*6;?A ■ ■ ■ ; 1 i ZOOLOGICA 6 6 SCIENTIFIC CONTRIBUTIONS OF THE NEW YORK ZOOLOGICAL SOCIETY VOLUME XXVI 1941 Numbers 1-30 PUBLISHED BY THE THE ZOOLOGICAL PARK, SOCIETY NEW YORK NEW YORK ZOOLOGICAL SOCIETY General Office: 630 Fifth Avenue, New York City OFFICERS President, Fairfield Osborn First Vice-President, Alfred Ely Chairman, Executive Committee & Second Vice-President, Laurance S. Rockefeller Treasurer, Cornelius R. Agnew General Director, Zoological Park and Aquarium, Allyn R. Jennings Assistant General Director, Harry Sweeny, Jr. Zoological Park Raymond L. Ditmars, Curator of Reptiles and Insects Lee S. Crandall, Curator of Birds Claude W. Leister, Curator of Mammals and Educational Activities Leonard J. Goss, Curator of Health William Bridges, Editor and Curator of Publications Aquarium in the Zoological Park Charles M. Breder, Jr., Director Christopher W. Coates, Aquarist Ross F. Nigrelli, Pathologist G. M. Smith, Research Associate in Pathology Homer W. Smith, Research Associate in Physiology Myron Gordon, Research Associate in Genetics Department of Tropical Research William Beebe, Director John Tee-Van, General Associate Gloria Hollister, Research Associate Jocelyn Crane, Technical Associate Consultant, Jean Delacour SCIENTIFIC STAFF Editorial Committee Fairfield Osborn, Chairman Allyn R Jennings William Beebe Charles M. Breder, Jr. Harry Sweeny, Jr Jean Delacour William Bridges CONTENTS. PAGE Part 1. May 26., 1941. 1. A New Crayfish from San Luis Potosi, Mexico. (Decapoda, Astacidae). By Horton H. Hobbs, Jr. (Text-figure 1) 1 2. A New Corydoras from Brazil. By F. R. LaMonte 5 3. Notes on Plumage Changes in the Bald Eagle. By Lee S. Crandall. (Plates I-IV) 7 4. External Characters of Six Embryo Nurse Sharks, Ginglymostoma cirratum (Gmelin). By William Beebe. (Plates I & II; Text-figures 1-4) 9 5. A Papillomatous Disease of the Gallbladder Associated with Infection by Flukes, Occurring in the Marine Turtle, Chelonia mydas (Linnaeus). By G. M. Smith, C. W. Coates & R. F. Nigrelli. (Plates I-IV; Text-figure 1) . . . . 13 6. Eastern Pacific Expeditions of the New York Zoological Society. XXIII. Polychaetous Annelids from the West Coast of Mexico and Central America. By Aaron L. Treadwell. (Text-figures 1-21) 17 7. Plankton of the Bermuda Oceanographic Expeditions. X. Polychaetous An- nelids from Bermuda Plankton, with Eight Shore Species, and Four from Haiti. By Aaron L. Treadwell. (Text-figures 1-9) 25 3. Caudal Skeleton of Bermuda Shallow Water Fishes. V. Order Percomorphi: Carangidae. By Gloria Hollister. (Text-figures 1-20) 31 ). Description of an Egg of the Long-tailed Bird of Paradise. By Lee S. Crandall. (Plate I) 47 ). On the Uterine Young of Dasyatus sabinus (LeSueur) and Dasyatus hastatus (De Kay). By C. M. Breder, Jr., & Louis A. Krumholz. (Text-figures 1 & 2) . ' 49 11. Additional Social and Physiological Aspects of Respiratory Behavior in Small Tarpon. By Arthur Shlaifer 55 12. Notes on Mexican Snakes of the Genus Trim ere surus. By Hobart M. Smith. . 61 Part 2. September 8, 1941. 13. The Life History and Bionomics of the Trematode, Zygocotyle lunata (Param- phistomidae). By Charles H. Willey. (Plates I-IV) 65 14. Eastern Pacific Expeditions of the New York Zoological Society. XXIV. Fishes from the Tropical Eastern Pacific. [From Cedros Island, Lower California, South to the Galapagos Islands and Northern Peru.] Part 1. Lancelets and Hag-fishes. By William Beebe & John Tee-Van. (Text- figures 1 & 2) 89 15. Eastern Pacific Expeditions of the New York Zoological Society. XXV. Fishes from the Tropical Eastern Pacific. [From Cedros Island, Lower California, South to the Galapagos Islands and Northern Peru.] Part 2. Sharks. By William Beeb^ & John Tee-Van. (Plates I & II; Text- figures 1-34) 93 PAGE 16. Correlations Between Structural Eye Defects and Behavior in the Mexican Blind Characin. By Charles M. Breder, Jr., & Edward B. Gresser. (Plates I— IV ; Text-figures 1 & 2) 123 17. On the Species of Otus scops. By Jean Delacour. (Text-figures 1-10) 133 18. A New Genus of Kaleege Pheasants. By Marquess Hachisuka 143 Part 3. October 31, 1941. 19. Eastern Pacific Expeditions of the New York Zoological Society. XXVI. Crabs of the Genus Uca from the West Coast of Central America. By Jocelyn Crane. (Plates I-IX; Text-figures 1-8) 145 20. Eastern Pacific Expeditions of the New York Zoological Society. XXVII. A Study of Young Sailfish ( Istiophorus ). By William Beebe. (Plates I-V; Text-figures 1-9) 209 21. On the Reproduction of Opsanus beta Goode & Bean. By Charles M. Breder, Jr. (Plates I & II) 229 22. On the Reproductive Behavior of the Sponge Blenny, Paraclinus marmoratus (Steindaehner). By Charles M. Breder, Jr. (Plates I— III ; Text- figure 1 ) 233 23. The Chromatophores of Fundulus hetcroclitus in Polarized Light. By A. M. Shanes & Ross F. Nigrelli. (Plates I— III) 237 24. New Races of Alaudidae and Timaliidae from Northern Thailand. By H. G. Deignan 241 25. Respiratory Behavior in Fishes Not Especially Modified for Breathing Air Under Conditions of Depleted Oxygen. By Charles M. Breder, Jr. (Plate I) 243 26. Eastern Pacific Expeditions of the New York Zoological Society. XXVIII. Fishes from the Tropical Eastern Pacific. [From Cedros Island, Lower California, South to the Galapagos Islands and Northern Peru.] Part 3. Rays, Mantas and Chimaeras. By William Beebe & John Tee-Van. (Plates I-IV ; Text-figures 1-40) 245 27. Erotylidae of Kartabo, Bartica District, British Guiana. (Coleoptera.) By C. H. Curran. (Plate I; Text-figure 1) 281 Part 4. December 29, 1941. 28. Further Studies on the Light Sensitivity and Behavior of the Mexican Blind Characin. By C. M. Breder, Jr., & Edward B. Gresser. (Plate I; Text-figures 1 & 2) 289 29. Eastern Pacific Expeditions of the New York Zoological Society. XXIX. On the Growth and Ecology of Brachyuran Crabs of the Genus Ocypode. By Jocelyn Crane. (Plates I & II; Text-figures 1-7) 297 30. High Speed Photographs of Flying Fishes in Flight. By H. E. Edgerton & C. M. Breder, Jr. (Plates I-VIII) 311 Index to Volume XXVI 315 ZOOLOGICA SCIENTIFIC CONTRIBUTIONS OF THE NEW YORK ZOOLOGICAL SOCIETY 0- VOLUME XXVI Part 1 Numbers 1-12 - 'h» PUBLISHED BY THE SOCIETY THE ZOOLOGICAL PARK, NEW YORK May 26, 1941 CONTENTS PAGE 1. A New Crayfish from San Luis Potosi, Mexico. (Decapoda, Asta- cidae). By Horton H. Hobbs, Jr. (Text-figure 1) 1 2. A New Corydoras from Brazil. By F. R. LaMonte 5 3. Notes on Plumage Changes in the Bald Eagle. By Lee S. Cran- dall. (Plates I-IV) 7 4. External Characters of Six Embryo Nurse Sharks, Ginglymostoma cirratum (Gmelin). By William Beebe. (Plates I & II; Text- figures 1-4) 9 5. A Papillomatous Disease of the Gallbladder Associated with In- fection by Flukes, Occurring in the Marine Turtle, Chelonia my das (Linnaeus). By G. M. Smith, C. W. Coates & R. F. Nigrelli. (Plates I-IV; Text-figure 1) 13 6. Eastern Pacific Expeditions of the New York Zoological Society. XXIII. Polychaetous Annelids from the West Coast of Mexico and Central America. By Aaron L. Treadwell. (Text-figures 1-21) 17 7. Plankton of the Bermuda Oceanographic Expeditions. X. Poly- chaetous Annelids from Bermuda Plankton, with Eight Shore Species, and Four from Haiti. By Aaron L. Treadwell. (Text-figures 1-9) 25 8. Caudal Skeleton of Bermuda Shallow Water Fishes. V. Order Percomorphi: Carangidae. By Gloria Hollister. (Text- figures 1-20) 31 9. Description of an Egg of the Long-tailed Bird of Paradise. By Lee S. Crandall. (Plate I) 47 10. On the Uterine Young of Dasyatis sabinus (Le Sueur) and Dasyatis hastatus (De Kay). By C. M. Breder, Jr., & Louis