asl INGTMDerrOranalraySaa. des.
-/a-1- 2 les eieraie
52 FISHES FROM BRAZIL 173. Plagioscion auratus (Castelnau).
Several specimens from 4 to 15 inches in length were collected at Para in the market. These are all very much darker than the other two species at hand, with the exception of one 6 inches long, and, to a less degree, one twice as long. ‘The darkest ones are almost black on the back, the head, and all of the fins, and dark dusky on the lower parts. The lightest one is slightly dusky on the back, the dorsals and caudal; the balance is without color, the color being as in the other two species. This species may be known from Plagioscion surinamensis by the smaller elliptical eye, not so near the anterior profile; the wider, more convex, interorbital space, which is from half again as wide as the vertical diameter of the eye to twice as wide. The snout is a little more produced, and the lower jaw is always included. ‘The length of the pectoral is equal to the distance from the tip of the snout to the preopercle. The first ray of the ventral, in all sizes, is filamentous and projects beyond the other rays about one diameter of the eye. The gill-rakers number 10 or 11. Other differences are indicated under the description of Plagioscion surinamensis and in the tables ofmeasurements.
TABLE OF MEASUREMENTS EXPRESSED IN HUNDREDTHS OF LENGTH. Length in mm. to caudal base............. Headmn LO0thsiotdengthierssc aces eine Leng thtofmeaxillanyae anaemic seitacieraciete Wong diameteriotieve tee eons Width ofiinteror britain is eisai, tens Bengtosnoutiyoeinaeesas)weagslenaras Lengthvofpectoralsieeiiaeiy nice cine aiaae Length of ventral without filament Lensth' of second analispine wns. hale Width of second anal spine................ Length of fourth dorsal spine............. engthvohicaudalernateeeeemecEeiaceeie:
Number of/dorsaliraysanseie noone ence Numberofjanaliraysiven-cmeeetereciceries 174. Plagioscion surinamensis (Bleeker). A few small specimens and one large one, a little over a foot in length, were collected at Para. They are all very light in color; only slightly dusky above and on the edges of the dorsal fins. This species appears from literature to be much more rare than the other two here reported upon, and it was not found nearly so abundantly in the market during our stay at Para.
EDWIN CHAPIN STARKS 53 It may be readily known from Plagioscion auratus (Castelnau) by the larger, rounder eye nearer to the anterior profile; by the narrower interorbital space, which is less in width opposite the middle of the eye than the vertical diameter of the eye; by the snout which does not overhang the mouth so much, the lower jaw being scarcely, or not at all included; by the longer pectoral, which is equal in length to the distance from the tip of the snout to a point on the opercle midway between the preopercle and the gill opening; by the first ray of the ventral, which, though a trifle longer than the other rays, does not end in a filament. It has 2 or 3 more gillrakers on the lower limb of the arch, or 12 or 13; the preorbital is a little narrower, the second anal spine a little longer, and the scales on the interorbital space are ctenoid a little farther forward. Little difference exists between these two species in the character of the teeth. The character of the interorbital space will the most readily separate them. Most of these differences are well shown by the drawing of the type of Plagioscion auratus and the drawing published by Steindachner of Plagioscion surinamensis (Fische-Fauna des Magdalenen-Stromes. Pl. I), though there are also many differences there shown that do not exist owing to the inaccurate drawing of the former species.
TABLE OF MEASUREMENTS EXPRESSED IN HUNDREDTHS OF LENGTH. Length in mm. to caudal base............. 250 128 15 Meadiin l00ths of length s5.... 22.2.0... 5. 32 34 32 Meneeniormaxillary: jo 60sesscss deo seis ees 15 15 | 15 Won GIAIMETEE OL CVE... dues ck ele p se cee ee 6 7 7 Width of interorbital space............... 5 5 6 WenpGnvormsnOm Gere cits cle tie deere tvs ele se elaine vi 8 7 WENO Gh PECtOTaly. oh 5.5.0 vis scne esc a ecco 26 25 24 Went hvotevermtralere i cscs ark « ccre ose aso os 25 22 23 enethiof second anal spine...........4...- 14 14 15 Width of second anal spine...........:-..' 2 2 2 Length of fourth dorsal spine............. 14 13 15 WenoGivotieatidallys ies Sa 3).ci.c ae sistsneeatiale 23 25 26 INtimibeniordorsalirays: 20.0.2. a05 eee Se 33 X-I, 32 X15 33 INumiberOtranalinayS:\).o. 4c4 2 <ctdeecla ae «6 TENG II, 6 IBIAS 175. Bairdiella armata Gill.
A single specimen was taken in the market at Para. Comparing it with specimens from Panama the snout is not so sharp, the spines of the preopercle are more blunt and wider at the base, the mouth is a little more oblique and more curved, the gill-rakers below the angle of the first arch number 16 rather than 13, and the second anal spine is a trifle longer. It appears probable that with more material the Atlantic form may be separable
54 FISHES FROM BRAZIL from the Pacific, but all of the differences are too slight to warrant their separation with the material at hand.
176. Micropogon opercularis (Quoy and Gaimard).
Three specimens were taken at Para.
Comparing them with specimens of Micropogon furniert from Jamaica the snout is sharper and projects farther over the mouth, the head is not so deep and is more depressed above the eyes. The scales scarcely differ in size, there being 6 or 7 in a vertical series between the first dorsal spine and the lateral line, 8 or 9 between the first anal spine and the lateral line, and 50 lateral line pores. There are 26 or 27 rays in the dorsal in the Brazilian specimens, and 28 or 29 in the specimens from Jamaica. Famity CICHLID/.
177. Aequidens tetramerus (Heckel).
A half dozen specimens were taken in the market at Para. 178. Cichlosoma bimaculata (Linnezus).
