Atlantic Flyingfishes of the Genus <I>Cypselurus</I>

ATLANTIC FLYINGFISHES OF THE GENUS CYPSELURUS, WITH DESCRIPTIONS OF THE JUVENILES1

JON C. STAIGER Institu/e of Marine Science, University of Miami

ABSTRACT The Atlantic species of the exocoetid genus Cypselurus are reviewed and six species recognized: Cypselurus coma/us, C. cyanopterus, C. exsiliens, C. furcatus, C. heterurus, and C. pinna/ibarba/us. A diagnosis and synonymy is presented for the genus, as well as dis- cussions of ontogeny and range. The splitting of Cypselurus and recognition of the genus Cheilopogon by Parin (1961) is not accepted. Two keys to Atlantic species are presented; one for specimens under 100 mm standard length, the other for larger specimens. For each species distinctive characters, description, and synonymy are given, and ontogeny, range and reproductive biology are discussed. Particular attention is paid to the ontogenic changes in pectoral and pelvic fin coloration with regard to identification. The placement of C. luetkeni in the synonymy of C. heterurus and of C. lineatus in the synonymy of C. pinnatibarba/us is upheld.

lNTRODUCTION AND ACKNOWLEDGMENTS The systematics of the flyingfishes (Exocoetidae) have been confused since the description of a second species in 1766. Though more than 60 names had been proposed for the Atlantic forms by 1935, investigators studying large series of specimens have recognized no more than 20 as valid. Bruun (1935: 5-8) listed the synonymy of the group and recognized 17 species. Breder (1938: 56) reduced this number to 16. Much of the difficulty encountered in describing exocoetids results from the presence of usually considerable growth changes, such that the adults of some species bear little resemblance to juveniles in color pattern and presence of mandibular barbels. Most early workers had so few specimens that no complete developmental series could be assembled, and many new species were named for the juvenile forms of existent species. Breder was the first to describe developmental series of many Atlantic species, but he did not have enough specimens to construct a key that would function for individuals of all sizes. This paper reviews in a comparative manner the Atlantic species of Cypselurus in the light of the large number of specimens at my disposal, describes in detail some of the ontogenic changes of each species, and pre- sents a key adequate to identify juvenile as well as adult specimens.

IContribulion No. 639 from The Marine Laboratory, lnstitute of Marine Science University of Mia,mi, Florida, This paper is a modification and enlargement of a thesi~ submitted In partial fulfillment of the degree of Master of Arts, Boston University, Boston, Massa- chusetts. 1965] Staiger: Atlantic Flyingfishes 673 This work was undertaken at Boston University at the suggestion of Dr. Robert H. Gibbs, Jr., and I am very grateful to him for his continuing helpful criticism and valuable suggestions. The original study was enlarged and the manuscript considerably revised under the supervision of Dr. C. Richard Robins of the Marine Laboratory, Institute of Marine Science, University of Miami, and I express my gratitude to him for his patience, advice, and criticism. For the loan of specimens I am indebted to Mr. Jack W. Gehringer of the Bureau of Commercial Fisheries, Biological Labora- tory, Brunswick, Georgia, Dr. Giles W. Mead of the Museum of Compara- tive Zoology, Harvard University, and Dr. Carter R. Gilbert of the University of Florida. Specimens at the National Museum were examined through the courtesy of Dr. Leonard P. Schultz. Dr. Thomas W. McKenney advised me in the preparation of the fin dia- grams and Mr. William P. Davis assisted invaluably in the preparation of the photographs. This study was supported in part by a grant-in-aid from the National Science Foundation (NSF-G-19324).

MATERIALS Specimens examined were from the following sources: Boston University (BU) (these specimens are now at the United States National Museum); Museum of Comparative Zoology, Harvard University (MCZ); Marine Laboratory, University of Miami (UMML); Bureau of Commercial Fisher- ies Biological Laboratory, Brunswick, Georgia (BLBG); United States National Museum (USNM); University of Florida (UP).

METHODS All measurements were taken with dial calipers. The lengths were read to 0.1 mm and are presented as percentages of standard length. The methods of measurement are taken from Bruun (1935: 13), except for new ones used here. I. Standard length. Tip of snout to end of hypural plate. 2. Dorsal and anal-fin hei!lhts. Insertion of longest ray to its tip. 3. Pectoral and pelvic-fin lengths. Insertion of first ray to tip of longest ray. 4. Prepectorallength. Tip of snout to insertion of pectoral fin. 5. Interorbital width. Least clistance across the ossified portion of the or- bits, not including any soft parts. 6. Barbel length. Tip of barbel to point at which the posterior edge of the barbel meets the skin of the lower jaw. 7. Fir~t pectoral ray length. Tip of first well developed pectoral ray to its origin. 8. Di~tance between first and second pectoral rays. Measured at the distal end of the first ray. 9. Distance between the second and third pectoral rays. Measured along the same line as the measurement for the first-second ray distance. 674 Bulletin of Marine Science [15(3) Meristic characters were examined with a binocular dissecting micro- scope. Counts were made according to Hubbs & Lagler (1958: 19-24). 1. Dorsal and anal-fin rays. The last ray is frequently bifurcate to its base, but was always counted as one ray. All elements were counted. 2. Pectoral rays. The first pectoral ray is small and difficult to see and is omitted in counts, conforming with the practice of previous workers. 3. Vertebrae. Counts were taken from specimens stained with alizarin red-S and cleared in glycerin, or from X-rays. The hypural plate was counted as one vertebra. Caudal and precaudal vertebrae were not counted sepa- rately. 4. Predorsal scales. Counted along the mid-dorsal line from end of occiput to dorsal fin origin. 5. Postdorsal lateral scales. Counted mid-laterally from the first row of scales behind the dorsal origin to the hypural plate. 6. Scales above the lateral line. Counted along the diagonal scale row from beside the first dorsal ray to, but not including, the lateral line scale. Palatine teeth were exposed by cutting through the angle of the jaw, and were counted as present only when they could be seen as well as palpated. Strength of palatine dentition was based on the abundance of teeth.

Genus Cypselurus Swainson, 1839 Cypsilurus Swainson, 1839: 187, 296 (original description; illustration; type- species Exocoetus appendiculatus Wood, 1825 = E. comatus Mitchill, 1815; misprint for Cypselurus according to Opinion no. 26 of the Inter- national Commission on Zoological Nomenclature).-Bruun, 1935: 11-13 (description; illustration; key)-Smith, 1950: 125 (description; illustration; key). Cypselurus. Lowe, 1840: 38 (description).-Agassiz, 1846: (correction of Cypsilurus Swainson, 1839 and Cyprilurus DeKay, 1842; fide Breder, 1938: 38).-Jordan & Gilbert, 1882: 380 (description).-Breder, 1928: 23 (synonymy) .-Jordan, Evermann, & Clark, 1930: 201 (listed).- Parr, 1930: 20-25 (description; illustration).-Fowler, 1936: 423, 1240 (description; key; synonymy).-Breder, 1938: 38 (description; illustration; key; synonymy).-Hubbs & Kampa, 1946: 205 (description; key). -Lozano Rey, 1947: 619 (description; illustration; key).-Parin, 1960: 16-19 (description; key) .-Smith, 1961: 125 (description; illustration; key).-Parin, 1961: 163-165 (description; illustration). Cheilopogon Lowe, 1840: 38 (original description; type-species Cypselurus pulchellus Lowe, 1840 = Cypselurus pinnatibarbatus (Bennett, 1831); Madeira) .-Hubbs & Kampa, 1946: 207 (description as subgenus; key; illustration) .-Parin, 1961: 167 (description; key; illustration). Cyprilurus. DeKay, 1842: 231 (misprint for Cypselurus). Ptenichthys MUller, 1843: 312 (original description; type-species Exocoetus !urcatus Mitchill, 1815) .-Parin, 1961: 173-174 (description; key; illustration) . Exonautes Jordan & Evermann, 1896a: 322 (original description; type- species Exocoetus exsiliens (Linnaeus, 1771) ). 1965] Staiger: Atlantic Flyingfishes 675 Ptenonotus Ogilby, 1908: 13 (original description; type-species Exocoetus cirl'iger Peters, 1877).-Ogilby, 1914: 13 (verbatim reprint of 1908 above). Eucypsilurus Bruun, 1935: 84 (original description; type-species Cypsilurus heterurus (Rafinesque, 1810)). Poecilocypsilurus Bruun, 1935: 84 (original description; type-species Cypsil- urus callopterus (GUnther, 1866)). Procypsilurus Bruun, 1935: 84 (original description; type-species Cypsilurus exsiliens (Linnaeus, 1771)). Maculocoetus Whitley & Colefax, 1938: 287 (original description; type- species M aculocoetus suttoni Whitely & Colefax, 1938; Gilbert Islands). -Parin, 1961: 168-170 (description; illustration; key). Abeichthys Parin, 1961: 171-173 (original description; type-species Exocoe- tus agoo (Temminck & Schlegel); key; illustration). Poecilocypselurus. Parin, 1961: 165-167 (description; key; illustration). Procypselurus. Parin, 1961: 168 (description; key; illustration). Diagnosis.-Pectoral fins greater than 45 per cent of standard length (except in very small juveniles, shorter than about 15 mm); reaching at least to the dorsal fin midpoint (except in very small juveniles). Pelvic fins longer than 26 per cent of standard length; reaching at least to anal fin midpoint. Anal fin with shorter base than dorsal fin; beginning well behind dorsal-fin origin; usually with fewer rays than dorsal fin, never more. First pectoral ray simple, second bifurcate; in specimens less than 15 mm standard length all rays usually simple. Anterior dorsal and anal rays longest. Eggs demersal, with long tendrils. Teeth present on both jaws. Adults rec- tangular in cross section. Ontogeny.-In all species certain structures exhibit proportional increase or decrease with growth. Snout length, pectoral-fin length, and greatest depth increase in size proportional to standard length. Length of head, pelvic-fin, and barbel, height of dorsal and anal-fins, and diameter of fleshy orbit all decrease in size proportionally to standard length. The barbel length at first increases in proportion to standard length, but after a maximum length is reached at a standard length of about 50 mm it decreases until it disappears. The shape of the pectoral fin changes as its length increases. In small juvenHes the fin is short and wide, giving a somewhat rounded appearance. As the fish grows the fin increases in length and decreases in width proportionally, ultimately yielding the pointed fin of the adult. Range.- The species of Cvpselurus occur in the tropical and temperate waters of the Atlantic, Pacific, and Indian Oceans. Only two of the Atlantic species, C. exsiliens and C. turcatus, occur in mid-ocean; C. comatus, C. cyanopterus, C. heterurus, and C. pinnatibarbatus are found within 400 miles of land. This requirement for land can be met by small islands 676 Bulletin of Marine Science [15(3) as well as by the coasts of continents. Cypselurus is reported as far north as 42°N in the western North Atlantic (c. heterurus and C. exsiliens; BU specimens) and 59°N in the eastern North Atlantic (C. heterurus; Bruun, 1935: Oslo Fjord), and as far south as 23°S in the western South Atlantic (C. exsiliens; Bruun, 1935: 95) and 35°S in the eastern South Atlantic (c. pinnatibarbatus; Bruun, 1935: 95).

Discussion.-Parin (1961) recently has revised the genera of Exocoetidae. He divides Cypselurus among two genera. Of the Atlantic species Cypse- lurus comatus is retained, but the other five species are referred to Cheilo- pogon Lowe, 1840. Cypselurus cyanopterus and C. exsiliens are placed in the subgenus Procypsilurus Bruun, 1935, Cypselurus pinnatibarbatus is the only species in the subgenus Cheilopogon, and Cypselurus furcatus and C. reterurus are placed in the subgenus Ptenichthys MUller, 1843. Cypselurus is characterized by Parin as having a comparatively deep and wide body, with maximum depth 3.9-6.0 times in body length (16.7-25.6 per cent of standard length) and maximum width 4.1-7.4 times in body length (13.5-24.4 per cent of standard length). Cheilopogon is characterized as being relatively shallow and narrow, with maximum depth 5.0-7.0 times in body length (14.3-20.0 per cent of standard length) and maximum width 6.5-9.3 times in body length (10.8-15.4 per cent of standard length). My data extend the range of body depth in Cheilopogon to 13.1-20.7 per cent of standard length and its width range to 10.3-17.3 per cent of standard length. Thus there is considerable overlap between the two genera in regard to these two characters. Parin uses the length of the pectoral fin as a differentiating character also. He states that in Cypselurus the pectoral fin extends just to the end' 'Ofthe dorsal-fin base, while in Cheilopogon it extends beyond the end of the dorsal-fin base and sometimes beyond the end of the caudal-fin base. A perusal of the descriptions of the different subgenera of Cheilopogon in Parin's paper demonstrates that several of these groups (Maculocoetus, Abeichthys, and Ptenichthys) possess pectoral fins that reach nearly to the end of the dorsal-fin base and sometimes beyond it. This is essentially the condition described for Cypselurus. There would appear to be, therefore, a gradation from the pectoral fin configuration of Cypselurus into the configuration of Cheilopogon. In Cypselurus the first pectoral ray is simple and the rest are branched. In most species of Cheilopogon the same condition exists, but in the single 'species of the subgenus A beichthys the first two pectoral rays are simple. Another supposed difference between the two genera lies in their ontogeny. The "Parexocoetus stage" is a developmental stage in which the young have a high black dorsal fin. The young of Cypselurus have no "Parexocoetus stage," while the young of all the subgenera of Cheilopogon except Ptenichthvs have this stage. Adults of Cypselurus have a low, gray 1965] Staiger: Atlantic FLyingfishes 677 dorsal fin, while adults of CheiLopogon have low to medium-height dorsal fins that are either uniformly gray or have a dark posterior spot. In other words, the dorsal fin of adults of some species of Cheilopogon looks like that of adults of CypseLurus. The only definite difference, therefore, is the presence of the "Parexocoetus stage," which is absent in the subgenus Ptenichthys of Cheilopogon, as well as in the genus CypseLurus. The anal fins of both CypseLurus and Cheilopogon originate under the third to ninth dorsal rays and have from 0-5 fewer rays than the dorsal fin. Parin also used the secondary canals of the lateral-line system as a differentiating character. CypseLurus has a few short secondary canals on each lateral-line scale. These canals are in a single layer and do not anastomose. In Cheilopogon the secondary canals are either rare or abundant. If abundant then they are in one or a few layers and often form branches that anastomose. The condition in CypseLurus is therefore intermediate between the two configurations in Cheilopogon. In CypseLurus the jaw teeth are of medium size, while those of Cheilo- pogon may be either small or fairly large. Cypselurus is, therefore, intermediate between the two conditions in Cheilopo{?on. According to Parin, the jaw teeth of CypseLurus usually have lateral cusps, which are often large enough to make the teeth appear tricuspid. In Cheilopogon the teeth are unicuspid, but may have supplementary crests. Parin, however, states (1961: 130) that the number of tricuspid teeth in CypseLurus may be small, although lateral crests are distinct on most of the teeth. For the subgenus Ptenichthys of CheiLopogon he states (1961: 131) that the teeth are frequently provided with supplementary cusps. Parin's illustrations (1961: 132-133) depict a much mOre extreme difference in cusp development between species of Cypselurus and Cheilopogon than is indicated by Bruun's illustrations (1935). This character thus appears to grade from one nominal genus into the other, and does not provide generic distinction. There are no differences between the genera in number of gill rakers, morphology of the skull, and morphology of the eggs. The young of CypseLurus have one median barbel or none, while young of Cheilopogon have two narrow barbels or one wide fringed one. A division on this basis is strictly arbitrary. In every character used by Parin to distinguish CypseLurus from Cheilo- pogon, CypseLurus is either intermediate, broadly overlapping, or similar to CheiLopogon. Recognition of more than a single genus seems unwar- ranted, and all species are herein included in the genus Cypselurus.