A. Krumholz. (Text-figures 1 & 2) 49 11. Additional Social and Physiological Aspects of Respiratory Behavior in Small Tarpon. By Arthur Shlaifer 55 12. Notes on Mexican Snakes of the Genus Trimeresurus. By Hobart M. Smith 61 ZOOLOGICA SCIENTIFIC CONTRIBUTIONS OF THE NEW YORK ZOOLOGICAL SOCIETY 1. A New Crayfish from San Luis Potosi, Mexico.1 (Decapoda, Astacidae). Horton H. Hobbs, Jr. (Text-figure 1). The new crayfish here described was taken by the New York Aquarium Cave Expedition to La Cueva Chica, a limestone cave in the State of San Luis Potosi, Mexico. An interesting ac- count of this cave was published by Mr. William Bridges in the Bulletin of the New York Zo- ological Society (Vol. XLIII, No. 3) in June, 1940. In this paper Mr. Bridges mentions the occurrence of crayfish in the cave, and at my request Dr. Charles M. Breder, Jr., of the New York Aquarium, and Mr. Marshall B. Bishop, of the Peabody Museum, have kindly lent me the 34 crayfish taken from the cave, and given me permission to describe them. Upon ex- amining this series I find that while they are closely allied to one of the most common species in the United States, they are distinct. During December, 1939, an expedition from the Department of Biology of the University of Florida collected in the states of Nuevo Leon and Tamaulipas, Mexico. A fine series of an apparently new subspecies of Cambarus bland- ingii was taken one mile north of Villa Juarez, Tamaulipas, but was not described since the collection contained no first form males. After having compared specimens of bland- ingii acutus from the southern United States with the Cueva Chica and Tamaulipas bland- ingii, I find that the specimens from Tamau- lipas are intermediate between these two, and that the Cueva Chica material represents a new race of blandingii. Cambarus blandingii cuevachicae, new subspecies. Diagnosis. Cambarus blandingii cuevachicae is a member of the subgenus Ortmannicus (i. e., the first pleopod of the first form male termi- nates in four distinct parts, and hooks are present on the ischiopodites of the third and fourth pereiopods). Rostrum broad with small lateral teeth. Areola practically obliterated (hardly broad enough to bear a single row of punctations). Carapace strongly granulate ex- cept along median dorsal surface. First pleopod of first form male similar to that of C. blandingii acutus. (See Text-fig. 1, B and D). Holotypic Male (Form I). Body subovate; compressed laterally. Abdomen narrower than thorax (1.54-1.91 cm. in widest parts, respec- tively). Width of carapace greater than depth in re- gion of caudodorsal margin of cervical groove. Greatest width of carapace about midway be- tween caudodorsal margin of cervical groove and caudal margin of cephalothorax. Areola almost obliterated in middle (hardly broad enough to bear a single row of puncta- tions) ; cephalic section of carapace about 1.7 times as long as areola (length of areola 35.2% of entire length of carapace) . Upper surface of rostrum deeply excavate; margins slightly convex distad of base, taper- ing and forming minute tubercles at base of acumen. Acumen short and broad, extending to base of distal segment of peduncle of anten- nule. Upper surface punctate; lateral margins with setiferous punctations almost to tip of apex. Subrostral ridge not evident in dorsal view. Postorbital ridges terminate cephalad in small tubercles, not spiniform. Suborbital angle absent. Branchiostegal spines present as large acute tubercles. •A contribution from the Department of Biology, Uni- versity of Florida, Gainesville, Florida. 1 2 Zoologica: New York Zoological Society [XX vl: 1 Surface of carapace strongly granulate except on dorsomedian surface of cephalic region; here, with setiferous punctations; one larger acute tubercle on either side in place of lateral spine, which is flanked by two or three slightly smaller ones. Abdomen slightly shorter than carapace (3.6- 3.79 cm.). Anterior section of telson with one spine in each posterolateral corner. Epistome broader than long ; subminaret shaped ; margins not raised ; faveolus present at base; small obtuse tubercle on median cephalic border. Eyes well developed. Antennules of the usual form with a well devel- oped spine on ventral surface of basal segment. Antennae reaching to caudal margin of telson. Antennal scale broad (broadest in middle) with a moderately well developed spine on outer dis- tal margin, reaching to distal segment oi peduncle of antennule. First right pereiopod long and relatively slender. Hand entirely tuberculate. A single row of eight strong tubercles along inner margin of palm subtended dorsad by two weaker rows and a few scattered tubercles. Very weak ridge present on upper surface of immovable finger. Another weak ridge present on outer distal mar- gin of same finger. Movable finger of right chela with minute tu- bercles along distal half of inner margin (crowded on distal third) . Lateral margin convex laterad ; two distinct rows of tubercles present on lateral margin along proximal two-thirds; an upper row of 24, arising near base of finger, and a lower row of 17 tubercles originating at base of distal three-fourtlis. A few scattered tubercles pres- ent on all surfaces near base of finger. Proximal two-fifths of outer margin with five tubercles. Distal upper, lower, and mesial surfaces setose punctate. Immovable finger of right chela with minute denticles on distal two-thirds (crowded on distal third). Mesial margin concave laterad; two distinct rows of tubercles on mesial margin: the upper row with 20 tubercles along the proximal two-thirds (the fifth is decidedly larger and more conspicuous than the others) and the lower row of 13 along the proximal three- fourths (the seventh is decidedly the largest of this row). When the lingers are brought to- gether the large tubercle on the upper side of finger lies over the movable finger and the large one on the lower row extends beneath it. Carpus longer than wide, much shorter than inner margin of palm of chela; shallow irregular furrow above. Mesial and upper mesial sur- faces tuberculate ; four large spike-like tubercles on mesial distal half; upper lateral, lateral, and ventral surfaces with setiferous punctations. Merus with dorsal, ventral, distal mesial, and distal lateral surfaces tuberculate; proximo- mesial and proximolateral surfaces with a few scattered punctations. Fourteen tubercles in a row on upper margin. Lower mesial margin with a row of 19 tubercles, and lower lateral margin with a row of 15 tubercles. Hooks on ischiopodites of third and fourth pereiopods. Both hooks are slender and simple. Coxopodite of fourth pereiopod with a large knob-like protuberance on caudomesial margin. First pleopod extending almost to coxopodite of second pereiopod when the abdomen is flexed. Tip terminating in four distinct parts. The mesial process is subspiculiform and is directed distad and laterad; is exceeded by the other processes distad. The cephalic process is corne- ous and blade-like, somewhat rounded distad, and extends distad of the other processes; it partially overhangs the central projection. The caudal process, the least conspicuous of the four distinct terminals, lies along the caudo-lateral margin of the central projection and extends barely distad of the mesial process. This (the caudal) process is corneous, and its tip is acute. The central projection is the largest of the terminal elements; it is corneous, concave mesiad and somewhat twisted. This projection is made up of two parts, the centro-cephalic which