A large number of specimens were taken at Ceara Mirim, Lake Papary, Ceara and Lake Extremoz. The largest is 614 inches in total length. Among them are 3 specimens which have only 3 anal spines, or one less than the usual number.
There are two scales between the upper lateral line and the soft dorsal, without the pore-bearing scale, but with a small scale at the base of the fin. The lateral lines are separated by 2 scales besides the pore-bearing scales. The upper lateral line covers 18 scales, and the lower one 6. On specimens 11% inches in length the tip of the ventral scarcely reaches past the front of the anal, and the longest rays of the soft dorsal and the anal do not reach to opposite the middle caudal rays. In specimens 6 inches in length the ventrals reach well past the anal spine, and the soft dorsal rays reach beyond the tips of the caudal rays, while the anal rays do not quite reach to the tips of the caudal rays. In none of them do the ventral rays reach past the middle of the anal (‘‘Ventral often extending to the posterior end of the anal.” Regan Ann. Mag. Nat. Hist. ser. 7 XVI, p. 68). Following is the color taken from life. The back is dark green growing light bluish on the side of the body. The breast and belly are dusky bluish. There are 7 or 8 dark cross bars wider than the spaces between them. A dark spot is usually present below the posterior part of the eye, sometimes conspicuous, or sometimes nearly or quite absent. A larger, more
EDWIN CHAPIN STARKS 55 definite and darker spot is just below the lateral line midway between the front of the eye and the base of the caudal. A smaller one is on the upper part of the caudal peduncle, usually ringed with bluish body color. A dark band, more or less conspicuous, runs back from the eye through the middle spot to the spot on the caudal peduncle. It is often interrupted by the middle spot, and is usually more conspicuous in front of it than behind it. The fins are very dark, sometimes nearly black; the pectorals are always lighter; the caudal, soft dorsal, and anal are usually with numerous light spots of the bluish body color, but these are absent when the fins are nearly black.
179. Cichlosoma temporale (Giinther).
A single specimen of this species was taken in the market at Para. It agrees in all essential respects with current descriptions. 180. Crenicichla lepidota Heckel.
Specimens from Lake Papary, Lake Extremoz and Ceara Mirim are referable to this species, though it has hitherto been known only from Southern Brazil, Paraguay and Buenos Ayres. ‘They agree very well with the description published by Mr. Regan (Proc. Zool. Soc. Lond. 1905, p. 158). They resemble specimens of Crenichichla saxatalis from Para, but have larger scales, which number from 40 to 45 below the lateral line, instead of from 50 to 56. ‘They are deeper, being from 26 to 30 hundredths of the length (without the caudal) instead of from 23 to 25 hundredths. The head is usually larger and the suborbital a little deeper. The anal rays are usually 8 or 9, while in Crenicichla saxatilis they are occasionally 9, but nearly always 10. They have 7 or 8 large spots along the side, which are nearly always evident and usually conspicuous.
The following life colors, taken from the specimens from Lake Papary, show three color phases. In the first specimen the ground color is greenish slate, with a series of large black blotches along the sides edged with pearly spots. Similar small pearly spots are scattered over the side of the body. On the head a dark band runs back from the eye, giving place to an irregular pearly stripe on the side of the body. A small black spot is present at the base of the upper caudal rays. All of the vertical fins are edged with dusky and the anal and caudal are inconspicuously spotted with white. The dorsal is bluish with metallic spots. The pectoral and ventrals are light yellowish, and the caudal is tinged with the same color.
In the second specimen the ground color is very dark olive-green above, growing lighter below with yellowish tinges on the side of the' body and slate
56 FISHES FROM BRAZIL color on the belly. Pearly green spots are scattered over the side, but these do not ring the dusky blotches. A black streak runs back from the eye, and a small black spot is present below the eye. The dorsals are very dark green, almost black, with a few inconspicuous greenish pearly spots scattered over them. The caudal and anal are tinged with yellow and bordered with dusky, and the ventrals and pectoral are light dusky brown.
In the third type of color the ground color on the upper parts is as described for the second specimen, but lighter and growing to white below, while the belly is tinged with wine-color. The shoulder spot is slightly ringed with light green. The usual streak runs back from the eye, and the usual side blotches are present, but there are no pearly spots. The dorsal is black toward the tips of the rays, and a very conspicuous milk-white band with bluish tinges runs near its edge, but leaves a clear-cut black border. The base of the dorsal is yellowish. ‘The anal is dusky yellowish green, and the caudal is similar but blackish at the base and edged with dusky below, while on the upper edge the black bordered white band of the dorsal is continued. The pectoral and ventral are yellowish.
181. Crenicichla saxatilis (Linnzus).
Several specimens were taken in the market at Para, the largest 10 inches in length. Among them are two that differ from the others, but the difference seems to be due to disease or lack of nourishment. ‘The larger one is covered with large, black, irregular blotches different on the different sides of the body. ‘The teeth have disappeared except a few on the sides of the upper jaw, and a spongy, papilose growth of tissue has taken their place. Inside of the mouth are several copopods. The lower jaw projects more than usual and a large symphyseal knob has developed. ‘The body is a little more slender and the eye is a little larger. The other is somewhat smaller and less emaciated. Only one or two black spots are present; the lower jaw is not at all projecting, and the spongy growth has not replaced the teeth to so great an extent.
The following is the life color. The upper parts are very dark green shading below to drab posteriorly, to wine-red on the posterior part of the belly, and to white on the breast. The side of the head is bright yellow and the ventral part is white. A dark shoulder spot is irregularly ringed with silvery pigment and has silvery spots scattered over it and about it for a short distance. A dark streak runs through the eye to the opercular flap. A dark spot is present at the base of the upper caudal rays. The entire length of the dorsal is black toward the tips of the rays, with a vivid white