KEY TO THE SPECIES I. Key to specimens shorter than 100 mm standard length: Ia. Dorsal fin usually uniformly pigmented: fin height greater than 20 per cent of standard length. . 2 678 Bulletin of Marine Science [15(3) Ib. Dorsal fin either without pigment or with pigment appearing as a number of blotches on a clear background: fin height less than 20 per cent of standard length. . 4 2a. Barbel single, semicircular, subdivided into 15-20 fringe-like appendages. 39-47 predorsal scales. 50-52 vertebrae...... ~ Cypselurus pinnatibarbatus 2b. Barbels paired, not divided into fringe-like appendages. 21-41 predorsal scales. 43-46 vertebrae. . 3 3a. Barbels longer than 60 per cent of standard length. Both caudal lobes darkly pigmented. Pectoral fin as in Figure 6, A-B. Rear edge of dorsal fin rounded. 30-41 predorsal scales...... Cypselurus cyanopterus 3b. Barbels shorter than 40 per cent of standard length. Upper caudal lobe very lightly pigmented. Pectoral fin as in Figure 8, A-D. Rear edge of dorsal fin straight and nearly vertical to body. 21-30 predorsal scales. . Cypselurus exsiliens 4a. Barbel single, longer than 40 per cent of standard length. Pectoral and pelvic fins evenly pigmented, without pattern, as in Figure 2a. Dorsal fin unpigmented. Caudal fin without pigment or evenly, lightly pigmented. . Cvpselurus comatus 4b. Barbels paired, shorter than 40 per cent of standard length. Pectoral and pelvic fins with patterns of dark spots and bands, as in Filwres 14, 15, 17, and 18. Dorsal fin with dark blotches. Caudal fin with a number of pigment blotches on both lobes .... 5 5a. First pectoral ray longer than 38 per cent of standard length. Distance between the first and second pectoral rays (measured at the end of the first) less than 1.7 times the distance between the second and third. Barbels longer than 16 per cent of standard length (sometimes shorter in specimens less than 35 mm in standard length). Pectoral-fin coloration as in Figure 14, A-F with the posterior edge of the clear band forming a straight line at right angles to the body. . Cypselurus furcatus 5b. First pectoral ray shorter than 38 per cent of standard length. Distance between first and second pectoral rays (measured at end of first) greater than 1.7 times the distance between the second and third. Barbels shorter than 16 per cent of standard length. Pectoral-fin coloration as in Figure 17, A-F, with the posterior edge of the clear band not forming a straight line (if a straight line is formed then the sides of the clear band converge toward the edge of the fin). . Cypselurus heterurus

I I. Key to specimens longer than 100 mm standard length: 1a. A prominent dark spot on the posterior half of the dorsal fin 2 Ib. No dark spot on dorsal fin. . 4 1965] Staiger: Atlantic Flyingfishes 679 2a. Pectoral fin dark, with obvious pale crossband, as in Figure 8, D or E. Upper caudal lobe much lighter than lower. 21-30 predorsal scales. . Cypselurus exsiliens 2b. Pectoral without a pale crossband, or with a vague crossband. Caudal lobes uniformly pigmented. 30-46 pre dorsal scales 3 3a. Pectoral fin bluish or brown in preservation; without an obvious pale posterior margin; without a pale crossband, as in Figure 6C. 30-41 predorsal scales. 43-46 vertebrae. .. Cypselurus cyanopterus 3b. Pectoral fin grayish in preservation; with an obvious pale posterior margin; with a vague crossband in some specimens. 39-46 predorsal scales. 50-52 vertebrae. . Cypselurus pinnatibarbatus 4a. Pectoral fin uniformly pigmented, as in Figure 2 B. Palatine teeth present. Premaxillary teeth tricuspid. . Cypselurus comatus 4b. Pectoral fin with a pale crossband. No palatine teeth. Premaxillary teeth unicuspid. . 5 5a. Pectoral grayish in preservation. Pale pectoral crossband widest at mesial edge, in specimens longer than 150 mm standard length ending well before anterior edge and triangular in shape, as in Figure 17, F-H. Pectoral with a narrow pale posterior margin. First pectoral ray generally shorter than 38 per cent of standard length. . Cypselurus heterurus 5b. Pectoral fin nearly black in preservation. Pale pectoral crossband widest at anterior edge, always reaching to, or nearly to, the anterior margin, not triangular, as in Figure 14 G. Pectoral with a wide pale posterior margin. First pectoral ray generally longer than 38 per cent of standard length. . Cypselurus furcatus

Cypselurus comatus (Mitchill, 1815) Clearwing Flyingfish Figures 1, 2, and 3. Table 1. Exocoetus comatus Mitchell, 1815: 448-449 (original description; illustrated; New York).-Valenciennes, in Cuvier & Valenciennes, 1846: 133-135 (description; South Carolina) .-Gunther, 1866: 286 (listed; South At- lantic).-Liitken, 1876: 399 (fide Breder, 1938: 51). Exocetus comatus. DeKay, 1842: 231 (description; illustration; New York). -Storer, 1846: 188 (description; New York). Exocoetus appendiculatus Wood, 1825: 283 (original description; illustration) . Exocaetus bahiensis Ranzani, 1842: 362-365 (original description; fide Bruun, 1935: 6, 52). Exocoetus heterunts. Jordan & Meek, 1886: 59-61 (misidentification in part; description; key) .-Jordan & Evermann, 1896b: 735-736 (misidentification in part; description). Cypselurus vitropinna Breder, 1927: 20-21 (original description; illustration: West Indies).-Parr, 1930: 25, 27 (description; Bahamas).-Breder & 680 Bulletin of Marine Science [15(3)

\0 """ V) ..... '

'- ~ ~ :-9 U ;:!'" ~ :::'" '"::\ Z c cu ..::\ III ~ .~ ~ ::: '§ cu.. ::: ~ l:: <::l .~ ::: CI c ;>., - ..::\ ti .~ III ~ -::: i>: '" '" ~ cj cj cj -.cj cj cj 1965] Staiger: Atlantic Flyingfishes 681 Nichols, 1930: 7-9 (description).-Nichols & Breder, 1930: 5,7 (key; synonymy) .-Breder & Nichols, 1934: 40, 42, 44 (description; illustration). Cypselurus antarei Beebe & Hollister, 1933: 83-85 (original description; illustration; 21°50'N, 63°32'W). Cypsilurus comatus. Bruun, 1933: 382 (listed) .-Bruun, 1935: 52-56, pi. 2 (description; illustration; key; Atlantic). Cypselurus comatus. Breder & Nichols, 1934: 40, 42 (discussion).-Breder, 1938: 51-55 (description; key; illustration; Western Atlantic) .-Parin, 1961: 165 (listed). Cypselurus alienus. Herre, 1935: 392-393 (original description; South coast of Cuba) .-Herre, 1936: 11, 13 (listed; illustration; West Indies). Material examined.-Gulf Stream: UMML (4, 169.5-179.7 mm) Ba- hamas, 5 mi. W of Bimini, 6 July 1961.-UMML (3, 182.0-204.4 mm) Bahamas, off Cat Cay, 21 May 1962.-UMML 13771 (2,31.1-46.5 mm). -UMML (1, 43.8 mm) 25°20'N, 80000'W, 29 August 1962.-UMML (1, 96.2 mm) 22°35'N, 78°08.5'W, 29 November 1951.-BLBG (1, 43.8 mm) 28°19'N, 79°26'W, 24 April 1953-BLBG (1, 18'0.8 mm) 27°40'N, 79°18'W, 23 June 1954.-BLBG (1, 60.1 mm) 30021'N, 79°26'W, 27 June 1954.-UMML 13770 (2, 34.7-55.5 mm).-BLBG (1, 27.4 mm) 27°40'N, 79°18'W, 12 September 1954.-USNM 94722 (1, 31.4 mm) Cuba. Bahamas: BLBG (1, 40.4 mm) 26°25'N, 76°40'W, 13 June 1954.- BLBG (1,29.4 mm) 26°25'N, 76°40'W, 24 January 1954.-BLBG (1, 60.1 mm) 20043'N, 73°33'W, 22 December 1961. Atlantic Ocean: BU (1,82.4 mm) 20oIO'N, 60005'W, 10 December 1956. Caribbean Sea: BU (1, 151.0 mm) between Grand Cayman and Swan Islands, 15 January 1950.-USNM 170321 (3, 182.8-195.1 mm). Gulf of Mexico: BU (1, 242.5 mm) 20050'N, 93°00'W, 25 November 1956. Distinctive characters.-Juveniles of Cypselurus comatus may be distinguished from Atlantic congeners by their single, simple median barbel. Most other species have paired barbels. C. pinnatibarbatus has a median barbel, but it is short and fringed. The adults are distinguished by their low, clear dorsal fin (held in common with C. heterurus and C. furcatus), uniformly pigmented pectoral fins (unique within the group with a low dorsal fin), and tricuspid premaxillary teeth. No other Atlantic species of Cypselurus has teeth as obviously tricuspid as does C. comatus. Description.---:-Cypselurus comatus is a small species. The maximum size recorded is 242.5 mm standard length, but the average adult seems to be shorter than 200 mm standard length. C. cyanopterus, a large species, has a maximum recorded size of 333 mm standard length (Bruun, 1935: 40) . The general body forms of the juveniles and adults are illustrated in Figure 1. 682 Bulletin of Marine Science [15(3)

FIGURE 1. Cypselurus comatus. A, Juvenile 96.2 mm SL.-B, Adult 173.5 mmSL.

Palatine teeth are weak, i.e., small and few in number in all specimens examined. All jaw teeth examined had two extra cusps which were shorter than the central cusp, but much larger than the accessory cusps of the other species. Meristic data (range of study material 27-242 mm standard length, extremes in parentheses from the literature): dorsal-fin rays 10-14; anal-fin rays 8-9; dorsal minus anal-fin rays 1-4(5); pectoral-fin rays 12-16; predorsal scales 22-30 (33); postdorsal lateral scales 14-21; total gill rakers (first arch) 18-23 (24); scales above lateral line 5-8; vertebrae 40-43 (44). Morphometric data (expressed as per cent of standard length, range of study material 27-242 mm standard length, extremes in parentheses from the literature): preanal length 77.9-80.8 (81.5) per cent; predorsallength 65.4-76.8 per cent; prepelvic length 58.8-66.3 per cent; prepectorallength 23.8-26.5 per cent; head length 23.7-27.6 per cent; snout length 3.7-7.0 per cent; fleshy orbit diameter 7.1-9.9 per cent; interorbital width (7.5) 7.7-9.8 per cent; pectoral-fin length 57.5-70.9 per cent; pelvic-fin length 25.6-41.3 per cent; dorsal-fin height 7.3-15.6 per cent; anal-fin height 6.0-12.9 per cent; dorsal-fin base 17.3-19.3 per cent; anal-fin base 9.0-11.4 1965] Staiger: Atlantic Flyingfishes 683 per cent; greatest depth (17.8) 18.1-21.2 per cent; caudal peduncle depth 6.8-8.1 per cent; body width (13.1) 13.9-15.3 per cent; depth behind eyes 15.4-16.1 per cent. Pectoral fin grayish, pigment denser distally. Posterior margin and last few rays pale. Pelvic fins grayish, the pigment denser distally. Dorsal fin lightly pigmented along the rays with no pattern. Anal fin clear. Caudal fin quite heavily pigmented in adults. Body of adults light silver ventrally, dark dorsally. Juvenile coloration is discussed in the next section. Ontogeny.-The young of C. comatus have a single, median mandibular barbel (see Fig. 2) that is fleshy, with two lateral folds that extend for its

A B [:

·f/" ..··· -.-.'"

D E FIGURE 2. Cypselurus comatus. A-B, Pectoral fin, 31.1 mm SL, base line 2 mm; 191.8 mm SL, base line 10 mm.-C, Barbel, 71.1 mm SL, base line 5 mm.- D-E, Pelvic fin, 31.1 mm SL, base line 1 mm; 204.4 mm SL, base line 1 mm. entire length, and tapers distally, to a point. The barbel undergoes a spectacular development. A specimen of 55.5 mm standard length had a barbel of 100.8 per cent of standard length. Breder (1938:52, Table 14) lists a specimen of 71 nun standard length with a barbel 102.8 per cent of standard length. This appears to be its maximum relative size. The relative length of the barbel decreases up to about 160 mm standard length and then disappears, with no barbel reported for specimens over 170 mm standard length. The structure of the barbel does not change with growth. The development and regression of the barbel is shown in Figure 3. 684 Bulletin of Marine Science [15(3) 160 0 0 0 140 0 0 0 0 0 00 0 120 00 0 o 0 0 0 00 100 • 0 • 0 -.i 0 0 vi 0 0 80 0 0~ 00 .J 0 0 q; • -e 60 0 •• • • co • • 0 CD It, • • 40 • t, • 0 • IJ. • t.lJ. •• Ii' 0 0 C:. 20 b. 0 C:.r::, b. Ab.tP

0 t, f\ I ..---. 0 20 40 a:> 80 100 120 140 160 l80 200 S. L. in mm FIGURE 3. Development and regression of barbel in relation to SL. Solid circles, Cypselurus coma/us; open circles, C. cyanopterus; triangles, C. exsiliens.

The dorsal fin undergoes a reduction in height with growth, as in all the species. The pelvic fins become relatively smaller with growth. Speci- mens smaller than 150 mm standard length have pelvic fins longer than 35 per cent of standard length while in larger specimens they are less than 35 per cent of standard length. In juveniles the fins are fairly heavily pigmented, darkest posteriorly (Fig. 2 D). With growth this pigment disappears, so that by 160 mm standard length they are clear except for some pigment along the proximal portions of the rays and along the anterior edge of the fin (Fig. 2 E). The pectoral fin maintains the same pigmentation throughout development (Fig. 2 A and B). The caudal fin is pale in juveniles, but darkens with growth. In a specimen of 82.4 mm standard length the caudal was almost devoid of pigment, in a specimen of 151.0 mm standard length it was darker, and in a specimen of 242.5 mm standard length it was nearly black, with both lobes uniformly pigmented. Above 45 mm standard length the juvenile body color pattern is like the adult, with a silvery ventral surface and dark dorsal surface. Below 45 mm standard length the body is rather uniformly dark. 1965] Staiger: Atlantic Flyingfishes 685 Range.-Cypselurus comatus is found in the tropical Atlantic Ocean throughout the Gulf of Mexico and the Caribbean Sea. Most specimens have been collected off Florida, the Bahamas, and the Antilles. Bruun (1935: 56, 96, text-fig. 28) reported one specimen from the coast of Brazil at 11oS, 36°W. All the specimens examined were collected within 300 miles of land. C. comatus can therefore be considered a tropical, coastal species. Reproductive biology.-According to Bruun (1935: 55) sex generally could not be determined in fish shorter than 170 mm standard length and the smallest ripe fish he identified was 178.5 mm in standard length. The breeding season extends from February to April (Breder, 1938: 52). Ovarian eggs are about 1.3 mm in diameter and have filaments, as do those of the other species of Cypselurus (Breder, 1938: 52 fig. 27).

Cypselurus cya:;opterus (Valenciennes, 1846) Margined Flyingfish Figures 3, 4, 5, and 6. Table 1