arises just cephalad of the caudal proc- ess on the lateral surface, and the centro-caudal which is somewhat larger and forms the mesial face of the central projection. These two proc- esses are fused along an oblique line visible in lateral and caudo-lateral views only. On the lateral surface of the appendage a conspicuous knob bearing long setae arises at the base of the cephalic process. Male Paratype (Form II). Differs from the male holotype in the following respects: Height of carapace greater than width (1.50-1.31 cm.). Postorbital ridges terminate cephalad in small acute tubercles. Branchiostegal spines moder- ately strong. Antennae reach anterior section of telson. Chelae smaller with less well defined tubercles, double row present on lateral margin of movable finger only. First pleopod termi- nating in four distinct parts; none corneous nor as sharply defined as in the first form male. (See Text-fig. 1, F and M). Hooks on ischio- podites of third and fourth pereiopods much reduced. Female Allotype. Differs in the following points from the first form male, holotype: Width of carapace slightly less than depth. Right subrostral ridge barely evident just caudad of base of acumen. Branchiostegal spines moderately strong. Lengths of abdomen and carapace subequal. Anterior section of telson with two spines in right posterolateral corner and none in the left (broken). Epistome slightly broader and shorter than in male. Antennae reaching fifth abdominal segment. Chelae smaller and tubercle count differing — only one row of tubercles on each opposing margin. 3 1941] Hobbs: A New Cray fish from San Luis Potosi, Mexico Text-figure 1. Cambarus blandingii cuevachicae, new subspecies. Pubescence has been removed from all structures illustrated. Figures not otherwise indicated are of C. blandingii cuevachicae. A, mesial view of first pleopod of C. blandingii acutus from Louisiana; B, mesial view of first pleopod of holotype; C, right chela of holotype; D, lateral view of first pleopod of holotype; E, lateral view of first pleopod of C. blandingii acutus from Louisiana; F, mesial view of first pleopod of male, form II; G, annulus ventralis of allotype; H, antennal scale of male paratype ; I, epistome of male paratype; J, ischiopodites of third and fourth pereiopods of first form male paratype; K, lateral view of carapace of holotype; L» dorsal view of carapace of holotype; M, lateral view of first pleopod of male, form II, Zoologica: New York Zoological Society Annulus ventralis subovate, elongate along transverse axis. Fossa projects beneath median dextral wall; sinus curving sinistrad slightly sinistrad of midventral line where it turns caudad to cut the caudal margin of the annulus. A shallow longitudinal furrow is present sinis- trad of the midventral line across the face of the annulus. Sternum just cephalad of an- nulus bearing a low tubercle on either side of the midventral line. Sternum just caudad of annulus modified into a low flat semioval struc- ture. Measurements. Holotype: carapace, height 1.71, width 1.86, length 3.79 cm.; areola, width .02, length 1.38 cm.; rostrum, width at base 1.59, length 1.89 cm.; abdomen, length 3.6 cm.; right chela, length of inner margin of palm 1.42, width of palm 1.14, length of outer margin of hand 4.23, length of movable finger 2.53 cm. Female Allotype: carapace, height 1.87, width 1.85, length 3.83 cm.; areola, width .07, length 1.36 cm.; rostrum, width at base .64, length .89 cm.; abdomen, length 3.72 cm.; right chela, length of inner margin of palm .92, width of palm 1.00, length of outer margin of hand 2.86, length of movable finger 1.85 cm. Type Locality. La Cueva Chica, a limestone cave, about one mile northeast of Pujal, San Luis Potosi, Mexico. These specimens were collected above the first waterfall. “Tumbled rocks alternated with shallow, narrow puddles. The going was difficult, but in the second puddle we forgot all about that, for Bishop spotted another crayfish. This time there was no es- cape, for the pool was scarcely a yard wide and twice as long. A couple of grabs and he had the creature. [XXVI: 1 “It was not blind. Lighter in color than the normal crayfish of the outside waters, it was fully eyed and of the common local species. That was a disappointment, but it went into the pickling jar anyway.”2 The temperature of the water was 80° F. The male holotype and female allotype and a second form male paratype are deposited in the United States National Museum. Of the paratypes, a male, form I, one male, form II, and a female are in my personal collection; one male, form I, seven males, form II, ten females, four immature males, and six immature females are in the Peabody Museum, New Haven, Con- necticut. Relationships. Cambarus blandingii cueva- chicae is most closely allied to C. blandingii acutus. Specimens collected one mile north of Villa Juarez, Tamaulipas, Mexico, seem to be somewhat intermediate between these two, par- ticularly in respect to the width of the areola. Remarks. I have been unable to discover any peculiarities in this race that seem to be associated with cave life. Bridges has pointed out that these specimens were of lighter color than the “normal crayfish of the outside water.” I have seen only preserved specimens so I can- not attest this observation. On the hairy parts of the ventral surface of these crayfish were found numerous ostracods, Entocythere cambaria Marshall, and many branchiobdellid worms, Cambarincola macro- donta Ellis.3 2 Bull. N. Y. Zool. Soc. 43 (3): 84-85. May-June, 1940. 3 Dr. Clarence R. Goodnight kindly identified this bran- chiobdellid for me. 1941] LaMonte: A New Corydoras from Brazil 5 2. A New Corydoras from Brazil. F. R. LaMonte American Museum of Natural History. Through the courtesy of the New York Aquari- um, we have secured three living specimens (approximately 15 mm.-18 mm. standard length), and seven preserved specimens (13 mm.-19 mm. standard), of a Brazilian Corydoras apparently hitherto undescribed. These fishes formed part of a collection brought to New York by Mr. Auguste Rabaut, and col- lected by him toward the end of December, 1940, in a tributary of the Amazon. The collector never heard a name given the stream, but he reached it after seven days by river boat west from Manaos, and four more days’ journey north toward the Colombian border. The country was flooded swamp, with heavy vegetation. Mr. Rabaut reports that the water was extremely acid. The specimens appear to represent two color patterns of the same species, the change in pattern probably occurring with growth. Each pattern, in certain lights, shows traces of the other. Of the three living fishes, the two larger are the “striped pattern;” of the preserved speci- mens, six are the “half and half” pattern, and the 19 mm. specimen is striped. Corydoras rabauti, new species. Description of Type. Number 15644, American Museum of Natural History, collected Decem- ber, 1940, in a tributary of the Amazon River, by A. Rabaut. Differences in body proportions of the 19 mm. striped specimen are noted in parentheses. Length to base of caudal, 15 mm. A rather chunky fish, with a short, rounded snout; the bony interorbital contained twice in the snout. There is no black hastate spot at the base of caudal, and the width of the naked area on the breast between the coracoids is equal to the diameter of the eye. The suborbital does not cover the entire cheek. The coracoids do not completely surround the pectoral bases on the surface of the body. The dorsal does not reach the adipose when laid back. The depth of the suborbital is equal to the diameter of the eye; the dorsal spine equal to the length of the snout. The flesh of the abdomen is smooth, without platelets. Snout, 2.1 in the head, its profile convex, sloping gently; interorbital, 2 (1.8); eye, 3.1. Body depth in standard length, 3 (2.7); width, 3.4 (3.1); head, 2.7 (3); snout to origin of dorsal, 1.8(2). Depth of caudal peduncle in body depth, 1.4. Lateral scutes spinous, 18-20/17-20 (20/21). Lateral line pores are visible on the first two or three scutes only. The abdomen is not granular and there are no platelets. Fontanel elongate, oval, reaching two-thirds the way to the anterior margin of eye and, pos- teriorly, to the supra-occipital process. Occipital process very narrowly triangular, not reaching azygous predorsal scute. Rictal barbels just reaching the gill-opening. Dorsal I, 8; the spine 1.8 (1.5) in the head. The fin is rounded, the first few rays longer than the spine. Distance of the dorsal fin from the adipose, 1.3 in the dorsal base. Anal I, 6, reach- ing past origin of caudal. Pectoral passing origin of ventrals to about midway in their length; ventrals barely reaching anal. Caudal widely forked, tip of lower lobe slightly rounded. Color in Life. Head region to short distance in front of dorsal, pinkish-orange. A dull, dark blue streak along anterior margin of eye socket, a short, very narrowly triangular metallic blue streak along posterior margin of pre-opercle. All fins except caudal, which is hyaline, are light orange, the dorsals and ventrals darkest. An ink-blue area, very slightly lighter ventrally, occupies the posterior two-thirds of the body, extending, with a concave anterior marginal outline, from in front of the dorsal origin to in front of the ventrals, entirely around the body, and back to the caudal base. On the caudal peduncle, the dark blue area becomes a short, thin line, margined by two square patches of metallic light blue. The eye is blue. In pre- servative, the lighter fins are hyaline; the dorsal and ventrals brown; the pinkish-orange body areas are faintly pink, buff or dirty light brown according to the original intensity of color; the lighter blue areas disappear as do the metallic patches; the ink-blue areas are darker and more dense. In the 19 mm. striped pattern fish (A. M. N. H. Number 15645), the ink-blue area is confined 6 [XXVI: 2 Zoologica: New York Zoological Society chiefly to a band, curving upward from the upper half of the caudal peduncle, following the midline of the body, but above it, and rising to a short distance in front of the dorsal origin. There are traces of bluish below this stripe on the entire posterior half of the body. The rest of the body is pinkish-orange, darkest on the head. The metallic light blue patches are as in the other fish. There is a trace of bluish on the middle rays of the caudal, not extending to the margin. In preservative, the body of this fish is creamy, slightly darker on the head, and the ink-blue areas are darker than in life. The species is very near eques (Steindachner, 1877, Sitz. K. Akad. Wien, volume 74, Abt. 1, p. 140. fig.), from which it differs in: a wider interorbital; a wider naked area on breast; a narrower suborbital which does not cover the entire cheek; a larger eye; longer barbels, and an apparently different color pattern. 1941] Crandall: Notes on Plumage Changes in the Bald Eagle 7 3. Notes on Plumage Changes in the Bald Eagle. Lee S. Crandall Curator of Birds, New York Zoological Park. (Plates I-IV). In August of 1930 a young Southern Bald Eagle ( Haliseetus leucocephalus leucocephalus (Linnaeus)), just out of the nest and barely able to fly, was captured on Little St. Simon’s Island, Georgia, by Philip Berolzheimer, of New York. In December of the same year, the young bird was brought north by Albert C. Benninger, of Brooklyn. It was kept in Forest Park, Queens, New York City, until June, 1931, when it was lent to the Girl Scout camp at Iona Island, Rock- land County, New York. On September 2, 1931, the bird was presented to the New York Zoo- logical Park and became the subject of the fol- lowing notes. In past years, we had made several attempts to check the progressive plumages of young Bald Eagles but always some catastrophe over- took the bird before the investigation had gone very far. Consequently, on the arrival of this specimen, no notes were made. However, in the autumn of 1933, we were able to plan indoor winter accommodation for the Bald Eagles and were so confident that greater longevity could be obtained that notes on our young individual were made. On October 10, 1933, the bird being three years old and in its fourth year, the body plumage was blackish-brown in general, the breast and lower parts much mixed with white. Forehead and throat were streaked with white, and the tail (Fig. 1) was heavily blotched with gray but with no pure white. The iris was dark brown and the beak greenish-black. On January 29, 1934, it was noted that the white streaks on the forehead had spread over the entire crown. The throat had become nearly clear white and a band of white extended to the nape. Between this band and the mottled area of the crown, was a parallel dark band, extending from the lores to the nape and passing through the eyes. During the following summer, this pattern was entirely lost. On September 2, 1934, the head, neck and throat were grayish-white, clearest on the throat but elsewhere blotched with blackish-brown, more heavily on the hind neck. The tail (Fig. 2) was extensively blotched with gray but still without clear white. The iris was light brown and the beak had become light olive, with a hint of yellow. The voice, too, was changing, for while the bird still began his call with the harsh notes of immaturity, he ended with the clear challenge of the adult. It was during the autumn of 1935, when the bird was five years old and in its sixth year, that it assumed what might be considered its first adult plumage. On November 1 of that year, the head and neck were recorded as pure white. The tail (Fig. 3), also, was now predominately white, although still heavily blotched with blackish-brown. The iris had acquired the pale yellowish-white of the adult and the bill, too, had become clear, pale yellow. The harsh notes had disappeared from the voice, which was now entirely clear. The tail grown during the summer of 1936 (Fig. 4) was more clearly white but still showed a considerable flecking of blackish-brown, par- ticularly on the shafts. On November 1, 1936, it was noted that while upper and lower tail coverts were mainly white, both were heavily stippled with brown. In November of 1937, the upper and lower tail coverts were recorded as entirely white. The tail of this year (Fig 5) showed a considerable reduction in the extent of dark stippling, though this was still present. In 1938 (Fig. 6), 1939 (Fig. 7) and 1940 (Fig. 8), the markings on the rectrices became progressively reduced. But it was not until 1940, when the bird was ten years old and in its eleventh year, that all stippling disappeared from the vanes. Even then, as may be seen in the photograph of the tail in situ (Fig. 8), faint dark markings were still present on the shafts of some feathers. It seems probable that there may be a con- siderable individual variation in the extent of dark markings on the shafts of the rectrices. A Bald Eagle received at the Zoological Park on March 1, 1927, had the head and tail “white.” Examination of this bird’s tail, after its death on June 17, 1940, when it must have been at least eighteen