Exocoetus cyanopterus Valenciennes, in Cuvier & Valenciennes, 1846: 97-98 (original description; Brazil) .-GUnther, 1866: 294 (listed; coast of Brazil).-GUnther, 1869: 239 (listed; Madeira).-Melliss, 1875: 110 (St. Helena).-Jordan & Meek, 1886: 63 (description).-Jordan &. Ever- mann, 1896b: 739 (description). Exocoetus vermiculatus Poey, 1860: 300-301 (original description; Cuba). Exocoetus albidactylus Gill, 1863: 167 (original description; Caribbean). Exocoetus bahiensis. Gunther, 1866: 293 (Atlantic Ocean) .-Jordan & Meek, 1886: 64 (description; key).-Jordan & Evermann, 1896b: 739- 740 (description; key).--Osorio, 1898: 199' (listed; Annobon Island).- Norman, 1935: 56 (listed). Exocoetus parrae Poey, 1868: 385 (original description; Cuba). Exocaetus bahiensis. Osorio, 1895: 246 (listed; Annobon Island). Cypsilurus bahiensis. Jordan & Evermann, 1898: 2836 (listed) .-Jordan & Seale, 1906: 211 (listed; Samoa).-Smith, 1950: 126 (description; illustration; key; South Africa). Cypsilurus cyanopterus. Jordan & Evermann, 1898: 2836 (listed) .-Ever- mann & Marsh, 1900: 104 (description; Puerto Rico).-Bruun, 1935: 40- 44 (description; illustration; key; Atlantic Ocean).-Norman, 1935: 58 (listed; South Atlantic) .-Marshall, 1950: 170 (description; Cocos-Keel- ing lslands).-Poll, 1953: 188-190 (description; illustration). Cypselurus bahiensis. Breder, 1927: 22-23 (description) .-Jordan, Ever- mann, & Clark, 1930:202 (listed).-Nichols & Breder, 1930:1, 5, 7-8 (description; illustration; Caribbean Sea).-Parr, 1930:26 (description; Ba- hamas).-Fowler, 1931: 319 (listed; Fiji, Santa Cruz Islands).-Breder & Nichols, 1934: 40, 44 (discussion; illustration) .-Fowler, 1934a: 432 (listed; Natal) .-Fowler, 1934b: 394 (listed, Santa Cruz Islands, New Hebrides).-Fowler, 1936: 431-432 (description; Cape Verde Islands).- Smith, 1961: 126 (description; illustration). CypseIurus monroei Nichols & Breder, 1928: 432-433 (original description; 686 Bulletin of Marine Science [15(3) illustration; Coconut Grove, Florida).-Breder Nichols 1930: 1-3 (description; illustration; South of Dry Tortugas, Florida) .-Nichols & Breder, 1930:5,7 (key; synonymy).-Breder & Nichols, 1934:40, 42, 44 (discussion; illustration). Cypselurus lineatus. Nichols & Breder, 1930: 3-4, 6, 8 (description; illustration; synonymy; Cape Verde Islands; misidentification). Cypselurus cyanopterus. Jordan, Evermann, & Clark, 1930: 202 (Iisted).- Fowler, 1936: 1244 (description; Canary Islands) .-Breder, 1938: 39-46 (description; key; illustration; North Atlantic) .-Parin, 1960: 44-45 (description; key; Northwestern Pacific Ocean) .-Cadenat & Marchal, 1963: 1254-1255 (description; Ascension Island). Cheilopogon cyanopterus. Parin, 1961: 168 (listed). ? Exocoetus solitarius. Bennett, 1860; 15 (original description: St. Helena). Material examined.-Gulf Stream: BLBG (1, 22.1 mm) 34°34'N, 75°40'W, 24 July 1960.-BLBG (1, 80.1 mm) 29°56'N, 8001O'W, 12 May 1959.-BLBG (2, 48.0-72.6 mm) 32°53'N, 77°04'W, 8 November 1953.-UMML 13772 (3, 19.5-68.3 mm).-BLBG (1,96.2 mm) 30°18' N, 80012'W, 29 July 1963.-BLBG (1, 90.3 mm) 25°11'N, 79°56'W, 15 May 1959.-BLBG (5,29.6-60.8 mm) 34°34'N, 75°40'W, 15 August 1961.-BLBG (1,24.3 mm) 35°00'N, 75°19'W, 18 July 1960.-BLBG (1, 21.5 mm) 32°39'N, 76°46'W, 10 August 1953.-BLBG (2, 20.9- 22.6 mm) 29°55'N, 80038'W, 12 June 1960.-BLBG (1, 37.3 mm) 31°00'N, 79°25'W, 14 September 1954.-UMML 12773 (1,113.7 mm). -MCZ 34712 (1, 146.5 mm).-BU (1,68.6 mm) 34°45'N, 73°41'W, 10 October 1957.-UMML 13774 (2, 25.1-25.2 mm).-BU (1, 80.6 mm) 37°45'N, 79°41'W, 7 October 1957.-BU (1, 171.5 mm) 36°42'N, 7'ooOO'W, 3 October 1957.-BU (1, 175.2 mm) ca. 39°N, 68°W, 23 Sep- tember 1957.-BU (1, 179.6 mm) 39°07'N, 65°58'W, 21 September 1957.-BU (1, 82.6 mm) 39°28'N, 69°30'W, 30 October 1956.-UMML 13779 (1,41.5 mm).-UMML 13780 (1,44.7 mm). Caribbean Sea: UMML 13777 (1, 42.5 mm). Gulf of Mexico: UF 3857 (1, 88.7 mm) .-UF 7245 (2, 108.7-241.0 mm).-UF 3673 (3, 225.7-241.0 mm).-BU (1, 25.1 mm) 28°55'N, 87°57'W, 27 February 1959.-USNM 158847 (1 ).-UMML 13778 (1,73.4 mm).-UMML 13775 (1,15.7 mm).-UMML 13776 (1, 103.1 mm).-BU (1, 286.9 mm) 20012'N, 91°59'W, 12 July 1957.-BU (1, 200.9 mm) ca. 21°N, 91°W, 21 November 1956.-BU (1, 187.9 mm) 20050'N, 93°00'W, 25 November 1956.-BU (3, 199.6-223.7 mm) 20010'N, 92°25'W, 24 November 1956.-BU (1, 37.9 mm) 23°10'N, 92°00'W, 11 May 1954.-BU (1, 24.6 mm) 21°42'N, 93°35'W, 22 May 1954. Spain: BU (1,297.1 mm) 37°09'N, 9°42'W, 28 November 1957. Distinctive characters.- The most distinctive juvenile character is the extreme development of the paired mandibular barbels, which exceed 100 per cent of standard length in specimens from 20 to 83 mm standard 1965] Staiger: Atlantic Flyingfishes 687 length (longer than in any other Atlantic species of Cypselurus). The high dorsal fin (9.4-34.2 per cent of standard length) and bandless pectoral fins are also characteristic. Adults are distinguished by a combination of high, pigmented dorsal fin, evenly pigmented caudal fin, evenly pigmented blue pectoral fins, and low number of vertebrae (43-46). The high dorsal fin separates C. cyanopterus from C. comatus, C. furcatus, and C. heterurus. The presence of pigment on both caudal lobes distinguishes it from C. exsiliens, as do the uniformly pigmented pectoral fins. The low number of vertebrae distinguishes it from C. pinnatibarbatus.

Description.-Cypselurus cyanopterus is one of the larger Atlantic flyingfishes. The largest specimen I examined was 297.1 mm standard length; Bruun (1935: 40) lists the largest specimen he examined as 333 mm standard length. The general body form of the juveniles and adults is illustrated in Figure 4. Palatine teeth were present in all specimens examined, varying considerably in size and number. The strength of palatine dentition, previously used as a distinguishing character, is too variable to be valid. Premaxillary teeth are somewhat curved and sometimes possess one or two protuberances

FIGURE 4. Cypselurus cyanopterus. A, Juvenile 73.4 mm SL.-B, Adult 225.7 mm SL. 688 Bulletin of Marine Science [15(3)

-i ~ vi 25 . t:J." ~ ~ D o CT.t:J. f1 •• i20 .~~~.o a Ot:J. J~~ a ~ 15 a co O~ 0 % <5 • 00 00 ••• o 0 00 rO:> '1:) 0 o· .... o • o 10 0 0 0000 0 • o 00 o 0 •

5 o 50 100 150 200 250 300 350 S.L. in mm FIGURE 5. Dorsal-fin height in relation to SL. Solid circles, Cypselurus cyanopterus; triangles, C. exsiliens; open circles, C. heterurus. beside the main point, but these do not compare in size or frequency of occurrence with those seen in C. comatus. Dorsal fin higher than 12 per cent of standard length (see Fig. 5), except in some specimens larger than 200 mm standard length, with a rounded posterior edge; darkly pigmented in all stages of development, but the position of the pigment changes with ontogeny. Meristic data (range of study material 15-297 mm standard length, extremes in parentheses from the literature): dorsal-fin rays 11-14; anal-fin rays 8-11; dorsal minus anal-fin rays 1-5; pectoral-fin rays 12-16; predorsal scales 30-41; postdorsal lateral scales 16-26; total gill rakers (first arch) 16-28; scales above lateral line 6-9; vertebrae 43-45 (46). Morphometric data (expressed as per cent of standard length, range of study material 15-297 mm standard length, extremes in parentheses from the literature): preanal length 72.0-80.7 per cent; predorsallength 63.1- 74.5 per cent; prepe1vic length 56.0-61.0 (61.6) per cent; prepectoral length 20.1-25.2 per cent; head length 22.0-27.4 per cent; snout length 3.0-6.7 per cent; fleshy orbit diameter 4.9-8.5 per cent; interorbital width 6:0-10.1 per cent; pectoral-fin length 45.9-78.5 per cent; pelvic-fin length (26.3) 26.5-50.2 per cent; dors'l1-fin hei~ht 9.4-35.2 per cent; anal-fin height 5.6-16.6 per cent; dorsal-fin base 18.5-23.2 per cent; anal-fin base 11.6-15.7 per cent; greatest depth 13.1-19.3 per cent; caudal peduncle depth 6.0-8.4 per cent; body width 10.3-15.9 per cent; depth behind eyes 10.3-12.7 per cent. 1965] Staiger: Atlantic Flyingfishes 689 Pectoral fins darkly pigmented, not banded; posterior few rays usually devoid of pigment. Pigment density changes with ontogeny. Pelvic fins dark, the pattern changing with ontogeny. Body generally dark dorsally, silver ventrally. No vertical dark banding in juveniles, as seen in C. jllrcatus, C. heterurus, C. exsiliens, and C. pinnatibarbatlls.

Ontogeny.- The most striking ontogenic change undergone by this species is the development and regression of the mandibular barbels. The barbels are long and fleshy, tapering evenly to their ends (see Fig. 8 D). Each barbel has a pair of pigmented folds laterally along its entire length, with the lateral fold about twice as wide as the mesial one. The smallest specimen examined (15.7 mm standard length) possessed barbels that were 85.1 per cent of standard length. The barbels increase proportionally until the fish is about 70 mm standard length (in a specimen of 68.3 mm standard length the barbels were] 54.2 per cent of standard length). After this the barbels decrease in length proportionally until they are lost between 190 and 200 mm standard length. The largest specimen reported with barbels was ]90.0 mm standard length with barbels 24.9 per cent of standard length (Bruun, 1935: 41, Table 28). See Figure 3 for a comparison of the development with C. comatlls and C. exsiliens. The pelvic fins become relatively shorter with development. The proportions range from 50.2 per cent of standard length in a 44.7 mm specimen, 39.1 per cent for 103.1 mm, 31.8 per cent of standard length for 199.6 mm, to 27.1 per cent of standard length for a specimen 297.1 mm long. The pelvics also undergo a color change with ontogeny. In small specimens (ca. 30 mm standard length) the fin is faintly banded (see Fig. 6 E). There are three dark bands; a narrow one at the fin root, a wide one in the middle of the fin, and a narrow one along the posterior edge of the fin. The background between these dark arcs or bands is lightly pigmented. As the fish grows these dark bands disperse so that the background becomes darkly pigmented. This pigment is concentrated primarily on the fin rays and next to them, so that the fin membrane is clear in many places (Fig. 6 H). With further development most of these clear areas disappear leaving the fin almost completely dark, as in Figure 6 G. At about 100 mm standard length the dark pigment begins to withdraw from the proximal end of the fin, resulting in a pattern like Figure 6 H. By ] 70 mm standard length the proximal third of the fin is clear, and this clearing continues until the entire fin is clear by 280 mm standard length. The dorsal fin decreases in height proportionally with development. Figure 5 compares this decrease with those for C. exsiliens and C. furcatus. In small juveniles, about 20 mm standard length, the dorsal fin is about 14 per cent of standard length. This proportion increases with growth until by about 100 mm standard length the fin is at its highest-ca. 25 per cent of 690 Bulletin of Marine Science [15(3)

F ~,-~ G

FIGURE 6. Cypselurus cyanopterus. A-C, Pectoral fin, 29.6 mm SL, base line 2 mm; 58.7 mm, base line 5 mm; 225.7 mm SL, base line 25 mm.-D, Barbel, 113.7 mm SL, base line 10 mm.-E-H, Pelvic fin, 29.6 mm SL, base line 1 mm; 58.7 mm SL, base line 5 mm; 73.4 mm SL, base line 5 mm; 103.1 mm SL, base line 5 mm. standard length. By 175 mm standard length the fins have decreased to about 18 per cent of standard length and by 250 mm standard length they are only apout 12 per cent of standard length. The dorsal-fin pigmentation also changes with growth. In specimens shorter than 100 mm standard length the entire fin is dark. This pigment is gradually lost from the anterior part of the fin, until in the largest specimens the pigment is 1965] Staiger: A~lantic Flyingfishes 691 concentrated in a spot on the posterior margin. The posterior profile of the fin is rounded in juveniles and small adults, but by 300 mm standard length this profile becomes straight. The pectoral fin does not develop a band as is seen in C. heterurus and others, but there is a definite change in pattern with ontogeny. Figure 6, A-C illustrates this change. In small juvenile specimens the fin is rather uniformly lightly pigmented, except for a clear area along the posterior edge of the fin (Fig. 6 A). With development this pigmentation darkens and a number of clear areas appear on the fin, especially along the anterior edge (Fig. 6 B). As further development occurs these clear areas disappear and the fin assumes an overall dark pigmentation, with a small clear margin along the mesial edge of the fin (Fig. 6 C). The caudal fin is weakly pigmented in small juveniles. By about 45 mm standard length the fin has developed a dark background with a number of clear spots on it. These spots diminish in size until the fin is uniformly pigmented at about 150 mm standard length. Range.-Cypselurus cyanopterus is found in tropical waters on both sides of the Atlantic Ocean. It occurs throughout the Gulf of Mexico and the Caribbean Sea, and in the Gulf Stream to 39°N. It has been collected in the eastern Atlantic at the Canary Islands, off the coast of Spain (37°N), and in the Gulf of Guinea. In the western South Atlantic it occurs as far as 19°5, off Brazil. Like C. comatus and others of the genus, C. cyanopterus can be considered a coastal species, in that it occurs no farther than 400 miles off shore. It has been taken at St. Peter and St. Paul Rocks, which bears out Bruun's contention that even small oceanic islands can provide the requirements for a coast that these flyingfishes seem to need (Bruun, 1935: 96, text-Fig. 27). Reproductive biology.-Above 250 mm standard length sex may be determined easily (Bruun, 1935: 44). No ripe specimens have been reported.

Cypselurus exsiliens (Linnaeus, 1771) Bandwing Flyingfish Figures 3, 5, 7, 8, and 9. Table 1. Exocoethus exsiliens. Linnaeus, 1771: 529 (original description; Carolina). Exocoetus exiliens. MUlier, 1776: 209 (Carolina).- Gmelin, 1789: 1400 (description, Carolina) .-Gmelin, 1818: 325-326 (description; illustration).-Valenciennes, in Cuvier & Valenciennes, 1846: 114-115 (description; New Jersey).-GUnther, 1866: 219 (description; key; Atlantic Ocean).-Jordan & Gilbert, ]882: 380 (description).-Jordan & Meek, ]886: 54-55 (description).-GUnther, 1909: 147 (description of type specimen) . Exocoetus bicolor Valenciennes, in Cuvier & Valenciennes, 1846: ]]] -] ]2 (origin a] description; Atlantic Ocean). 692 Bulletin at Marine Science [15(3) Exocetus exiliens. Storer, 1846: 188 (description; New York). Exocoetus nigricans. GUnther, 1866: 290 (description; Java).-Jordan & Meek, 1886: 63-64 (description).-Jordan & Evermann, 1896b: 737-738 (description) . Exocoetus melanurus. Jordan & Gilbert, 1882: 379-380, 904 (description; synonymized with E. exiliens p. 904). Exocoetus exsiliens. Lilljeborg, 1891: 466-472 (description) .-Jordan & Evermann, 1896b: 732 (description). Exonautes exsiliens. Jordan & Evermann, 1896a: 322 (new genus; type -species Exocoetus exsiliens) .-Jordan & Evermann, 1898: 2836 (listed). Cypsilurus nigricans. Jordan & Evermann, 1898: 2836 (listed). Cypselurus nigricans. Nichols, 1924: 63 (description; Galapagos Islands).- Jordan, Evermann, & Clark, 1930: 201 (listed). Cypselurus smithi Breder & Nichols, 1930: 4-6 (original description; illustration; North of the Bahamas) .-Nichols & Breder, 1930: 5, 7 (key; synonymy) .-Breder & Nichols, 1934: 40, 42, 44 (description; illustration). Cypsilurus exsiliens. Bruun, 1935: 44-47 (description; key; illustration; At- lantic Ocean). Cypselurus exsiliens. Fowler, 1936: 424-425, 1241 (description; West Africa) .-Breder, 1938 :46-48 (description; key; illustration; Western At- lantic) .-Imai, 1958: 45 (description; illustration; North Pacific) .-Parin, 1960: 33-34 (description; key; North Pacific). Cypselurus pinnatibarbatus. Fowler, 1936: 428- 429 (description; illustration; West Africa; misidentification). Cypselurus exiliens. Lozano Rey, 1947: 620-622 (description; illustration; key; Eastern Atlantic). Cheilopogon exsiliens. Parin, 1961: 168 (listed; illustrated p. 124, 169, 170). ? Exocoetus nigricans Bennett, 1840: 287 (original description; 5° N in both Atlantic and Pacific Oceans) .-Bleeker, 1866: 73 (description; Eastern Atlantic).-Martens, 1865: 28,34 (Atlantic Ocean, 8°N) Martens, 1867: 40 I.-Vaillant, 1919: 131 (listed; Western Mediterranean Sea). ? Exocoetus spilopus Valenciennes in Cuvier & Valenciennes, 1846: 118- 120 (original description; West Indies, Atlantic, Indian Oceans). Material examined.-Gulf Stream: UMML 14036 (2,88.6-110.5 mm).- BU (1,42.5 mm) 36°07'N, 73°25'W, 22 April 1960.-BLBG (1, 18.8 mm) 32°40'N, 78°16.5'W, 24 October 1962.-BLBG (1, 183.5 mm) 29°50'N, 80011'W, 16 June 1958.-BLBG (1, 18.2 mm) 33°49'N, 75°59'W, 10 May 1953.-BLBG (1, ]5.3 mm) 33°14'N, 76°25'W, 8 May 1953.-BU (1,93.3 mm) 39°07'N, 65°58'W, 21 September 1957. -USNM (1, 115.8 mm) Massachusetts, Cape Cod.-UMML 14038 (1, 33.5 mm).-UMML 14037 (1, 18.0 mm).-UMML 14035 (1, 105.3 mm). Bahamas: BLBG (1,42.7 mm) 26°30'N, 76°40'W, 12 June 1954. Caribbean Sea: USNM 30975 (1, 179.9 mm). Gulf of Mexico: UMML 4469 (1,52.7 mm).-UF 5210 (1, 126.4 mm). -BU (1,71.0 mm) 19°05'N, 94°10'W, 17 May 1954.-BU (1,99.6 mm) 26°40'N, 92°00'W, 8 May 1954.-BU (1, 45.2 mm) 19°]3'N, 95°34'W, 17 May 1954.-BU (1, 35.7 mm) 24°00'N, 96°50'W, 25 1965] Staiger: Atlantic Flyingfishes 693

May 1954.-BU (1, 182.9 mm) 22°IO'N, 90020'W, 23 May 1954.- USNM 159038 (1, 120.4mm). Atlantic Ocean: UMML 14039 (1, 60.5 mm). Bermuda: UMML 14034 (1,22.1 mm).