years old, showed heavy streaking of the shafts. [XXVI: 3 S Zoologica: New York Zoological Society Examination of the dates on which feathers were cast seems to indicate no definite procedure, except that in the last four years the outer pair dropped first and simultaneously or almost so. The earliest date for the casting of the first rectrice was April 19 (1934); the latest, June 21 (1934 and 1936). The earliest date for the casting of the last rectrice was July 30 (1934); the latest, September 25 (1938). Dates assigned to the mounted feathers indicate the time at which they were cast; in each case, they were grown in the previous year. Fig. 8 shows the tail of the living bird on December 12, 1940. The tail was grown in 1940. Summary. A Southern Bald Eagle ( Haliseetus leuco- ccphalus leucocephalus (Linnaeus)), hatched in 1930, was kept in the New York Zoological Park. In the autumn of 1935, when the bird was five years old and in its sixth year, the head and neck became pure white and the tail nearly so. It was not until 1940, when the bird was ten years old and in its eleventh year, that blackish stippling disappeared from the vanes of the rectrices. Acknowledgement. The dating and mounting of the feathers shown in the accompanying photographs were done by Herbert D. Atkin, who was a Keeper in the Department of Birds of the New York Zoological Park from 1905 to 1940, when he retired on pension. His painstaking cooperation made possible their presentation here. EXPLANATION OF THE PLATES. Plate I. Fig. 1. Haliceetus l. leucocephalus. 1933 tail feathers (bird 3 years old; in 4th year) molted in 1934. Fig. 2. 1934 tail feathers (bird 4 years old; in 5th year) molted in 1935. Plate II. Fig. 3. 1935 tail feathers (bird 5 years old; in 6th year) molted in 1936. Fig. 4. 1936 tail feathers (bird 6 years old; in 7th year) molted in 1937. Plate III. Fig. 5. 1937 tail feathers (bird 7 years old; in 8th year) molted in 1938. Fig. 6. 1938 tail feathers (bird 8 years old; in 9th year) molted in 1939. Plate IV. Fig. 7. 1939 tail feathers (bird 9 years old; in 10th year) molted in 1940. Fig. 8. Tail feathers of the living bird (10 years old; in 11th year) photographed in situ on December 12, 1940. CRANDALL PLATE I FIG. 1. FIG. 2. NOTES ON PLUMAGE CHANGES IN THE BALD EAGLE CRANDALL. PLATE I I. FIG. 3. FIG. 4. NOTES ON PLUMAGE CHANGES IN THE BALD EAGLE. CRANDALL. PLATE I I I FIG. 5. NOTES ON PLUMAGE CHANGES IN THE BALD EAGLE. CRANDALL PLATE IV. 1 ) Dald AGLE. AIL EATHERS Dec. /Z,/?AO. FIG. 8. NOTES ON PLUMAGE CHANGES IN THE BALD EAGLE. 1941] Beebe: External Characters af Six Embryo Nurse Sharks 9 4. External Characters of Six Embryo Nurse Sharks, Ginglymostoma cirratum (Gmelin).1 William Beebe (Plates I & 11; Text-figures 1-4). Six embryo nurse sharks, Ginglymostoma cirratum (Gmelin), taken from a single female, have been kindly loaned to me for the study of external characters, by Commodore William K. Vanderbilt. They are borrowed from his marine museum at Huntington, Long Island, and were obtained from Mr. Louis L. Mowbray. The parent was said to be about five feet in length, and was caught in Bermuda. It was not possible to stain or section the specimens, as I agreed to return the embryos in their present condition. The material consists of six embryos which I have designated as A to F. They are nicely graduated as to size and degree of development, and in total length measure from 73 to 152 mm. They are free, in the sense of being without shells or egg-cases, but each is attached by a short, twisted, umbilical-like extent of tissue with a large, rounded or oval mass of yolk. In his recent monograph on Chlamydoselachus (Bashford Dean Memorial Volume Archaic Fishes, Article VII, American Museum, N. Y.), Dr. E. W. Gudger writes, “The tropical, shallow- water nurse shark, Ginglymostoma cirratum carries in each greatly dilated uterus as many as 21 huge, thick-shelled eggs.” As regards the subsequent history of these uterine eggs, Gudger on strong circumstantial evidence believes that this shark is ovoviviparous, and that when the young are pretty well developed, they break out of their shells, and these latter are cast out while the embryos are retained in the uteri during further development. A photograph of one of the unbroken egg-cases, 140 mm. long, is shown on page 560 of the Memorial Volume. The total lengths of the six embryos under consideration are as follows: Embryo A, 73 mm.; B, 104; C, 107; D, 125; E, 138, and F, 152 mm. The relationship by weight of embryo to yolk is from 5.7% to 33%. In embryo A the fin-folds are still sufficiently in evidence to confuse any fin base measure- ments. The first dorsal is joined to the second 1 Contribution No. 612, Department of Tropical Re- search, New York Zoological Society. Contribution from Bermuda Biological Station for Re- search, Inc. dorsal, and this to the caudal, by the continuation of the fin in a groove beneath the surface profile of the body. The same is true of the anal which extends forward in a subdermal groove to the vertical of the pel vies. The translucent dorsal fins in this smallest embryo show thirteen in- cipient rays in the first, and ten in the second. External gills are present in all but embryo F, the largest. There is a gradual absorption in length of these filaments in the first five speci- Embryo A, length 73 mm. Side view with yolk sac (X%). Dorsal view of head showing external gill filaments from spiracle and gill-slits. (X6). 10 Zoologica: New York Zoological Society [XXVI: 4 mens. This is evident in the relationship per- centage between gill filaments and total lengths, 20%, 9.6%, 8%, 4.5%, and 3%. The smallest embryo shows the interesting condition of four external gill filaments projecting from the right spiracle, the longest of which measures 1.7 mm. in length. None are visible in the left spiracle. In the same smallest nurse shark embryo the anterior or first gill-slit shows twenty-six long filaments and a cluster of short ones at the top of the slit. A progressively larger number spring from the next three slits, while the fifth, which is directly below the fourth, has only four filaments, these, however, being of greater length than any of the rest. This general relationship holds good for the larger embryos. The climax of numerical development of ex- ternal gill filaments occurs in D, the 125 mm. individual. The next larger, with a total length of 138 mm., has lost all the filaments from the left first and second slits, although they are present on the right side. In E, of 138 mm., this asymmetry is reversed, the right anterior slits being free. The filaments are absorbed from be- low upward, the last to disappear being close to the dorsal ends of the slits. Those from the fourth and fifth slits are the final ones to vanish. F, of 152 mm. length, shows no external evi- dences of filaments. In the two youngest em- bryos the fifth slit is directly beneath the fourth, but from here on to the adult shark, it is slightly behind the fourth. As to color and pattern, the 73 mm. embryo is unspotted, as are full-grown sharks. Spots first appear, although weak and sparse, in the 104 mm. individual, and increase in strength and number up at least to embryos of 152 mm. length. In 300 mm., free-swimming nurse sharks the spots are small or obscured by the general dark pigmen- tation. A 385 mm. specimen taken on the Zaca expedition, is immaculate. Detailed examination shows that the pattern of spots can be quite asymmetrical (specimen C, with seven spots on the right side of the body, and fifteen on the left side); or