Distinctive characters.- The most distinctive character of the juveniles is the morphology of the paired barbels, which are unique among the Atlantic Cypselurus. Each barbel is short and flap-like, rather than cylindrical, which is the case with C. cyanopterus, C. furcatus, and C. heterurus. Throughout the development the upper caudal lobe is only very lightly pigmented, which separates it from all other Atlantic Cypselurus. Adults are distinguished by a combination of unpigmented upper caudal lobe, high dorsal fin with black spots (common also to C. cyanopterus and C. pinnatibarbatus), barred pectoral fin (separating it from C. cyanopterus), and intermediate number of predorsal scales (separating it from C. pinnatibarbatus, as shown in Table 1).

Description.-Cypselurus exsiliens is a medium sized species. The largest specimen I examined measured 183.5 mm standard length. Bruun (1935: 47) records the largest specimen as 236 mm standard length and states that the normal size of the adults is from 200 to 220 mm standard length. The general body form of the juveniles and adults is illustrated in Figure 7. Dorsal fin one of the highest (12.4-34.4 per cent of standard length) of any Atlantic species of Cypselurus. Dorsal edge of fin straight, especially in the small specimens, in which the dorsal edge is almost vertical to the body. Fin always darkly pigmented, the pattern varying with ontogeny. No palatine teeth. All jaw teeth examined were unicuspid. Bruun (1935: 46) felt that the absence of palatine teeth was perhaps an adult character, but I found no palatine teeth in juveniles. Meristic data (range of study material 15-182 mm standard length, extremes in parentheses from the literature): dorsal-fin rays 13-15 (16); anal-fin rays 8-10 (11); dorsal minus anal-fin rays 3-6; pectoral-fin rays 13-16; predorsal scales 21-30; postdorsal lateral scales 17-24; total gill rakers (first arch) 16-26 (28); scales above lateral line 5-8; vertebrae 43-44 (45). Morphometric data (expressed as per cent of standard length, range of study material l5-182 mm standard length, extremes in parentheses from the literature): preanal length 73.1-81.4 per cent; predorsal length 64.4- 70.9 per cent; prepelvic length 53.6-59.2 (59.4) per cent; prepectoral length 20.0-26.7 per cent; head length 20.5-27.3 per cent; snout length 3.7-6.4 (6.6) per cent; fleshy orbit diameter 4.0-9.9 per cent; interorbital width 6.4-10.6 per cent; pectoral fin length 52.8-75.0 (75.7) per cent; pelvic fin length (26.7) 33.4-49.7 per cent; dorsal fin height (12.4) 13.4- 34.4 per cent; anal fin height 6.2-17.9 per cent; greatest depth (14.9) 694 Bulletin of Marine Science [/5(3)

FIGURE 7. Cypselurus exsiliens. A, Juvenile 60.5 mm SL-B, Adult 183.5 mm SL.

16.6-20.2 per cent; body width 13.0-17.2 per cent; depth behind eyes 13.4-16.3 per cent. Coloration of the adult dark dorsally and silver ventrally. Small juveniles with a light background and six dark vertical bars. By 50 mm standard length three of the dark bars have faded appreciably, leaving the most noticeable bars at the pelvic fin roots, the anal fin base, and the caudal peduncle. Pectoral fins dark with a pale crossband; the posterior edge and first ray unpigmented or lightly pigmented (degree of pigmentation changing with ontogeny). Pelvic fin pigmentation varies with ontogeny-in adults the fin is virtually clear (see Fig. 9 A-D). Ontogeny.-The barbels of C. exsiliens are flat, densely pigmented along their edges, with the center pale or clear( see Fig. 8 F). The anterior edge is straight, the posterior serrate with a distal, flap-like appendage. In the smallest specimen examined (15.3 mm standard length) the barbel was 13.1 per cent of standard length. It increases proportionally to about 40 1965] Staiger: Atlantic Flyingfishes 695 per cent of standard length at about 40 mm standard length, decreases to about 12 per cent of standard length in specimens from 70 to 100 mm standard length and is lost between 105 and 116 mm standard length. This development and regression may be seen in Figure 3. The maximum proportional barbel length observed was 46.3 per cent of standard length in a 42.5 mm standard length specimen. Of all species examined C. exsiliens possesses the barbel for the shortest time. The dorsal fin is one of the highest of the species examined: over 20 per cent of standard length in small juveniles, reaching a peak of 34 per cent of standard length at 45.2 mm standard length (see Fig. 5). In juveniles the dorsal edge of the fin is nearly vertical; as the fin decreases in relative height the shape changes so that by 100 mm standard length the dorsal and posterior edges form a rounded profile. The entire fin is dark until about 100 mm standard length, after which the anterior half of the fin clears, leaving a dark spot on the posterior of the fin. In small juveniles the pelvic fin has a clear background with pigment concentrated along the fin rays and in a small band along the fin root and

I F FIGURE8. Cypselurus exsiliens. A-E, Pectoral fin, 18.8 mm SL, base line 2 mm; 42.5 mm SL, base line 5 mm; 60.0 mm SL, base line 10 mm; 105.3 mm SL, base line 10 mm; 183.5 mm SL, base line 25 mm.-F, Barbel, 60.0 mm SL, base line 5 mm. 696 Bulletin of Marine Science [15(3)

A B

FIGURE 9. Cypselurlls exsiliens. A-D, Pelvic fin, 33.5 SL, base line 5 mm; 60.0 mm SL, base line 5 mm; 110.5 mm SL, base line 10 mm; 183.5 mm SL, base line 10 mm. extending down the mesial edge of the fin (Fig. 9 A). As the size increases the postero-mesial corner of the fin darkens until it appears that the posterior two-thirds of the fin is dark. Figure 9 Band C illustrates this pattern phase. Above 140 mm standard length the fin is clear again with pigment concentrated on the fin rays, especially dense on the anterior half of the fin, as in Figure 9 D. The upper caudal lobe is clear to over 100 mm standard length; becoming lightly pigmented by 180 mm standard length, but it is never as dark as the lower lobe. In small juveniles the lower lobe is clear with a number of dark spots; above 100 mm standard length the entire lobe is densely pigmented. In specimens smaller than 50 mm standard length the pectoral fin has a lightly pigmented background with two dark crossbands (Fig. 8 A and B). The lightly pigmented area between the crossbands begins to clear at about 50 mm standard length. This clearing continues until by 100 mm standard length there is a clear band all the way across the fin (Fig. 8 D). During the period that the clear band forms, the two dark bands expand until the anterior and posterior areas of the fin are evenly, darkly pigmented 1965] Staiger: Atlantic Flyingfishes 697 (Fig. 8 D). With further growth the clear band decreases proportionally until by 120 mm standard length it no longer extends completely across the fin (Fig. 8 E). This final pattern is the adult configuration. In most of the specimens the first ray and last few rays are either without pigment or only lightly pigmented. Range.-Cypselurus exsiliens is found in the tropical waters of the western and central Atlantic Ocean. It is found throughout the Caribbean Sea and the Gulf of Mexico and in the Gulf Stream as far north as Cape Cod. C. exsiliens has been taken at Bermuda and in the central Atlantic at 30oN. In the central Atlantic it occurs as far east as 29°W and as far south as Rio de Janeiro, Brazil in the South Atlantic. C. exsiliens is one of the two oceanic species among the Atlantic Cypselurus. It often occurs within a few miles of shore, but it does not enter inlets and bays, as do C. heterurus and C. furcatus. Reproductive biology.-Bruun (1935: 47) states that sex could be determined from 190 mm standard length on and that the only ripe specimen he examined was a female of 218 mm standard length.

Cypselurus furcatus (Mitchill, 1815) Spotfin Flyingfish

Figures 10, 11, 12, 13, 14, 15, and 19. Table 1

Exocoetus furcatus Mitchill, 1815: 449 (original description; illustration, New York) .-Valenciennes, in Cuvier & Valenciennes, 1846: 135-136 (description; Gulf of Mexico).-Giinther, 1866: 286 (description; Atlantic and Indian Oceans).-Jordan & Meek, 1886: 61-63 (description; key).- Jordan & Evermann, 1896b: 737 (description; key; Rhode Island).- Bean, 1903: 334-335 (description; New York). Exocetus furcatus. DeKay, 1842: 231 (description; New York). Cypselurus furcatus. Jordan & Gilbert, 1882: 380-381 (description; Atlantic Ocean).-Breder, ]927: 20 (listed; West Indies).-Nichols & Breder, 1927: 61 (description; illustration; New York).-Fowler, 1928: 6, 83 (listed; description; Atlantic, Indo-Pacific Oceans) .-Nichols & Breder, 1928: 431, 447, 448 (listed; illustration; Atlantic Ocean) .-Breder & Nichols, 1930: 3-4 (description; illustration; West Indies).-Jordan, Evermann, & Clark, 1930: 201 (listed) .-Nichols & Breder, ] 930: 5, 7 (key; synonymy).-Parr, 1930: 25, 26 (description; illustration; Ba- hamas).- Breder & Nichols, 1934: 40, 44 (discussion; illustration).- Fowler, 1934c: 412 (description; St. Lucia Lake, Zululand).-Fowler, 1936: 427-428, 1241-1242 (description; listed; West Africa) .-Breder, 1938: 69-73 (description; illustration; key; synonymy; Western Atlantic Ocean). Longley & Hildebrand, 1941: 31-32 (discussion of coloration).- Parin, 1960: 35-37 (description, key; Northwestern Pacific).-Smith, 1961: 125-126 (description; illustration; South Africa) .-Smith, 1963: 9 (listed; illustration; Seychelles Islands). 698 Bulletin of Marine Science [15(3) Cypsilurus furcatus. Jordan & Evermann, 1898: 2836 (listed) .-Bruun, 1935: 58-60 (description; illustration; key; synonymy; Atlantic Ocean).- Smith, 1950: 125-126 (description; illustration; South Africa). Cypselurus heterurus. Nichols & Breder, 1930: 7 (misidentification in part; fide Breder, 1938: 69).-Breder & Nicho]s, 1934: 44 (misidentification in part; fide Breder, 1938: 69). Cheilopogon furcatus. Parin, 1961: 132-133, 174 (listed; illustration). ? Exocoetus nuttallii LeSueur, 1821: 10 (original description; illustration; Gulf of Mexico).

Material exomined.-Gulf Stream: UMML 13786 (1, 40.6 mm).- UMML 13791 (1,45.7 mm).-UMML 13792 (1, 33.3 mm).-UMML 13796 (1,47.0 mm).-UMML 13782 (1, 130.5 mm).-UMML 13784 (1, 149.7 mm).-BU (2,34.7-62.2 mm) 36°42'N, 70000'W, 3 October 1957.-BU (1, 34.9 mm) 41°30'N, 66°15'W, 7 June 1957.-BU (I, 86.6 mm) 35°50'N, 72°35'W, 19 Ju]y 1958.-UMML 13794 (2, 45.9- 57.1 mm).-UMML 13789 (5, 60.1-83.9 mm).-BU (1, 82.6 mm) 39°07'N, 65°58'W, 14 September 1957.-UMML 13795 (1,69.5 mm). -BU (1, 57.4 mm) 38°00'N, 65°25'W, 1 May 1960.-BLBG (1, 45.8 mm) 33°15'N, 76°23'W, ]5 May 1959.-BLBG (3, 30.7-45.8 mm) 31°28'N, 78°42'W, 24 October 1953.-BU (1, 27.0 mm) 32°50'N, 77°27'W, 21 April 1957.-MCZ (1, 82.7 mm) Florida, W. Palm Beach, 14 May 1920.-USNM 116707 (2, 106.5-112.0 mm) .-USNM 37792 (1, ca. 124 mm).-USNM 21628 (1, 201.4 mm).-BU (1, 87.3 mm) 36°40'N, 71°05'W, 25 June 1957.-USNM 118884 (2, 34.4-43.5 mm). -UMML 13790 (1,35.9 mm). Bahamas: UMML 13788 (1, 43.1 mm) .-BLBG (1, 104.6 mm) 26°30' N, 76°40'W, 18 April 1954.-BLBG (1, 59.0 mm) 26°30'N, 76°40'W, 23 January 1954.-BLBG (1,34.6 mm) 26°30'N, 76°40'W, 24 January 1954.-BLBG (4,8.8-14.8 mm) Tongue of the Ocean, 28 January 1954. -BLBG (1, 18.8 mm) 25°30'N, 77°10'W, 27 January 1954.-BLBG (1, 61.0 mm) 26°30'N, 76°40'W, 13 June 1954. Caribbean Sea: MCZ 34747 (1, 34.4 mm).-BU (5, 39.2-56.0 mm) 19°12'N, 62°05'W, 1 May 1957.-BU (2, 47.7-69.4 mm) 19°39'N, 58°03'W, 30 April 1959. Gulf of Mexico: UMML 4468 (1, 87.3 mm).-UMML 7723 (1, 24.3 mm).-BU (2, 72.5-148.0 mm) 24°54'N, 96°05'W, 27 May 1954.- UMML 13783 (3,24.1-117.3 mm).-UMML 13785 (1, 138.1 mm).- BU (1, 136.7 mm) ca. 25°N, 91oW, 29 November 1956.-UMML 13781 (2, 27.7-54.8 mm).-BU (2, 47.0-116.6 mm) 26°40'N, 92°00'W, 8 May 1954.-USNM 185914 (1, 170.4 mm).-USNM 185803 (1,211.6 mm).-USNM 185786 (1, 19.0 mm).-USNM 157817 (1,186.9 mm). Bermuda: BU (2, 25.7-30.0 mm) 32°00'N, 63°20'W, no date.-UMML 13793 (1, 33.4 mm) .-BU (2, 28.2-32.9 mm) 32°57'N, 64°19'W, 13 October 1955.-BU (3, 43.2-65.7 mm) 33°N, 64°W, 12 October 1955. 1965] Staiger: Atlantic Flyingfishes 699 60

o o 0 em 00 0 0c? 0

Ql 0<00 0 ~o 00 ~ 20 ooOJO8 0 0 o .0 tP.. 0.••• .,.. • ~~ ~ 0 o OJ ().I~b'·. .., •••~. ~ • o .• • -...... ••••••• 0 o 00 , --- ••.• o-O- __ .~._-O.----.--O---r-O o 20 40 60 80 100 120 140 160 180 200 S.L.inmm FIGURE 10. Development and regression of barbel in relation to SL. Solid circles, Cypselllrlls furcatus; open circles, C. heterurus.

-BU (2,67.5-79.5 mm) 32°57'N, 64°19'W, 13 October 1955.-USNM (1, 32.6 mm) 30027'N, 66°05'W, 19 February 1914. Distinctive characterS.-Juveniles (smaller than 150 mm standard length) are distinguished from other species of Cypselurus by a combination of characters. In common with C. heterurus and C. comatus they have a low dorsal fin (less than 20 per cent of standard length), but, unlike C. comatus, the fin is unevenly pigmented. Compared with C. heterurus the mandibular barbels are longer (usually longer than 16 per cent of standard length; see Fig. 10); the first pectoral ray is longer (usually longer than 38 per cent of standard length; see Figure 11); and the distance between the first and second pectoral rays is usually less than 1.7 times the distance

•• 50 •• \\e ,.. . .\ ... • 0 • ••• • • a: ••• • 0 40 • • .0 • • ~ 0 •• u 00 o 0 -Ql 0 a.. 35 <0 ~o 00 0 -:;; ~ «,oO~oCOOOo ii: o ~o 0 30 o 0 0 ~ CO

25 o 50 100 150 200 S.L. in mm FIGURE 11. First pectoral-fin ray length in relation to SL. Solid circles, Cypselurus furcatus; open circles, C. heterurus. 700 Bulletin ot Marine Science [15(3)

IJ) » 500 o a:: 0

400

o 0 300 ~ 0 o 0 0 0& a a 0 200 Cb~ ogDo~O 0 _I,,- - 0a 100 _ -ti_ ~ce :- -~o a Cl a o 50 100 150 200 S. L in mm

FIGURE 12. Distance between first and second pectoral rays as per cent of distance between second and third pectoral rays in relation to SL. Solid circles, Cypselurus furcatus; open circles, C. heterurus. between the second and third (usually greater than 1.7 in C. heterurus; see Fig. 12). The distinctive pectoral fin color pattern serves to distinguish juvenile and adult C. furcatus from C. heterurus. Adults (greater than 150 mm standard length) are distinguishable by their low, unpigmented dorsal fin (less than 12 per cent of standard length); long first pectoral ray (greater than 38 per cent of standard length); and dark pectoral fin with distinct light crossband, widest distally. Description.-Cypselurus furcatus is a large member of the genus. The largest specimen I examined measured only 211.6 mm standard length, but Bruun stated that his largest specimen was 293 mm standard length, and that the common adult range was 230-260 mm standard length (Bruun, 1935: 60). Figure 13 illustrates the general body form of the adults and juveniles. No palatine teeth. Pre-maxillary teeth very small, without accessory cusps. Meristic data (range of study material 8-211 mm standard length, extremes in parentheses from the literature): dorsal-fin rays 11-14 (15); anal-fin rays 8-12; dorsal minus anal-fin rays 1-5; pectoral-fin rays 14-17; predorsal scales 22-35; postdorsal lateral scales 13-22; total gill rakers (first arch) 16-24; scales above lateral line 6-9; vertebrae 44-46. Morphometric data (expressed as per cent of standard length, range of study material 8-211 mm standard length, extremes in parentheses from the literature): preanal length 72.3-78.5 per cent; predorsal length 64.5- 74.3 per cent; prepelvic length 51.9-58.7 per cent; prepectoral length 1965] Staiger: A t[antic Flyingfishes 701 20.0-27.2 per cent; head length 20.5-26.4 per cent; snout length 2.3-8.5 per cent; fleshy orbit diameter (7.'0) 7.3-12.2 per cent; interorbital width 6.9-9.1 per cent; pectoral-fin length 31.9-74.8 per cent; pelvic-fin length (29.4) 32.8-58.2 per cent; dorsal-fin height (7.5) 9.5-19.5 per cent; anal-fin height 6.8-15.9 per cent; dorsal-fin base 19.1-23.7 per cent; anal-fin base 13.2-16.7 per cent; greatest depth 15.5-20.7 per cent; caudal peduncle depth 5.2-8.1 per cent; body width (12.5) 12.6-15.8 per cent; depth behind eyes 12.8-16.6 per cent. Adults pale silver ventrally and dark dorsally. Dorsal fin sparsely pigmented in adults, frequently appearing completely clear. Pelvic fins of the adults lightly pigmented with a vague, pale crossband. Fins nearly clear in specimens longer than 200 mm standard length. Adult's pectoral fins dark with a clear crossband extending from the mesial edge obliquely towards curved anterior edge. Edges of crossband curving laterally, so that it is narrow proximally, widest distally. Adult's caudal fin darkly pigmented over its entire surface. The juvenile coloration is discussed in the next section. Ontogeny.- The changes in dorsal-fin height parallel those of C. heterurus, but the fin is somewhat higher.

FIGURE 13. Cypselurus furcatus. A, Juvenile 72.5 mm SL.-B, Adult 149.7 mm SL. 702 Bulletin of Marine Science [15(3) The mandibular barbels each consist of a fleshy shaft arising from the front of the mandible and gradually tapering to a point. A dark membrane extends posteriorly the length of the shaft and decreases in width from

FIGURE 14. Cypselurus furcatus. A-G, Pectoral fin, 16.4 mm SL, base line 2 mm; 31.1 mm SL, base line 5 mm; 44.1 mm SL, base line 10 mm; 57.1 mm SL, base line 10 mm; 72.1 mm SL, base line 10 mm; 82.6 mm SL, base line 15 mm; 134.7 mm SL, base line 20 mm. 1965] Staiger: Atlantic Flyingfishes 703 root to tip (Fig. ] 5 D). The barbels are relatively longest (about 33 per cent of standard length) at from 35 to 65 mm standard length, from which they decrease in relative length slowly until they disappear. They are retained longer than in C. heterurus (Fig. 10): one specimen of 138 mm standard length had barbels 16 per cent of standard length; the smallest specimen without barbels was 100 mm standard length, while in C. heterurus a specimen of 79 mm standard length had already lost them. Juveniles have a pale body with six vertical pigmented bars just as in C. heterurus. C. furcatus, however, attains the adult pattern earlier (l15 mm standard length) than does C. heterurus (125 mm standard length). The pectoral fins of very small specimens (under 25 mm standard length) show the pattern diagrammed in Figure 14 A. This pattern consists of two pigmented areas on a clear background. The anterior pigment spot or blotch covers the first ray, the fin root, and extends out onto the fin for a quarter of its length. The second pigmented area is a band that extends from the mesial edge of the fin across to the tip of the fin. The anterior edge of this band is straight throughout the development of the adult's pattern, and serves to differentiate the juveniles of C. furcatus from the juveniles of C. heterurus. With growth the pigment blotch on the

~ ' ....""'.:.;.., "c. ~ :;::'. c;

<:, D ~ ,..:.: I >: ...... • :; C ~ .,,;?' t------f

FIGURE 15. Cypselurus jurcatus. A-C, Pelvic fin, 31.1 mm SL, base line 3 mm; 44.1 mm SL, base line 5 mm; 134.7 mm SL, base line 10 mm.-D, Barbel, 75.0 mm SL, base line 5 mm. 704 Bulletin of Marine Science [15(3) anterior of the fin breaks up into a number of smaller blotches which eventually reunite to form the anterior portion of the adult's pattern. This development can be followed easily in Figure 14. The band on the posterior portion of the fin maintains its integrity throughout development. Its development consists essentially of a spreading both anteriorly and posteriorly to cover more than half of the fin by the time the adult's pattern is attained. After about 60 mm standard length the characteristic V-shaped clear band starts to develop near the mesial edge of the fin. This V-shaped clear area develops into the clear band that is characteristic of the pectoral fin of the adult of C. furcatus. In the adult the sides of the clear band diverge so that the band is narrowest at the mesial edge of the fin and widest at the anterior edge. This pattern is seen in Figure 14 G. This is in contrast to the pattern in adult C. heterurus (Fig. 17 H), where the clear band is widest mesially. During the course of development processes from the pigment on the posterior portion of the fin often extend to the posterior edge of the fin, as in Figure 14 C-D. In the adult, however, the posterior margin of the fin is clear over a wide area. This wide, clear posterior margin has been used by previous workers to differentiate C. furcatus from C. heterurus, but this can be misleading, because the posterior edge of the fin can become so tattered on occasion as to nearly obliterate the clear posterior margin. The pelvic fins at 25 mm standard length are clear with pigment at the fin base and three large pigment spots on the rest of the fin, as in Figure 15 A. These three spots enlarge and merge as the fish grows. The spot at the fin base also enlarges with growth, but it is separated from the three other spots by a clear arc. By 130 mm standard length the pattern is similar to Figure 15 C, with the clear arc interrupted by pigment along fin rays two through four. With further growth the pigment gradually fades away posteriorly until in the large adults (200 mm standard length) the pelvic fins are clear except for some light pigment along the rays proximal to the fin base. Thus, the pelvic-fin pattern of the adult C. furcatus is similar to that of C. heterurus illustrated in Figure 18 G. The dorsal fin at 25 mm standard length has a slight amount of pigment on the anterior few membranes. By 35 mm standard length several spots appear along the dorsal edge of the fin. These spots at first increase in size, but then gradually decrease so that by 150 mm standard length most of the fin is clear. The anal fin develops a small spot of pigment on its posterior edge, but this disappears by 150 mm standard length to leave the fin clear. The caudal-fin pattern in juveniles consists of a spot on the tip of the upper lobe, another on the distal half of the lower lobe, and a third about a quarter of the way out on the lower lobe. These spots enlarge with growth so that by 115 rom standard length the entire fin is dark. Range.--Cypselurus furcatus is found in the tropical waters of the central 1965] Staiger: Atlantic Flyingfishes 705 and western Atlantic Ocean. It occurs throughout the Gulf of Mexico and the Caribbean Sea and in the Gulf Stream to 42°N, and is also common about Bermuda. In the central Atlantic it is found to 30oN. The eastern edge of its range appears to be west of the Cape Verde Islands (at 15°N, 29°W). It has not been taken along the west coast of Europe nor has it been recorded from the South Atlantic. C. furcatus is one of the two oceanic species of Cypselurus in the Atlantic. Unlike C. exsiliens, however, C. furcatus occurs close inshore, for instance in Biscayne Bay, Florida, as well as far out to sea. Reproductive biology.-Sexually mature females have never been reported. Breder (1932: 22) felt that the breeding season might be continuous, because he found small juveniles in the Caribbean Sea in February and from May to July. Bruun (1935: 60) reports a ripe male from November. I have small juveniles from the Gulf Stream and the Gulf of Mexico for the late fall through the summer. All this supports a contention of a long, though not necessarily continuous breeding season. Cypselurus heterurus (Rafinesque, 1810) Atlantic Flyingfish Figures 5, 10, 11, 12, 16, 17, 18, and 19. Table 1 Exocetus heterurus Rafinesque, 1810: 58-59 (original description; Gulf of Palermo). Exocoetus procne Filippi & Verany, 1859: 61 (original description; Tunis; J fide Breder, 1938: 56). Exocoetus maculipinnis Vinciguerra, 1883: 577 (original description; Bay of Naples Sicily, Canary Islands). Exocoetus heterurus. Jordan & Meek, 1886: 59-61 (description; key).- Jordan & Evermann, 1896b: 735-736 (description; key; Western At· lantic).-Bean, 1903: 331-332 (description; New York). Exocoetus furcatus. Jordan & Meek, 1886: 61-63 (misidentification in part). Exocoetus robustus. Jordan & Meek, 1886: 61 (description: key). Exocoetus lutkeni Jordan & Evermann, 1896b: 736 (original description; key; Cuba). Cypsilurus lutkeni. Jordan & Evermann, 1898: 2836 (Iisted).-Bruun, 1935: 56-58 (description; key; illustration; synonymy; North Atlantic). Cypsilurus heterurus. Jordan & Evermann, 1898: 2836 (listed) .-Bruun, 1935: 61-64 (description; illustration; key; synonymy; North Atlantic).- Bruun, 1938: 295 (description; Oslo Fiord).-Poll, 1953: 190-191 (description; illustration; 25°N, 17°W). Cypselurus lutkeni. Smith, 1907: 167 (description; illustration; North Caro- lina) .-Breder & Nichols, 1930: 6-7 (description; illustration; West Indies) .-Jordan, Evermann, & Clark, 1930: 201 (listed) .-Nichols & Breder, 1930: 5-6 (key; synonymy) .-Breder & Nichols, 1934: 40, 44 (discussion; illustration).-Fowler, 1936: 1234 (description; West Africa).-Breder, 1938: 56-62 (synonymized with C. heterurus). Cypselurus heterurus. Breder, 1927: 20 (listed). Nichols & Breder, 1927: 61 (description; illustration; New York).-Breder, 1928: 23 (listed).- Jordan, Evermann, & Clark, 1930: 201 (listed) .-Nichols & Breder, 706 Bulletin of Marine Science [15(3) 1930: 5 (key) .-Parr, 1930: 25-27 (description; illustration; Bahamas). -Breder & Nichols, 1934: 40, 44 (discussion; illustration).-Fowler, 1936: 1242-1243 (description; illustration; West Africa) .-Breder, 1938: 56-69 (description; illustration; key; synonymy; Western Atlantic).- Lozano Rey, 1947: 625-629 (description; illustration; Eastern Atlantic, Mediterranean) .-Bigelow & Schroeder, 1953: 172 (description; illustration; Gulf of Maine).- Imai, 1958: pis. 14-18 (illustration; Japan).- Springer, 1959: 166-167 (description).-Abe, 1960: 149 (listed; Japan). Cypselurus luetkeni. Springer, 1959: 166-167 (description; Gulf of Mexico). Cheilopogon heterurus. Parin, 1961: 105,112,173,174 (listed; illustration). Cheilopogon lutkeni. Parin, 1961: 174 (listed).

Material examined.-Gulf Stream: UMML 13812 (1, 78.1 mm).-BU (1, 70.8 mm) 40000'N, 68°45'W, 15 September 1957.-BU (1, 125.6 mm) 34°45'N, 73°41'W, 10 October 1957.-BU (1,89.2 mm) 37°08'N, 68°14'W, 18 February 1961.-BU (3, 71.0-131.4 mm) 39°07'N, 65°58'W, 14 September 1957.-BU (3, 77.2-104.2 mm) 38°00'N, 65°25'W, 17 February 1961.-BU (3, 107.8-144.8 mm) 39°07'N, 65°58'W, 21 September 1957.-BU (13, 23.6-127.9 mm) 39°28'N, 69°30'W, 30 October 1956.-UMML 13799 (4, 37.9-50.8 mm).-BU (31, 63.5-140.5 mm) 38°17'N, 70000'W, 25 June 1957.-BU (9, 23.4- 98.2 mm) 37°45'N, 71°49'W, 7 October 1957.-BU (1, 90.6 mm) 38°13'N, 66°48'W, 23 October 1956.-BU (1, 86.5 mm) 32°05'N, 71°45'W, 26 March 1957.-USNM 83748 (1,77.8 mm).-USNM 94694 (2, 98.1-105.1 mm).-USNM 94705 (3, 122.4-148.6 mm).-USNM 94716 (2, 35.7-67.5 mm).-USNM 118883 (1, 74.6 mm).-USNM 178269 (1,190.0 mm).-USNM 120570 (1,196.5 mm).-MCZ 37990 (1,214.5 mm).-MCZ 37139 (1, 179.1 mm).-USNM 161337 (1,47.6 mm).-USNM 154013 (4,194.5-232.0 mm).-USNM 94718 (2, 195.8- 232.1 mm).-BLBG (1, 142.2 mm) 30021'N, 79°26'W, 27 June 1954.- BLBG (1,98.6 mm) 30017'N, 80011'W, 15 October 1953.-BU (1,97.5 mm) 39°17'N, 71°49'W, 6 October 1957.-BLBG (1, 84.6 mm) 27°00' N, 79°18'W, 8 November 1953.-BLBG (1,90.7 mm) 31°41'N, 80°35' W, 21 October 1953.-BLBG (2, 115.2-118.4 mm) 29°40'N, 80023'W, 14 October 1953.-UMML 138'02 (2, 104.9-]31.2 mm).-UMML 13804 (4,36.5-50.7 mm).-UMML 13787 (1,101.5 mm).-UMML 4439 (4, 40.7-107.2 mm). Bahamas: BLBG (2, 54.3-69.4 mm) 26°30'N, 76°40'W, 18 June] 954. -BLBG (1, 63.5 mm) 24°32'N, 77°18'W, 7 October 1953.-BLBG (1, 157.6 mm) 24°28'N, 77°28'W, 8 October 1958. Caribbean Sea: USNM 154016 (3, 177.3-209.0 mm). Gulf of Mexico: USNM 158184 (1, 122.3 mm).-USNM 185803 (1, 169.1 mm).-USNM 185808 (3, 114.5-189.4 mm).-USNM 185915 (3, 114.0-164.4 mm).-UMML 13811 (3, 23.6-37.0 mm).-UMML 13800 (3,18.5-57.3 mm).-BU (3,30.9-125.0 mm) 24°54'N, 96°05'W, 27 May 1954.-UMML 13798 (5,30.9-115.4 mm).-BU (1, 107.5 mm) 1965J Staiger: A tlantie Flyingfishes 707 26°05'N, 95°25'W, 27 May 1954.-BU (1, 97.4 mm) 26°40'N, 92°00'W, 8 May 1954.-BU (3, 87.8-199.4 mm) Gulf of Mexico, 250 miles from Clearwater, Florida, 30 June 1948.-BU (1, 194.6 mm) 20012'N, 91°59'W, 26 November 1956.-BU (1, 191.9 mm) ca. 25°N, 89°W, 20 November 1956.-BU (1, 170.4 mm) 20034'N, 95°37'W, 20 May 1954.-BU (1, 145.6 mm) 24°22'N, 92°00'W, 10 May 1954. -BU (1, 207.3 mm) 22°lO'N, 96°20'W, 23 May 1954.-BU (5, 98.5- 147.8 mm) 25°30'N, 92°00'W, 10 May 1954.-BU (2, 15.6-23.2 mm) 26°40'N, 92°00'W, 8 May 1954.-BU (1, 190.0 mm) 20012'N, 91°59'W, 12 July 1957.-BU (3, 57.1-60.3 mm) 20olO'N, 92°25'W, 24 November 1956.-BU (2, 78.7-84.7 mm) 24°00'N, 96°40'W, 25 May 1954.-USNM 158422 (2, 186.3-206.9 mm) .-USNM 157987 (1,97.5 mm).-BU (2, 89.7-103.9 mm) 20034'N, 95°37'W, 20 May 1954.-USNM 157817 (5, 153.5-216.0 mm).-BU (3,50.0-70.4 mm) 24°22'N, 92°00'W, 10 May 1954. Brazil: USNM 10409 (1, 336.8 mm). Bermuda: USNM 178434 (1, 196.5 mm).-USNM 178152 (1, 166.7 mm) .-USNM (2, 68.5-81.0 mm) Bermuda, no date.-BU (2, 72.2- 77.4 mm) 33°N, 64°W, 12 October 1955. Atlantic Ocean: USNM 57837 (1, 317.3 mm). Distinctive characters.-Juveniles of C. heterurus are distinguished from C. exsiliens, C. cyanopterus, and C. pinnatibarbatus by their low (less than 20 per cent of standard length), unevenly pigmented dorsal fin. The presence of short, paired barbels (less than 16 per cent of standard length) separates them from C. comatus and C. furcatus. The short first .pectoral ray (less than 38 per cent of standard length) and its wide separation from the second ray (greater than 1.7 times the distance between the second and third rays) serve to further differentiate C. heterurus from C. furcatus. The best separation of juveniles is based on the pectoral-fin pattern. Juveniles of C. furcatus have a pattern consisting of two pigment spots along the first ray, one at the fin base, and one along the mesial edge separated by a clear V-shaped cleft from the large posterior pigment area. The rear edge of the clear V-shaped cleft continues straight out to the anterior edge of the fin. Juveniles of C. heterurus have two pigment spots along the first ray and one spot at the fin base, but there is no mesial spot in front of the large posterior pigment area and the anterior edge of this area is not straight in juveniles (shorter than 90 mm standard length). In large specimens the low (less than 12 per cent of standard length), unpigmented dorsal fin separates them from C. cyanopterus, C. exsiliens, and C. pinnatibarbatus. The short first pectoral ray (less than 38 per cent of standard length) and the triangular crossband that is widest mesially serve to separate them from C. comatus and C. furcatus. Description.-Cypselurus heterurus is one of the larger Atlantic flying- 708 Bulletin of Marine Science [15(3) fishes. The largest specimen examined for this paper measured 336.8 mm standard length. Bruun (1935: 64) states that the majority of specimens he examined lay in the range of 260 to 300 mm standard length. Figure 16 illustrates the general juvenile and adult configuration. Dorsal fin of adults low, generally less than 12 per cent of standard length. Dorsal-fin pigmentation sparse, frequently lacking in specimens longer than 150 mm standard length.