symmetrical (specimens C, E and F, with numerous transverse bands on the head and body), or somewhat irregular (specimen G). Coppery color is ap- parent in the irides of the 152 mm. individual. The smallest and largest of seven specimens (including the six embryos, and a 385 mm. free- swimming young shark), are females. Claspers are evident in all the rest. In the 104 mm. embryo they are 4.5 mm. long, wholly attached 1941] Beebe: External Characters of Six Embryo Nurse Sharks 11 to the inner edge of the pelvic fins, and extending to within 1 mm. of the posterior rim of these fins. There is a gradual increase in size until in the 385 mm. shark the claspers are 13 mm. long, 9 mm. below the posterior fin border, and free for 6 mm. of their length. As the cirri or tentacles develop, their tips curve around and down, into the corners of the mouth, inside the lateral flap. This condition also holds in the young, free-swimming sharks, the tentacles being thus partly hidden and pro- tected when not in use. In the course of embryonic development of Ginglymostoma from an embryo of 73 mm. to a full-grown shark of 2,090 mm. total length, we find there are six characters showing a percentage of increase relative to the total length, and seven- teen which are on the minus side. The body depth is somewhat greater in the adult, the general flattening being especially noticeable in unborn individuals. The head, both in length and width, snout, preoral, interorbital and in- ternarial, all show relative reduction in the adult, as much as 3.7% in head length, and 4.4% in the snout. The mouth width, on the contrary, is slightly greater. The eye, length of tentacle and of spiracle are markedly less in the full grown shark. The heights of the three median fins, first and second dorsals and anal enlarge steadily from early em- bryo onwards, even reaching a 4.6% increase. The pectorals exceed these, registering width and length increases of 5.7% and 7.6%. There is considerable reduction in antero- posterior measurements from snout to first and second dorsals, as well as to pectorals, pelvics and anal fins. The bases of all the median fins are relatively less in the adult, partly because of the slow disappearance of the fin-fold. 12 Zoologica: New York Zoological Society [XXVI: 4 Text-figure 4. Embryo F, length 152 mm. Right and left side views with yolk sac. We may safely deduce that throughout the life of this shark there is an increasing need for large pectorals and a general flattening of the body in relation to the bottom-living habits, while correlated with this is a reduction of the head, with its tentacles, eyes and spiracles, and a relative shortening of the whole body. In fact, Ginglymostoma tends in development rather to- ward a bottom-living, ray-like type, than to a mid-water, shark-like organism. Developmental characters showing a plus per- centage in relation to total length are these : depth 2%, pectoral length 7.6%, pectoral width 5.7%, first dorsal height 4.6%, second dorsal height 2.3%, anal height 1.3% and mouth width 1%. Characters showing minus percentage in develop- ment are head 3.7%, head width 1.7%, snout 4.4%, interorbital 4.6%, eye 2%, snout to first dorsal 4.6%, snout to second dorsal 3%, snout to pectoral 4.6%, snout to pelvics 5.2%, snout to anal 3.6%, first dorsal base 2.1%, second dorsal base 2.3%, anal base 3.1%, snout to mouth 2.6%, tentacle length 2.5%, internarial 1.3% and spiracle length .5%. EXPLANATION OF THE PLATES. Plate I. Fig. 1. Embryo B, length 101 mm. Shark, showing external gill filaments and yolk sac. Fig. 2. Embryo D, length 125 mm. Shark, showing right side and yolk sac. Plate II. Fig. 3. Embryo F, length 152 mm. Dorsal and right side views and yolk sac. Fig. 4. Embryo F, length 152 mm. Ventral view and yolk sac. BEEBE PLATE I FIG 1. FIG 2 EXTERNAL CHARACTERS OF SIX EMBRYO NURSE SHARKS, GINGLYMOSTOMA CIRRATUM (GMELIN). BEEBE PLATE I I. FIG. 3. FIG. 4 EXTERNAL CHARACTERS OF SIX EMBRYO NURSE SHARKS. GINGLYMOSTOMA CIRRATUM (GMELIN). 1941] Smith: A Papillomatous Disease of the Gallbladder 13 5. A Papillomatous Disease of the Gallbladder Associated with Infection by Flukes, Occurring in the Marine Turtle, Chelonia my das (Linnaeus).1 G. M. Smith Department of Anatomy, Yale School of Medicine, and the New York Aquarium & C. W. Coates & R. F. Nigrelli New York Aquarium (Plates I— IV ; Text-figure 1). In two earlier publications attention was called to certain cutaneous tumors occurring in the turtle Chelonia mydas (Linnaeus) (Smith & Coates, 1938, 1939). These growths, fibro- epithelial in character, are found distributed on the neck and in the axillary, inguinal, eyelid, conjunctival and corneal regions. As a rule the tumors have the morphological characteristics of benign growths ; rarely does the structure suggest a malignant change. In the majority of such tumors ova of a blood fluke can be demonstrated on microscopical study, giving rise to the hy- pothesis that blood flukes or ova may act as causative factors in the production of these lesions. In earlier studies ova occurring in cu- taneous tumors were identified as those coming from Hapalotrema constrictum (Leared). Recent studies of a number of these worms (Nigrelli, 1940), however, indicate that this blood fluke is the same as Distomum constrictum of Leared (1862) and the form described by Price (1934) as Learedius learedius. The form described as Ha- palotrema constrictum (Leared) Looss, 1899 ( = Mesogonimus constrictum Monticelli, 1896) is now designated as Hapalotrema mistroides (Monticelli). This form is also a blood fluke but found in an- other species of marine turtle Caretta caretta (Lin- naeus) (= Thalassochelys caretta (Linnaeus)). It must lie pointed out here, however, that the original identification was tentative and based on the shape and size of the eggs, which are strikingly similar. Relationships and identifications of worms have been based on egg characteristics by various taxonomists. Further studies of lesions in turtles as a result of fluke infections, carried on during the past year, have revealed a new form of papillomatous disease of the gallbladder of Chelonia mydas 1 Aided by a grant from the Marine Cancer Fund of Yale University School of Medicine. which may be the result of a fluke infection of this organ. A brief description of the gallbladder lesions and of the fluke itself follows. Infected gallbladders show a wide range of lesions, depending doubtless upon the intensity and duration of the infection and the suscepti- bility of the tissues of the host. Any part of the fundus of the gallbladder may show thickened papillomatous change in solitary patches (Plate I, Fig. 1; Plate II, Fig. 4) or in confluent irregular masses (Plate I, Fig. 2). A papillomatous hyper- plasia of the mucous membrane near the cystic duct’s entrance into the gallbladder may be the most conspicuous lesion, the result of infection by trematodes (Plate II, Figs. 4, 5,6). Inconsequence, the cystic end of the gallbladder may appear greatly thickened and encroaching on the lumen of the gallbladder (Plate IV, Fig. 12) almost to the point of producing a stenosis or obstruction. At times the entire wall of the gallbladder partici- pates in the lesion. In these circumstances the epithelium is thrown up into papillomatous folds (Plate III, Figs. 7, 8), the muscularis shows in- creased amounts of connective tissue and lymph- oid cell infiltration and the subserosa may be oedematous, thickened, and contain a large number of dilated capillaries both vascular and lymphatic. All sorts of irregularly