FIGURE 16. Cypselurus heterurus. A, Juvenile 69.4 mm SL.- B, Adult 163.8 mm SL. Palatine teeth lacking. Premaxillary teeth very small, even in the largest specimens. All teeth unicuspid. Meristic data (range of study material 15-336 mm standard length, extremes in parentheses from the literature): dorsal-fin rays 10-15; anal-fin rays 8-12; dorsal minus anal-fin rays 3-6; pectoral-fin rays 13-17; predorsal scales 22-35 (38); postdorsallateral scales 16-24; total gill rakers (first arch) 17-26; scales above lateral line 6-9; vertebrae 42-46 (49). Morphometric data (expressed as per cent of standard length, range of study material 15-336 mm standard length, extremes in parentheses from the literature): preanal length (74.7) 75.0-80.1 (81.6) per cent; predorsal length 65.3-71.7 per cent; prepelvic length 51.8-56.9 (58.2) per cent; prepectoral length 20.4-25.4 per cent; head length 20.0-27.1 per cent; snout length 3.2-8.7 per cent; fleshy orbit diameter (6.0) 6.3-11.5 per cent; interorbital width (6.9) 7.1-8.8 per cent; pectoral-fin length 39.3- 77.8 per cent; dorsal-fin height 7.8-23.8 per cent; anal-fin height 6.2-13.4 1965] Staiger: Atlantic Flyingfishes 709 per cent; dorsal-fin base 18.0-23.0 per cent; anal-fin base 10.7-13.5 per cent; greatest depth (14.5) 16.0-20.3 per cent; caudal peduncle depth 6.3-7.7 per cent; body width (11.8) 13.6-16.2 per cent. Adults (longer than 150 mm standard length) light silver ventrally and dark dorsally; dorsal pigment brown in preservation. Anal fin unpigmented in adults. Caudal fin darkly pigmented in adults. Pectoral fin of adults dark with pale crossband and narrow pale posterior margin; posterior half of fin darkest. Cross band widest mesially and tapering toward anterior edge of fin, triangular in shape. First ray often lightly pigmented. Pelvic fin in adults lightly pigmented with vague pale crossbands. Fin nearly clear in many specimens over 180 mm standard length.

Ontogeny.-The most striking ontogenic change involves the pectoral-fin color pattern. In specimens of about 20 mm standard length the fin has five pigment spots on a clear background (Fig. 17 A). Three of them are situated along the first ray. The fourth spot is large, and occupies most of the posterior half of the fin. The fifth spot, located along the mesial edge of the fin, is small and transitory. It is incorporated into the large posterior spot by 30 mm standard length (Fig. 17 B). Throughout development the three pigment spots on the anterior half of the fin remain fairly stationary but they do vary considerably in size. By 100 mm standard length the middle pigment spot and the proximal one at the fin base have merged. By 125 mm standard length all three anterior spots have merged to form a solid band of pigment along the length of the first ray. In Figure 17 H this merging is nearly complete, with just a small clear area left on the first ray. As can be seen from Figure 17 A to E the posterior pigment spot enlarges with development, spreads forward, and is variously broken up, but its anterior margin is not straight until the fish is about 90 mm standard length. At no time during the development of the adult's pectoral fin pattern is there any indication of a V-shaped, straight sided clear band between the anterior and posterior pigment spots. Up to 90 mm standard length the rear margin of the clear band is straight, but the forward margin is not consolidated until about 120 mm standard length. As with C. furcatus pigment processes extend from the posterior spot to the posterior edge of the fin during much of the course of development, but in the adult the posterior margin of the fin is clear. From about 1'00 mm standard length on the clear band assumes the triangular shape that is characteristic of the adult's fin pattern. From Figure 17 G and H it can be seen that the mesial edge of the clear band is widest and the sides of the clear band converge toward the anterior edge of the fin. In C. furcatus they are divergent (Fig. 14 G). By 150 mm standard length the adult's pattern is completely formed. It is similar to that shown in Figure 17 H, but the triangular clear band does not reach the anterior edge of the fin. In specimens below 100 mm standard length the distance between the 71'0 Bulletin of Marine Science [15(3)

:~: ••. or'

FIGURE 17. Cypselurus heterurus. A-G, Pectoral fin, 21.5 mm SL, base line 2 mm; 30.3 mm SL, base line 3 mm; 40.0 mm SL, base line 5 mm; 77.9 mm SL, base line 10 mm; 86.3 mm SL, base line 10 mm; 99.9 mm SL, base line 10 mm; 111.4 mm SL, base line 10 mm; 125.2 mm SL, base line 20 mm.

first and second pectoral fin rays is generally well over 1.7 times the distance between the second and third rays (see Fig. 12). In specimens larger than 100 mm standard length the distance between the first and second rays decreases markedly so that above this length a separation from C. furcatus on the basis of this character is impossible. As is true for all 1965] Staiger: Atlantic Flyingfishes 711 Atlantic species of Cypselurus the pectoral fin lengthens proportionally with growth but C. jurcatlls generally has a longer pectoral fin at any given standard length (Fig. 19).

--'''~.~''::~~'"':::::::::-'--'--- ~~:~- .. :~::-~::~

G H

FIGURE 18. Cypselurus heterurus. A-F, Pelvic fin, 21.5 mm SL, base line 3 mm; 30.3 mm SL, base line 3 mm; 40.0 mm SL, base line 3 mm; 58.7 mm SL, base line 5 mm; 92.7 mm SL, base line 5 mm; 125.2 mm SL, base line 10 mm; 220.] mm SL, base line 10 mm. H, Barbel, 71.0 mm SL, base line 5 mm. 712 Bulletin of Marine Science [15(3) The development of the color pattern in the pelvic fins is very similar to that in C. furcatus and it can be followed easily in Figure 18 A through G. The fin has a pigment spot at its base and a number of spots on the posterior half. A clear arc separates the two pigment areas. With development the posterior pigmented area enlarges until the fin is dark except

80 0

75 •• 0 •• • • 0 0 • • 70 \I. ·A· .0 0 .:c.., .. (J 1 • .~ 0 ~ 0 65 . .0. ~ • a.. 45 c9 • 40 0

35 •• • 30 0 50 100 150 200 250 300 350 S. L. in mm FIGURE 19. Pectoral-fin length in relation to SL. Solid circles, Cypselurus fureatus, open circles, C. heterurus.

for the clear are, which is eventually restricted to the mesial edge of the fin (Fig. 18 F). From this pigment maximum, at about 125 mm standard length, the pigment gradually fades away until, in the large adults (over 200 mm standard length), the fin is clear except for pigment on the proximal halves of the fin rays (Fig. 18 G). The dorsal fin of C. heterurus is quite high in small specimens (Fig. 5); 19651 Staiger: Atlantic Flyingfishes 713 up to about 70 mm standard length it averages about 17 per cent of standard length. Above 70 mm standard length the fin decreases proportionally to a mean height of about 11 per cent of standard length (at 150 mm standard length). Few specimens had fins as low as 10 per cent of standard length even in the largest examined. The paired mandibular barbels (morphologically identical to those of C. furcatus) are always small (Fig. 18); no specimen had barbels longer than 15 per cent of standard length. Maximum length is attained at about 25 mm standard length and is maintained to between 70 and 80 mm standard length, after which the barbels decrease in proportion slowly until they disappear, generally by 110 mm standard length. This development and regression of the barbels is compared with that of C. furcatus in Figure 10. The lower lobe of the caudal fin of small specimens is clear with a dark spot at its base and one on its distal half. The upper lobe is frequently devoid of pigment until about 55 mm standard length, at which time two spots appear in positions similar to those on the lower lobe. All these spots increase their area until at about 140 mm standard length the entire fin is darkly pigmented. The coloration of the body in juveniles consists of a light background with six vertical pigmented bands on the sides and ventral surface. The dorsal surface is pale, and bands are absent. This pattern changes gradually, with the back becoming darker and the belly silvery until by about 125 mm standard length the adult's pattern is attained. Range.-Cypselurus heterurus occurs in the tropical and temperate waters of the Atlantic Ocean and Mediterranean Sea. It has been taken in the western Atlantic from 42°N to 23°S and in the eastern Atlantic from 59°N to 4°N. It is very common throughout the Caribbean Sea and the Gulf of Mexico, and especially in the Gulf Stream and in the Straits of Florida. C. heterurus is also common about Bermuda. It is a coastal species, and has not been taken farther than 400 miles from land. Juveniles often come close inshore, and in the Miami area they have been dipnetted from piers in Biscayne Bay. Bruun (1935: 97-1938: 295) reports it twice from Oslo Fjord, farther north in the Atlantic than any other exocoetid. Reproductive biology.-Sex can be determined from 175 mm standard length (Breder, 1938: Table XV). Sexual maturity is reached at about 200 mm standard length. Heldt (in Bruun, 1935: 64) reports that the species breeds off the Atlantic coast of Morocco from June to July; Bruun (1935: 64) states that they breed in the Mediterranean Sea from May to August. Ovarian eggs range in size from 1.6-1.8 mm and are equipped with long adhesive tendrils (Breder, 1938: 67). Discussion.-Cypselurus luetkeni was described by Jordan & Evermann 714 Bulletin of Marine Science [15(3) from one specimen from Cape San Antonio, Cuba. The characters used in their key to differentiate it from C. hete/'Urus were (C. heteru/'Us in parentheses): base of anal 2 times (12/3) in base of dorsal; pectoral fin 12/5 (1 4/9) in length; and pelvic fin 2 8/9 (2 3/4) in length. Bruun (1935: 56) retained the species, but suggested that a reexamina- tion of type specimens might lead to a change in synonymy. He separated it from C. heterurus by the following characters: 27-33 predorsal scales (30-38 in C. heterurus); 7-8, usually 7, transverse scales (vs. 7-9, usually 8); and 46-47 vertebrae (vs. 47-49). He further considered C. /uetkeni to be a tropical form and C. heterurus to be subtropical. Breder (1938: 56) synonymized C. /uetkeni with C. heterurus, pointing out that the only adequate character in Jordan and Evermann's key (dorsal-anal base proportions) was inadequate because of the overlap he found between the two species, and that Bruun's characters also did not afford a separation. Breder supported his synonymy with figures demon- strating this overlap. Springer (1959: 166) provisionally reinstated C. luetkeni. Utilizing Bruun's and Breder's data he found separation in these characters: head length, snout length, diameter of eye, and pectoral-fin length. All exhibited overlap, but he considered the divergence as possibly great enough to be of specific value. The samples he compared in all four cases numbered no more than eleven specimens of each species. Head length exhibited 91 per cent average divergence; the other characters ranged from 79-86.3 per cent divergence. Jordan & Evermann (1896), Bruun (1935), and Springer (1959) present no data for either C. heterurus or C. luetkeni that do not lie within the ranges of the data presented here for the same characters (western Atlantic specimens only). When those characters that were emphasized by Springer are plotted against standard length, the data for C. heterurus do not deviate from the regression pattern of western Atlantic specimens. There is, then, no ground for considering C. /uetkeni as distinct from C. heterurus, even on a subspecific level.

Cypse/urus pinnatibarbatus (Bennett, 1831) Table 1. Exocoetus pinnatibarbatus Bennett, 1831: 146-147 (original description; Atlantic coast of North Africa) .-Gunther, 1866: 284 (description; Atlantic coast of North Africa) .-Metzelaar, 1919: 220 (listed; Dutch West Indies) .-Roule & Angel, 1921: 4 (listed; Azores). Cypselurus pulchellus Lowe, 1841: 38 (original description; type-species of Cheilopogon Lowe, 1841; Madeira).-Gi.inther, 1866: 285 (description: eouates it with Cypselurus solandri (Valenciennes, 1846» .-Bruun, 1935: 52, 77 (listed; referred to Cypselurus lineatus (Valenciennes, 1846». Exocoetus lineatus Valenciennes, in Cuvier & Valenciennes, 1846: 92-94 (original description; Canary Isian ds, Goree) .-Dumeri!, 1858: 264 1965] Staiger: Atlantic Flyingfishes 715 (Goree) .-GUnther, 1866: 287 (description; Madeira) .-Capello, 1872: 86 (St. Jago, Cape Verde; fide Breder, 1938: 49).-Goode, 1876: 76 (Bermuda; doubtful validity) .-Rochebrune, 1883: 139 (Goree, Dakar, Verde; fide Breder, ]938: 49).-Hilgendorf, 1888: 211 (Azores; fide Rreder, 1938: 49).-Blittikofer, 1890: 480). (fide Rreder, ]938: 49).- Steindachner, 1891: 361 (East of Arrecip; fide Breder, 1938: 49). - Steindachner, 1894: 75 (Grand Canary; fide Rreder, 1938: 49).-Jordan & Evermann, 1896b: 739 (description; Atlantic Ocean).-Osorio, 1898: 199 (listed; Cape Verde, St. Thome).-Pellegrin, 1905a: 28 (Madeira; fide Breder, 1938: 49) .-Pellegrin, 1905b: 137 (Cape Blanc, Cape Verde; fide Breder, 1938: 49).-Pellegrin, 1914: 30 (Canaries; fide Breder, 1938: 49).-Metzelaar, 1919: 220 (description; Gran Canaria). Cypsilurus lineatus. Jordan & Evermann, 1898: 2836 (listed) .-Bruun, 1935: 47-52 (description; key; synonymy; illustration; Eastern Atlantic).- Norman, 1935: 58 (listed; South At1antic).-Poll, 1953: 187-189 (description; illustration; Atlantic coast of North Africa). Exoeoetus fl/reatus. Roule & Angel, 1930: 85, pI. 4 fig. 110 (description; illustration; Azores, misidentification). Cypselurus minos Nichols & Breder, 1930: 1-3, 5, 8 (original description; illustration; Cape Verde Islands). Cypselurlls lineatus. Jordan, Evermann, & Clark, 1930: 202 (listed) .-Fow- ler, ]936: 429-431, 1243 (description; illustration; West Africa).-B!·eder, 1938: 47-52 (description; illustration; key; synonymy; Bermuda - doubts its validity) .-Lozano Rey, 1947: 623-625 (description; illustration; At- lantic coasts of Spain and Portugal). Cypselurus pinnatibarbatus. Hubbs & Kampa, 1946: 211-212 (description; synonymy) .-Abe, 1954: 195-202, 209-221 (description; illustration; synonymy; Japan).-Abe, 1960: 147-149 (description; Japan). Cheilopogon pinnatibarbatus. Parin, 1961: 105, 112, 132-133, 138, 171 (description; iIlustration; key).

Material examined.-None. Description from Bruun (1935), Breder (1938), Hubbs & Kampa (1946), Abe (1954, 1960), and Imai (1958).

Distinctive characters.-The .most distinctive juvenile character is the mandibular barbel, which is short, semicircular, and equipped with a number of fringe-like appendages radiating out from the basal portion. Adults may be distinguished by their high dorsal fin with a black spot (separating them from C. comatus, C. furcatus, and C. heterurus), high number of vertebrae (50-52) (separating them from all Atlantic Cypse- lurus), high number of predorsal scales (shared with C. cyanopterus), and gray pectoral fin with light posterior margin and vague light crossband which is often very difficult to see (separating it from C. cyanopterus). Description.-Cypselurus pinnatibarbatus is one of the largest Atlantic f1yingfishes. Bruun (1935: 52) mentions that most specimens were over 300 mm standard length, the largest measuring 378 mm standard length. Palatine teeth lacking in Atlantic specimens. The only specimen with obvious palatine teeth seen by Bruun came from the Cape of Good Hope. He suggested that it could be a member of an Indian Ocean subspecies 716 Bulletin of Marine Science [15(3) (1935: 96). Abe (1954: 196) states that C. pinnatibarbatus japonicus lacks palatine teeth. Premaxillary teeth unicuspid. Meristic data (study material from Bruun (1935) and Breder (1938»: dorsal-fin rays 12-14; anal-fin rays 10-12; dorsal minus anal-fin rays 0-4; pectoral-fin rays 14-16; predorsal scales 39-46; total gill rakers (first arch) 20-24; scales above lateral line 7-8; vertebrae 50-52. Morphometric data (study material from Bruun (1935) and Breder (1938), expressed as per cent of standard length): preanal length 76.1-83.1 per cent; predorsallength 71.4-77.6 per cent; prepelvic length 57.1-61.4 per cent; head length 20.3-21.8 per cent; snout length 4.6-5.9 per cent; fleshy orbit diameter 4.8-6.3 per cent; interorbital width 6.6-8.2 per cent; pectoral-fin length 65.5-75.7 per cent; pelvic-fin length 27.0-35.8 per cent; dorsal fin height 8.8-12.9 per cent; greatest depth 13.1-16.7 per cent; body width 10.7-13.4 per cent. The above ranges are based only upon measurements of adults, while data for the other species are based upon measurements of juveniles and adults. Body dark above, light below. Bruun (1935: 49) found no trace of the dark longitudinal stripe supposedly characteristic of the species. Abe (1954: 195) described C. pinnatibarbatus japonicus as blue-black dorsally (brown-black in formalin), whitish ventrally (brownish in formalin), with a dark blue longitudinal band on the side of the trunk. Pectoral fins gray with pale posterior margin and very light crossband; posterior few rays frequently without pigment. Pelvic fins gray, the pigment more highly concentrated proximally. Dorsal fin gray, a dark spot in center of fin. Anal fin clear. Caudal fin dark, the pigment more highly concentrated along the rays (juvenile fin light with several dark spots, as in C. cyanopterus) . Ontogeny.-A complete developmental series is lacking for the Atlantic forms of this species. Most of this discussion is based on work done on Pacific subspecies by Hubbs & Kampa (1946) and Abe (1954, 1960). The single median barbel, unlike that of C. comatus, does not grow to great proportions. The largest seen does not exceed the length of the head. It is present in specimens up to about 130 mm standard length. The dorsal fin in juveniles and small adults is very high (29 per cent of standard length in specimens of 100 mm standard length), has a rounded posterior margin, and is dark over most of its surface. This pigment is lost in the adult, resulting in a fin with a dark spot on its posterior margin. All specimens considered juvenile by Bruun had pectoral fins with pale crossbands, which he felt became indistinct with maturation. Abe (1954) has correlated the degree of distinctness of the crossband with the period of the day in which the specimen was collected: specimens caught at night showed an extremely indistinct crossband; those taken in the daylight 1965] Staiger: Atlantic Flyingfishes 717 showed a much more distinct crossband. If Bruun's collections were made at night with the aid of a night light this could account for his failure to recognize the crossband in the adults. Both juveniles and adults have a light margin of the posterior edge of the fin. The pelvic-fin pigmentation does not change with ontogeny. Range.-In the Atlantic Ocean Cypselurus pinnatibarbatus is a coastal species and has not been taken farther than 300 miles from shore. It is found in temperate as well as tropical waters. In the North Atlantic it is not found west of the Azores, but it has been taken in the South Atlantic at Rio de Janeiro, Brazil (Bruun, 1935: 95). Goode (1876) reported a single specimen from Bermuda, but Breder (1938: 49) doubts its validity. All of the other Atlantic records of C. pinnatibarbatus are from the west coast of Africa and the coast of Portugal. Cypselurus pinnatibarbatus is also found throughout the Indian and Pacific Oceans. Reproductive biology.-The smallest specimen whose sex was determined by Bruun (1935: 52) measured 292 mm standard length. No ripe specimens have been reported from the Atlantic.

Discussion.-Hubbs & Kampa (1946) synonymized Cypselurus lineatus (Valenciennes, 1846) with C. pinnatibarbatus (Bennett, 1831), indicating that C. pinnatibarbatus was undoubtedly the juvenile form of C. lineatus. Parin (1961) recognized six subspecies of Cheilopogon pinnatibarbatus. Cheilopogon pinnatibarbatus pinnatibarbatus is the Atlantic subspecies, the other five are confined to the Pacific and Indian Oceans.

SUMARIO

PECES VOLADORES DEL ATLANTICO DEL G:ENERO Cypselurus, CON DESCRIPCIONES DE LOS ESTADOS JUVENILES Se revisa el estatus sistematico de los peces voladores del Atlantico del genera Cypselurus y se reconocen seis especies: Cypselurus comatus, C. cyanopterus, C. exsiliens, C. furcatus, C. heterurus y C. pinnatibarbatus. Parin (1961) dividi6 el genero Cypselurus y situ6 cinco de las seis especies del Atlantico en el genero Chei/opogon. Su identificaci6n de Cheilopogon no es aceptada porque los caracteres usados para distinguir los dos generos no muestran efectivamente una separaci6n. Se presenta evidencia para apoyar el rechazo de Cheilopogon. Cypselurus /uetkeni, provisionalmente situado por Springer (1959) como sin6nimo de C. heterurus. Se present an evidencias que muestran que las dos especies no pueden ser validamente separadas, ni aun a nivel de sub-especie. De las seis especies de Cypselurus encontradas en el Atlantico sola- mente dos fueron encontradas a mas de 400 millas de la costa. Estas 718 Brilletin of Marine Science [15(3) dos especies son: C. exsiliens y C. furcatus, ambas viven en aguas tropicales. De las cuatro especies costeras, dos viven en aguas tropicalcs: C. comatus y C. cyanopterus, mientras que las dos restantes: C. heterurus y C. pinnatibarbatus, viven tanto en aguas templadas como tropicales. Todas las especies de Cypselurus muestran ciertos cambios ontogenic os en las estructuras durante el desarrollo. Estos cambios ontogenicos es- pecialmente aquellos pertenecientes al color, pueden ser valiosos para su identificaci6n. Los cambios ontogenicos en la pigmentaci6n de la aleta pectoral son usados para diferenciar especies facilmente confundidas como: C. heterurus y C. furcatus, y entre adultos de C. cyanopterus y C. pinnatibarbatus. Cypselurus comatus es una pequefia especie, can una longitud promedio de 200 mm aproximadamente. Las aletas pectorales y peIvicas de los adultos son grisaceas. La aleta dorsal es baja y ligeramente pigmentada y la aleta anal es clara. La aleta caudal es oscura. El numero de escamas predorsales varfa de 22 a 33. El numero de vertebras varfa de 40 a 44. Los juveniles tienen aletas pelvicas oscuras que se aclaran con el crecimiento. La aleta caudal se oscurece con el crecimiento. La barbilla mandibular media unica alcanza una longitud de mas del 100 por ciento del largo promedio, y se pierde al alcanzar longitud promedio de 160 mm aproximadamente. Cypselurus cyanopterus es una especie grande, can una longitud promedio de 300 mm. Las aletas pectorales de los adultos son uniformemente oscuras, mientras las aletas pelvicas son claras. La aleta dorsal es alta y tiene una mancha negra posterior. La aleta anal es clara. El numero de escamas predorsales varfa de 30 a 41. El numero de vertebras varia de 43 a 46. Los juveniles tienen aletas pelvicas oscuras que se convierten en claras con el crecimiento. Las aletas pectorales de los juveniles se oscurecen con el desarrollo y areas palidas aparecen y desaparecen. La aleta caudal se oscurece con el crecimiento. Las barbillas mandibulares pares alcanzan una longitud de mas del 150 por ciento de la longitud promedio y se pierden al alcanzar una longitud promedio de 200 mm. Cypselurus exsiliens es una especie de mediano tamafio, con una longitud promedio de 220 mm aproximadamente. Las aletas pectorales de los adultos son oscuras can una banda clara. La aleta dorsal es muy alta y tiene una mancha negra posterior. El lobulo caudal superior es claro, mientras el inferior es oscuro. El numero de escamas predorsales varfa de 21 a 30. £1 numero de vertebras varia de 43 a 45. Con el crecimiento las aletas pelvicas primero se oscurecen y despues desarrollan un fondo claro con algun pigmento a 10 largo de los radios. Las aletas pectorales de los juveniles son claras con dos bandas oscuras. Con el desarrollo las aletas se taman oscuras con una banda clara. Las barbillas mandibulares pares alcanzan una longitud aproximada del 40 por ciento de la longitud 1965] Staiger: Atlantic Flyingfishes 719 promedio y se pierden al alcanzar una longitud promedio de 120 mm. Cypselurus furcatus es una especie grande, con una longitud promedio de mas de 240 mm. Las aletas pectorales de los aduItos son oscuras con una banda clara oblicua que es estrecha en su parte proximal y ancha en la distal. La aleta dorsal es baja y clara en los aduItos. La aleta anal es clara. EI numero de escamas predorsales varia de 22 a 35. El numero de vertebras varia de 44 a 46. Las aletas pectorales de los juveniles tienen un numero de manchas oscuras en un fondo claro. Presenta una banda clara que tiene un caracteristico borde posterior recto. La aleta caudal oscurece con el crecimiento. Las barbillas mandibulares pares alcanzan una longitud de mas del 30 por ciento de la longitud promedio y se pierden al alcanzar una longitud aproximada de 120 mm. Cypselurus heterurus es una especie grande, que aIcanza mas de 280 mm de longitud promedio. La aleta dorsal en los adultos es baja y palida. La aleta anal es clara. Las aletas pectorales de los aduItos son oscuras con una banda oblicua palida. Esta banda es mas ancha proximalmente y es triangular en adultos grandes. La aleta dorsal es baja y palida. La aleta anal es clara. El numero de escamas predorsales varia de 22 a 38. El numero de vertebras varia de 42 a 49. Las aletas pectorales de los juveniles tiellen un numero de manchas oscuras en un fondo claro, como en C. furcatus, pero la banda clara no es de bordes rectos. La aleta caudal se oscurece con el desarrollo. Las aletas pelvicas de los juveniles, tienen un fondo claro con manchas oscuras. Con el desarrollo las aletas se aclaran. Las barbillas mandibulares pares aIcanzan una longitud de s610 15 por ciento de la longitud promedio y se pierden al aIcanzar una longitud promedio de 110 mm. Cypselurus pinnatibarbatus es una especie grande, que promedia bas- tante mas de 300 mm de longitud. La aleta dorsal en los adultos es alta y tiene una mancha oscura posterior. La aleta anal es clara. Las aletas pectorales de los adultos son oscuras con una vaga banda clara. Las aletas pelvicas de los adultos son grises. El numero de escamas predorsales varia de 39 a 46. EI numero de vertebras varia de 50 a 52. La aleta dorsal de los juveniles es oscura. Con el desarrollo el pigmento se con- centra en las parte porterior de la aleta. Las aletas pectorales de los juveniles son oscuras con una banda clara. La barbilla mandibular unica es dife- rente de las barbillas de las otras especies. En vez de ser filamentosa es semicircular, con apendices digitiformes pIanos. No excede del 20 por ciento de la longitud promedio y se pierde a los 130 mm de longitud promedio.

LITERATURE CITED

ABE, TOKIHARu 1954. Notes on the flying-fishes of Hachijo Island, with nomenclatorial remarks on the flying-fishesof the mainland of Japan and Hokkaido. 720 BuLLetin of Marine Science [J 5(3) II. Cypselurus pinnatibarhatus japonicus. With additional notes on Prognichthys agoo. Jap. J. Ichthyol., 3 (3-5): 193-202, 7 figs.; 3 (6): 209-222, 10 figs. 1960. Notes on fishes from the path of the "Kuroshiwo" with special ref- erence to the adaptation or preference of some fiying-fish for cool water. Rec. oceanogr. Wks. Jap., special no. 4: 147-150. AGASSIZ, LoUIS 1846. Nomenc1ator Zoologicus. Pisces. 8: 1-393. [Not seen.] BEAN, TARLETON H. 1903. Catalogue of the fishes of New York. N.Y. State Mus., Bull. 60, zoo1. 9: 784 pp. BEEBE, WILLIAM AND GLORIA HOLLISTER 1933. New species of fishes from the West Indies. Zoologica, N.Y., 12 (9): 83-88. BENNETT, EDWARD T. 1831. Characters of new genera and species of fishes from the Atlantic coast of northern Africa, presented by Capt. Belcher, R. N., Proc. Comm. Sci. and Corresp., Zool. Soc. London, 1: 145-148. BENNETT, F. D. 1840. Narrative of a whaling voyage round the globe. London, vol. 2: i-vii + ]-395 pp., text illust. BENNETT, GEORGE 1860. Gatherings of a naturalist in Australasia. London. i-xii + 1-456 pp., 8 pis., text il\ust. BIGELOW, HENRY B. AND WILLIAM C. SCHROEDER 1953. Fishes of the Gulf of Maine. Fish. Bull., U.S., 53 (74): i-viii + 1-577 pp., 288 figs. BLEEKER, PIETER 1866. Atlas Ichthyologique des Indes Orientales Neerlandaises, 6: 1-170, 46 pIs. BREDER, CHARLES M., JR. 1927. Scientific results of the first oceanographic expedition of the "Pawnee" 1925: Fishes. Bull. Bingham oceanogr. Coll.,l (1): 1-90. 1928. Scientific results of the second oceanographic expedition of the "Pawnee" 1926: Nematognathi, Apodes, Isospondyli, Synentognathi, and Thoracostraci from Panama to Lower California. Bull. Bingham oceanogr. Col1., 2 (2): 1-25. 1932. On the habits and development of certain Atlantic Synentognathi. Pap. Tortugas Lab., 28 (1): 1-35, 12 pis. 1938. A contribution to the life histories of Atlantic Ocean flyingfishes. Bull. Bingham oceanogr. Coli., 6 (5): 1-126, 48 figs. BREDER, CHARLES M., JR. AND JOHN T. NICHOLS 1930. West Indian forms of the flying fish, genus Cypselurus, with the description of a new species. Amer. Mus. Novit., 417: 1-9. 1934. On the significance of vertebral counts in exocoetid taxonomy. Proc. bioI. Soc. Wash., 47: 37-44. BRUUN, ANTON F. 1933. On the value of the number of vertebrae in the classification of the Exocoetidae. Vidensk. Medd. dansk naturh. Foren. Kbh.,94: 375-384. 1935. Flying-fishes (Exocoetidae) of the Atlantic. Dana Rep., 6: 1-106, 7 pis. 1938. A new occurrence of flying-fish (Cypselurus heterurus) in Oslo Fjord. NyU Mag. Naturvidensk., 78: 295-299. 1965] Staiger: Atlantic Flyingfishes 721 BUTTIKOFER, J. ] h~Ll. kelsebilder aus Liberia. Leiden, 2: i-viii + 1-510 pp., 32 pis., text illust. CADENAT, JACQUES AND E. MARCHAL 1963. Resultats des campagnes oceanographiques de la Reine-Pokou aux iles Sainte-Helene et Ascension. Poissons. Bull. Inst. franc;ais d'Afrique Noire, 25 (A, 4): 1235-] 315, 48 figs. CAPELLO, FELIX DE BRITO 1872. Primeira !ista dos peixes da i1ha da Madeira, Ac;ores, e das possess5es portuguezas d'Africa, que existem no Museo de Lisboa. J. Sci. Math. Phys. Nat. Lisboa, 4: 83-88. [Not seen.1 CUVIER, GEORGES AND ACHILLE VALENCIENNES 1846. Histoire naturelle des poissons. Paris, F. G. Levrault, 19: i-xix + 1-544 + [6 unnumbered] pp., pIs. 520-590. DEKAY, JAMES E. 1842. Zoology of New York, or the New York Fauna. 3. Reptiles and Amphibia. 4. Fishes. Albany, W. & A. White & J. Visscher, 1-415 pp., 79 pis. DUMERIL, AUGUSTE H. A. 1858. Reptiles et poissons de I'Afrique occidentale. Arch. Mus. Rist. nat., Paris, 10: 137-266. EVERMANN, BARTON W. AND MILLARD C. MARSH 1900. The fishes of Puerto Rico. Bull. U.S. Fish Comm., 20: 51-350,49 pIs. FILIPPI, FILIPPO DE AND JEAN B. VERANY 1859. Sopra alcuni pesci nuovi 0 poco noti del Mediterraneo. Mem. Acad. Sci. Torino., 2 series, 18: 187-199. [Not seen.] FOWLER, HENRY W. 1919. Notes on synentognathous fishes. Proc. Acad. nat. Sci. Philad., 71 (1): 2-15. 1928. The fishes of Oceania. Mem. Bishop Mus., 10: i-iii + 1-540 pp., 49 pIs., 82 figs.. 1931. The fishes of Oceania. Supplement 1. Mem. Bishop Mus., 11 (5): 313-381,7 figs. 1934a. Fishes obtained by Mr. H. W. Bell-Marley chiefly in Natal and Zulu- land in 1929 to 1932. Proc. Acad. nat. Sci. Philad., 86: 405-514, 53 figs. 1934b. The fishes of Oceania. Supplement 2. Mem. Bishop Mus., 11 (6): 385-466, 4 figs. 1934c. Nata] fishes obtained by Mr. H. W. Bell-Marley. Ann. Natal Mus., 7 (3): 412. 1936. The marine fishes of West Africa. Bull. Amer. Mus. nat. Rist., 70 (1): i-vii + 1-605 pp., 275 figs. GILL, THEODORE 1863. Descriptive enumeration of a collection of fishes from the Western coast of Central America, presented to the Smithsonian Institution by Captain John M. Dow. Proc. Acad. nat. Sci. Philad., (1864): 162- 184. GMELIN, JOHANN F. 1789. System a naturae, ed. 13. G. E. Beer, Lipsiae. 1 (3): ]033-1516. GOODE, GEORGE BROWN 1876. Catalogue of the fishes of the Bermudas. Based chiefly upon the collection of the United States National Museum. Bull. U.S. nat. Mus., 1 (5): 1-82. 722 Bulletin of Marine Science li5(3)