dilated glands are formed in the hyperplastic epithelium. Often glands penetrate down into the submucosa. However, malignant invasive changes have not been noted to date in approximately one hundred gallbladders examined which have shown the effects of fluke infection. Malignant change might readily be expected as a result of these chronic hyperplastic changes. In order to de- termine whether or not the papillomatous lesions do become malignant it will be necessary to have 14 [XXVI: 5 Zoological New York Zoological Society access to more extensive material than is avail- able at the present time. The majority of the flukes lie free in the dark green, thickened, often inspissated bile of the infected gallbladder. Flukes often are attached to the papillomatous parts of the mucous mem- brane (Plate IV, Fig. 11). The parasites may lie partly buried in the mucosa (Plate I, Figs. 1, 2). There can be little doubt that the presence of flukes at the site of characteristic lesions of the mucous membrane indicates that the underlying- cause of the disease arises in mechanical or chemical factors related to the activities of the infecting flukes. Ova surrounded by epithelioid or even giant cells may be found in the mucous membrane or other parts of the walls of the gall- bladder (Plate III, Fig. 9). Mucous production is abundant (Plate III, Fig. 10). The parasites from the gallbladder of Chelonia mydas have been identified as Rhytidodoides similis Price, 1939. A related species R. intesti- nalis was described in the same paper. According to Price, the identifications of “Both R. intesti- nalis and R. similis are in each case based on only two specimens; consequently it is not possible to determine the amount of variation within the species.” A collection of large numbers of the gallbladder form has afforded an opportunity to make just such a study for this species, at least (Text-figure 1). The worms are lanceolate in shape, tapering at both ends of the body and broadening consider- ably in the middle region. The posterior tip of the body can be extended into a minute “tail- like” appendage. When the flukes are examined in profile there is a characteristic hump on the dorsal side. In contracted specimens, the hump is more pronounced than in the extended forms. On the ventral surface, the acetabulum is promi- nent'and immediately anterior to this sucker may be seen the genital ridge and pore. The flukes are transparent and stained with bile. They measure 1.2 x .42-4.4 x 1.8 mm., with an average size of 2.2 X .97 mm. for fifty specimens. Cuticula without spines, smooth in expanded worms, with minute folds in contracted specimens. Oral sucker subterminal, measuring 140.2-342 microns in diameter (average, 293.4 microns), with lateral projections. The projec- tions are prominent in the smaller individuals, less so in larger forms but are nevertheless a constant feature. Acetabulum, 170.3-385.6 microns (average, 300.7 microns), pre-equatorial. Pre-pharynx lacking; pharynx 80-179 X 50-170 microns; esophagus of variable length, depending entirely on the state of contraction and expansion of the worms. Intestinal ceca simple, extending to the posterior extremity of the body. Excretory pore subterminal, dorsal; excretory vesicle Y- shaped, branching immediately behind the pos- terior testis, with the arms extending as far anterior as the pharynx. Genital aperture median, anterior to the ventral sucker and in the region of the cecal bifurcation. Cirrus pouch elongate piriform, muscular, situated over or slightly in front of the acetabulum, measuring 150-700 microns X 150-456 microns (average measurement 152 X 478.8 microns). Seminal vesicle and pars prostatica present. Testes in posterior third of the body, placed one behind the other or slightly oblique to the axis of the body, sometimes globular, sometimes sub-globular or ovoid in shape. Anterior testis usually smaller, measuring 102-420 microns X 250-456 microns; posterior testis 152-532 microns in diameter. In very small forms testes and vitellaria are poorly developed. Ovary globular, 91-280 microns in diameter (average, 250 microns), sub-median and pre-testicular in position. Ovarian complex, con- sisting of Mehlis’ gland, seminal receptacle, vital- line reservoir, and Laurer’s canal. Vitellaria consist of four groups of more or less elongated follicles; two small groups of follicles are in the anterior region of the body and present as two distinct masses, one on each side of the esophagus between the acetabulum and pharynx; the two larger groups extend from the acetabulum to the posterior extremity of the body. Main vitelline ducts pass medially into the ovarian complex at about the level of the posterior border of the ovary. Uterus fills the entire intercecal space between the ovary and acetabulum; metraterm present. Eggs yellowish, thick-shelled, embryo- nated, measuring on the average 36 X 72 microns. Comment. The histo-pathology of trematode infections among lower vertebrates is not well known. The condition reported above is in many respects not unlike liver fluke infections in mammals (see Craig & Faust, 1940). In sheep liver fluke, Fasciola hepatica Linnaeus, larvae are ingested with contaminated vegeta- tion. Excystment takes place in the duodenum where the parasites penetrate the wall and pass into the body cavity. They continue their migration, passing through the liver capsule directly into the liver substance. The mechanical irritation produced, in the case of heavier infec- tions, results in necrosis and fibrosis of the liver. Such an infection in man may cause cystic en- largement of the ducts, adenomata of the biliary epithelium, leucocytic and eosinophilic infiltra- tion, and eventually the development of scar tissue. Often, in their wanderings through the liver substance, the worms produce abscesses, and the mechanical damage causes atrophy of the liver and the portal vessels. Local lesions of this kind may be produced by other liver flukes such as the lancet fluke, Dicrocoelium dentriticum (Rudolphi), the cat liver fluke, Opisthorchis felineus (Rivolta), and the Chinese liver fluke, Clonorchis sinensis (Cobbold). In dicrocoeliiasis, sheep and other herbiverous mammals become infected by eating grass and other vegetation containing encysted larvae. When eaten, such larval forms excyst in the duodenum and find their way to the biliary passages, often producing hypertrophy of the 1941] Smith: A Papillomatous Disease of the Gallbladder 15 epithelium. Both opisthorchiasis and clonor- chiasis infections are brought about by eating fish poorly cooked, improperly cured, or raw, harboring the encysted stage. In clonorchiasis, the larvae, following excyst- ment, make their way up the common bile duct and migrate to the distal bile capillaries. Lesions produced by such infections have been described by Faust & Khaw (1927), Hoeppli (1933) and others. These include the proliferation of bili- ary epithelium, crypt formation in the bile duct, periepithelial fibrosis, periportal connective tissue hyperplasia and fibrous development around masses of eggs infiltrated in the liver substance. Although the papillomatous disease of the gallbladder of Chelonia mydas is stressed in the present contribution, there are also definite re- sponses in the liver and biliary ducts of these 16 [XXVI: 5 Zoologica: New York Zoological Society animals which will be discussed in a subsequent report. Summary. 1. Gallbladders of the marine turtle Chelonia mydas (Linnaeus) were found heavily infected with a fluke referred to as Rhytidodoides similis Price, 1939. 