GUNTHER, ALBERT 1866. Catalogue of the Physostomi, containing the families Salmonidae, Percopsidae, Galaxiidae, Mormyridae, Gymnarchidae, Esocidae, Um- bridae, Scombresocidae, Cyprinodontidae, in the collection of the British Museum. Cat. fishes Brit. Mus., 6: i-xv + 1-368 pp., 33figs. 1869. Report of a second collection of fishes made at St. Helena by J. C. Melliss, Esq. Proc. Zool. Soc. London, 1869: 238-239. 1899. Complete catalogue of Linne's private collection of fishes now in possession of the Linnean Society. Proc. Linn. Soc. London (Zool.), 1899: 15-38. 1909. The type of Exocoetus exiliens (L. Gmelin). Ann. Mag. nat. Hist., ser. 8, 3 (14): 147-149. HERRE, ALBERT C. 1935. New fishes obtained by the Crane Pacific Expedition. Field Mus. Nat. Hist. Zool. Ser., 18 (12): 383-438. 1936. Fishes of the Crane Pacific Expedition. Field Mus. Nat. Hist. Zool. Ser., 21: 1-472,50 figs. HILGENDORF, FRANZ M. 1888. Die Fische der Azoren. Arch. Naturgesch., 54: 205-213. [Not seen.] HUBBS, CARL L. AND ELIZABETH M. KAMPA 1946. The early stages (egg, prolarva, and juvenile) and the classification of the California flyingtish. Copeia, 1946 (4): 188-218, 5 figs. HUBBS, CARL L. AND KARL F. LAGLER 1958. Fishes of the Great Lakes region. Bull. Cranbrook lnst. Sci., 26: i-xiii + 1-213 pp., 251 figs., 44 pis. 1M AI, SADAHIKO 1958. Nihon Kinkaisan no Tobiuo no Kensaku. [Key of young and adult of flying-fishes collected from around waters of Japan.] In Nihonsan Gyorui no Chigyoki no Kenkyu. Studies on the eggs, larvae and juvenile of Japanese fishes. Series I. Kyushu Daigaku, Nogaku Bu, Suisan Daini Kyoshitsu. Second Laboratory of Fisheries Biology, Fisheries Department, Faculty of Agriculture, Kyushu University, Fukuoka, Japan: 26-44, pIs. 30-43. JORDAN, DAVID STARR AND BARTON WARREN EVERMANN 1896a. A Check-list of the fishes and fish-like vertebrates of North and Middle America. Rep. U. S. Comm. Fish, 21 (1894-1895): 207-590. 1896b. The fishes of North and Middle America. Bull. U. S. nat. Mus., 47 (1): i-iv + 1-1240 pp. 1898. The fishes of North and Middle America. Bull. U. S. nat. Mus., 47 (3): i-xxiv + 2183-3136 pp. JORDAN, DAVID STARR, BARTON W. EVERMANN, AND HOWARD W. CLARK 1930. Check list of the fishes and fishlike vertebrates of North and Middle America north of the northern boundary of Venezuela and Colombia. Rep. U. S. Comm. Fish., 1928, app. 10: 1-670. JORDAN, DAVID STARR AND CHARLES H. GILBERT 1882. Synopsis of the fishes of North America. Bull. U. S. nat. Mus., (16): 1-1018. JORDAN, DAVID STARR AND SETH E. MEEK 1886. A review of the American species of flying fishes (Exocoetus). Proc. U. S. nat. Mus., 1885, 8: 44-67. JORDAN, DAVID STARR AND ALVIN SEALE 1906. The fishes of Samoa. Bull. Bur. Fish., 25: 173-455, 110 figs., pIs. 38-53. 1965] Staiger: Atlantic Flyingfishes 723

LESUEUR, CHARLES A. 1821. Description of two new species of Exocoetus. J. Acad. nat. Sci. Philad., 2 (1): 8-11. LILLJEBORG, WILHELM 1891. Sveriges och Norges Fiskar. Malacopterygii, Plectognathi, Lopho- branchii. Uppsala, 2: 788 pp. LINNAEUS, CAROLUS 1771. Mantissa altera. Cum appendice Regni Animalis. Laurentii Salvii, Holmiae, 446 pp. LONGLEY, WILLIAM H. AND SAMUEL F. HILDEBRAND 1941. Systematic catalogue of the fishes of Tortugas, Florida. Pap. Tortugas Lab., 34: 1-331, 34 pis. LOWE, RICHARD T. 1840. On certain new species of fish from Madeira. Proc. Zool. Soc. London, 8: 36-39. 1841. Description of some new species of Madeiran fishes, with additional information relating to those already described. Ann. Mag. nat. Rist., 7: 92-94. LOZANO REY, LUIS 1947. Peces ganoidecs y fisostomos. Mem. Real Acad. Ciencias Madrid, 11: i-xv, 1-839. UiTKEN, CHRISTIAN F. 1876. Ichthyographiske bidrag. VI Bidrag til Flyvefishenes (Exocoeterens) diagnostik. Vidensk. Mcdd. dansk naturh. Foren. Kbh., (1877): 99- 114, 389-408. [Not seen.] MARSHALL, NELSON B. 1950. Fishes of the Cocos-Keeling Islands. Bull. Raffles Museum, 22: 170- 172. MARTENS, EDUARD VON 1865. Die preussische Expedition nach Ost-Asien. Zoologische Abtheilung. Berlin 1 (1): i-xii + 1-192 pp. 1867. Die preussische Expedition nach Ost. Asien. Zoologische Abtheilung. Berlin, 1 (2): 193-412, 15 pIs. MELUSS, JOHN C. l875. St. Helena. London. i-xiv + 1-426 pp., 62 pIs., 1 map, text ilIust. METZELAAR, JAN 1919. Report on the fishes, collected by Dr. J. Boeke, in the Dutch West Indies, 1904-1905, with comparative notes on marine fishes of Trop- ical West Africa. Rapport betreffende een voorloopig onderzoek naar den toestand van de Visscherij en de Industrie van Zeeproducten in de Kolonie Curacao: 1-314 pp., 64 figs. MITCHILL, SAMUEL L. 1815. The fishes of New York, described and arranged. Trans. Lit. Phil. Sec. New York, 1814 1: 355-492. MULLER, JOHANNES 1843. Beitdige zur Kenntniss der natlirlichen Familien der Fische. Arch. Naturgesch., 9 (1): 292-330, 381-384. MULLER, PHILIP L. S. 1776. Des Ritters Carl von Linne, Koniglich Schwedischen Leibarztes, etc. vollsUindigen Natursystems. 6 Supplements- und Register- Band. Nurnberg. [14 unnumbered] + 1-920 pp., 3 pIs. NICHOLS, JOHN T. ]924. A contribution to the ichthyology of the Galapagos. Zoologica, N.Y., 5 (4): 63-65, 1 fig. 724 Bulletin of Marine Science [15(3)

NICHOLS, JOHN T. AND CHARLES M. BREDER, JR. 1927. The marine fishes of New York and southern New England. Zool- ogica, N.Y., 9 (1): 1-192,261 figs. 1928. An annotated list of the Synentognathi. Zoologica, N.Y., 8 (7): 423-448. 1930. A key to the Atlantic species of the genus Cypselurus, with a new fiyingfish from the Cleveland Museum's "Blossom" expedition. Amer. Mus. Novit., 428: 1-8. NORMAN, JOHN R. 1935. Coast fishes part I. The South Atlantic. Discovery Rep., 12: 1-58, 15 figs. OGILBY, J. DOUGLAS 1908. On new genera and species of fishes. Proc. Roy. Soc. Queens]and, 21: 1-26. 1914. On new genera and species of fishes. Proc. Roy Soc. Queensland, 22: 13-39. OSORIO, BALTHAZAR 1895. Peixes de Dahomey. Jor. Sci. Acad: Lisboa, ser. 2, 3: 252-253. 1898. Da distribuicas geographica dos peixes e crustaceos colhidos nas possessoes portuguezas d'Africa occidental e existentes no Museo Nacional de Lisboa. Jor. Sci. Acad. Lisboa, 2 ser., 5: 185-202. PARIN, N. V. 1960. Tbe flying fishes (Exocoetidae) of the northwest Pacific. IPST Cat. No. 618. Translation of "Letuchie ryby (Exocoetidae) severo-zapad- noi chasti Tikhogo Okeana." Trud. Inst. Okeanol., 31: 205-285, 25 figs. 1961. The basis of classification of the fiying fishes (families Oxypor- hamphidae and Exocoetidae). Trud. Inst. Okeanol., 43: 92-183. (In Russian.) PARR, ALBERT E. 1930. Teleostean shore and shallow-water fishes from the Bahamas and Turks Island. Bull. Bingham oceanogr. Coli., 3 (4): 1-148. PELLEGRIN, JACQUES 1905a. Mission des pecheries de la cote occidentale d'Afrique. Poissons. Act. Soc. Linn. Bordeaux, 60: ]7-57. [Not seen.] 1905b. Mission des pecheries de la cote occidentale d'Afrique dirigee par M. Gruvel. Poissons. Bull. Soc. zool. Fr., 30: 135-141. [Not seen.] 1914. Missions Gruvel sur la cote occidentale d'Afrique (1905-1912). Ann. Inst. oceanogr., Monaco, 6 (4): 1-99. [Not seen.] POEY Y ALOY, FELIPE 1860. Poissons de Cuba. Memorias sobre la historia natural de la isla de Cuba, 2 (49): 115-356. 1868. Synopsis piscium Cubensium. Repertorio fisico-natural de ]a isla de Cuba, 2: 279-464. POLL, MAX 1953. Poissons. III.-Teleosteans Malacopterygians. Result. sci. Exped. oceanogr. beIge Atlant. Sud. 4 (3): 1-258, 104 figs., 8 pIs. RAFINESQUE, CONSTANTINE S. 1810. Caratteri di alcuni nuovi generi e nuovi specie di animali e piante della Sicilia. Palermo. 105 pp., 20 pIs. RANZANI, CAMILLO 1842. De novis speciebus piscium. Dissertatio quattuor. Novi Comment. Acad. Sci. Inst. Bononiensis, 5: 3-22, 307-338, 339-366. 1965J Staiger: Atlantic Flyingfishes 725 ROCHEBRUNE, ALPHONSE T. DE 1883. Fauna de la Senegambie. Act. Soc. Linn. Bordeaux, ser. 4, 1882, 6: 37-180. ROULE, LOUIS 1919. Poissons provenant des campagnes du Yacht Princesse-Alice (1891- 1903) et du Yacht HirondeIle II (1914). Result. Camp. sci. Monaco, 52: 1-190, 7 pIs. ROULE, LoUIs AND FERNAND ANGEL 1921. Notice preliminaire (He) sur les larves et les alevins de Poissons recueillis par S. A. S. Ie Prince de Monaco dans ses croisieres de 1896 a 1901 (indus). Bull. lnst. oceanogr. Monaco, 397: 1-8. 1930. Larves et alevins de poissons provenant des croisieres du Prince Albert ler de Monaco. Result. Camp. sci. Monaco, 79: 1-148, 6 pIs. SMITH, HUGH M. 1907. The fishes of North Carolina. N. Carolina Geol. and Econ. Survey, 2, Raleigh, 1-453, 188 figs., 21 pIs. SMITH, J. L. B. 1950. The sea fishes of southern Africa. Central News Agency Limited, Johannesburg. i-xvi + 1-550 pp., 1232 figs., 102 pIs. 1961. The sea fishes of southern Africa. Ed. 4. Central News Agency Limited, Johannesburg. i-xv + 1-580 pp., 1232 figs., 107 pIs. SMITH, J. L. B. AND MARGARET M. SMITH 1963. The fishes of the SeycheIles. Rhodes Univ., Grahamstown, 1-215, 98 pis. SPRINGER, VICTOR G. 1959. Cypselurus luetkeni (Jordan and Evermann) a synonym of Cypse- lurus heterurus (Rafinesque)? Copeia, 1959 (2): 166. STEINDACHNER, FRANZ 1891. Ichthyologische Beitrage. XV Sitzber. Akad. Wiss. Wien, 100 (1): 343-347. [Not seen.] 1894. Die Fische Liberia's. Notes Leyden Mus., 16: 1-96. [Not seen.] STORER, DAVID H. 1846. A synopsis of the fishes of North America. Mem. Amer. Acad. Arts Sci., 2: 253-550. SWAINSON, WILLIAM 1839. On the natural history and classification of fishes, amphibians, and reptiles. London, 2: i-vi + 1-448 pp., 135 figs. VAILLANT, LEON L. 1919. Appendice L. Roule; Poissons proven ant des campagnes du Yacht Princesse-Alice (1891-1903) et du Yacht Hirondelle II (1914). Result. Camp. sci. Monaco, 52: 1-190, 7 pis. VINCIGUERRA, DEClO 1883. Risultati ittiologici delle crociere del Violante." Ann. Mus. Stor. nat. Genova, 18: 465-590, pIs. 1-3. WHITLEY, GILBERT P. AND ALAN N. COLEFAX 1938. Fishes from Nauru Gilbert Islands, Oceania. Proc. Linn. Soc. N. S. Wales, 63: (3-4): 287. WOOD, WILLIAM W. 1825. Descriptions of four new species of the Linnaean genus Blennius and a new Exocoetus. J. Acad. nat. Sci. Phil ad., 4 (1): 283.