2. These flukes produce certain pathological changes in the tissues of the gallbladder. The outstanding feature of these lesions is a papillo- matous hyperplasia of the mucous membrane. 3. The parasites inducing these lesions are redescribed. References Cited. Craig, C. F. & Faust, E. C. 1940. Clinical Parasitology. Lea and Febiger, Phila. 772 pp., 244 text-figures. Faust, E. C. & Khaw, O. H. 1927. Studies on Clonorchis sinensis (Cobbold). Amer. Jour. Monogr. Series, (8), 339 pp. Hoeppli, R. 1933. Histological changes in the liver of sixty- six Chinese infected with Clonorchis sinen- sis. Chinese Medical Journal, JR': 1125- 1141. Leared, Arthur 1862. Description of a new parasite found in the heart of the edible turtle. Quart. J . Micr. Soc., London, n. s., 2: 168-170. Looss, A. 1899. Weitere Beitrage zur Kenntnis der Trema- toden-Fauna Aegyptens zugleich Versuch einer naturlichen Gliederung des Genus Distomum Retzius. Zool. Jahrb., Jena, Abt. f. Syst., 12: 521-784. Monticelli, F. S. 1896. Di un ematozoo della Thalassochelys caretta Linn. Internat. Monatschr. f. Anat. u. Physiol., Leipz. 13: 141-172. Nigrelli, R. F. 1940. Observations on Trematodes of the Marine Turtle, Chelonia mydas. Anat. Rec., 78, no. 4, Supph: 178. Price, Emmett W. 1934. New Genera and species of blood flukes from a marine turtle, with a key to the genera of the family Spirorchidae. J. Washington Acad. Sci., 24: 132-141. 1939. A new Genus and two new species of digenetic trematodes from a marine tur- tle. Proc. Helminth. Soc., Washington, 6: 24-25. Smith, G. M. & Coates, C. W. 1938. Fibro-epithelial growths of the skin in large marine turtles, Chelonia mydas (Linnaeus). Zoologica, 23: 93-98. 1939. The occurrence of trematode ova, Hapalo- trema consirictum (Leared), in fibro- epithelial tumors of the marine turtle, Chelonia mydas (Linnaeus). Zoologica, 24: 379-382. EXPLANATION OF THE PLATES. Plate I. Fig. 1. Strip of gallbladder mucous membrane showing to the right a small papilloma with several flukes attached to or partly buried in the mass. Fig. 2. Photograph of inside of gallbladder attached to the liver. Note confluent masses of papillomatous tissue; flukes lie attached or partly buried in the growth. Plate II. Fig. 3. Section of normal gallbladder containing bile. Fig. 4. Thickened, shrunken, infected gallbladder with several small papillomata near the entrance of the cystic duct. Figs. 5-6. Small contracted gallbladders infected by fluke, Rhytidodoides similis. The great thickening of the mucosa and submucosa has occurred near the cystic duct. Plate III. Figs. 7-8. Irregular hyperplasia of papillomatous regions of infected gallbladders. X 35. Fig. 9. Photo-micrograph of papillomatous region of gallbladder showing fragments of ova at points a and b. Epithelioid cells surround the small fragment at b. X 120. Fig. 10. Masses of mucous formed at the surface of papilloma. X 100. Plate IV. Fig. 11. Microscopic section of trematode Rhytido- doides attached to the hyperplastic epi- thelium of the gallbladder. Section of the parasite is at the level of the ventral sucker and the cirrus sac. X 40. Fig. 12. Photo-micrograph of the cystic end of infected gallbladder. The hyperplastic mucous membrane is thrown into countless folds which encroach upon the lumen of the gallbladder. Submucosa and muscularis are thickened. The subserosa seen in the upper part of the photograph is oedema- tous and contains numerous blood and lymphatic vessels. X 25. SMITH PLATE I. (Ill INI III! II 1! HI i mi in 1 IITI Nil mi mi i 2 3 4 1 i m ) o 7 FIG. 2. A PAPILLOMATOUS DISEASE OF THE GALLBLADDER ASSOCIATED WITH INFECTION BY FLUKES, OCCURRING IN THE MARINE TURTLE, CHELONIA MYDAS ( LINNAEUS ) . SMITH. PLAT FIG. 3. FIG. 4. FIG. 5. A PAPILLOMATOUS DISEASE OF THE GALLBLADDER ASSOCIATED WITH INFECTION BY FLUKES. OCCURRING IN THE MARINE TURTLE CHELONIA MYDAS ( LINNAEUS ) . SMITH PLATE III FIG. 7 FIG. 8. FIG. 9. FIG. 1 O . A PAPILLOMATOUS DISEASE OF THE GALLBLADDER ASSOCIATED WITH INFECTION BY FLUKES. OCCURRING IN THE MARINE TURTLE. CHELONIA MYDAS (LINNAEUSl SMITH PLATE IV. FIG. 1 1 . FIG 12. A PAPILLOMATOUS DISEASE OF THE GALLBLADDER ASSOCIATED WITH INFECTION BY FLUKES. OCCURRING IN THE MARINE TURTLE CHELONIA MYDAS ( LINNAEUS) . 1941] Treadwell: Polychaetous Annelids from Mexico 17 6. Eastern Pacific Expeditions of the New York Zoological Society. XXIII. Polychaetous Annelids from the West Coast of Mexico and Central America.1 Aaron L. Treadwell Department of Zoology, Vassar College (Text -figures 1-21). Contents. Introduction .17 Systematic Account. Family Amphinomidae Hermodice carunculata Kinberg 17 Notopygos ornata Grube 18 Chloeia euglochis Ehlers 18 Eurythoe, complanata Pallas 18 Eurythoe oculata sp. nov .18 Family Polynoidae Lepidasthenia picta Treadwell . 18 Lepidasthenia elegans sp. nov 19 Halosydna brevisetosa Kinberg 20 Family Sigalionidae Eupholoe nuda Treadwell 20 Polyodontes californicus sp. nov 20 Family Aphroditidae Aphrodita Linnaeus 21 Family Phyllodocidae Phyllodoce oculata Ehlers 21 Phyllodoce groenlandica Oersted . . . ..21 Eulalia magnapupula sp. nov. '. .21 Family Tomopteridae Tomoptcris opaca Treadwell . . 22 Family Typhloscolecidae Travisiopsis atlantica Treadwell 22 [This is the twenty-third of a series of papers dealing with the collections of the Eastern Pacific Expeditions of the New York Zoological Society made under the direction of Dr. William Beebe. The present paper is concerned with specimens taken on the Eastern Pacific Zaca (1937-1938) Expedition, which was made possible through the generosity of Mr. Templeton Crocker. For data on localities, dates, dredges, etc., refer to Zoologica, Vol. XXIII, No. 14, pp. 287-298.] Introduction. The following is a taxonomic account of 30 species of polychaetous annelids collected by Dr. William Beebe on the Eastern Pacific Zaca Ex- pedition in 1937-1938. Five new species are 1 Contribution No. 613, Department of Tropical Re- search, New York Zoological Society. Family Leodicidae Leodice longisetis Webster .22 Leodice paloloides Moore .22 Diopalra ornata Moore 22 Hyalinoecia juvenalis Moore .22 Arabella pacifica sp. nov 23 Family Amphictenidae Peclinaria gouldii Verrill 23 Family Opheliidae Ammotrypane bermudiensis Treadwell 23 Family Capitellidae Notomastus sp 23 Family Glyceridae Hemipodus mcxicanus Chamberlin 23 Family Chlorhaemidae Stylarioides sp 23 Family Terebellidae Terebella gorgonae Munro 23 Terebellides stroemi Sars 23 Family Sabellidae Sabella melanostigma Schmarda 23 Bibliography 24 included in the collection. They were collected from off Cedros Island, Lower California, on the north to Panama Bay on the south. The catalogue numbers all refer to specimens in the collections of the Department of Tropical Research of the New York Zoological Society. Systematic Account. Family Amphinomidae. Hermodice Kinberg. Hermodice carunculata Kinberg. Hermodice carunculata Kinberg, 1857, p. 13. Collected at Arriba Isthmus, Port Parker, Costa Rica, January 17, 1938, Cat. No. 3,882A. 18 Zoologica: New York Zoological Society [XXVI: 6 Notopygos Grube. Notopygos ornata Grube. Notopygos ornata Grube, 1856, p. 55. Collected at Station 203: D-9 (Port Parker, Costa Rica, January 22, 1938, 2 fathoms), Cat. No. 381,093. Chloeia Savigny. Chloeia euglochis Ehlers. Chloeia euglochis Ehlers, 1887, pp. 18-24; pi. 1, figs. 1, 2; pi. 2, figs. 1-8; pi. 3, figs 1M. Collected at Station 195: D-9 (Port Guatulco, Mexico, December 5, 1937, in 7 fathoms), Cat. No. 37,477. Collected at Port Guatulco, Mexico, December 5, 1937, in fish trap at 7 fathoms. Color plate Z-118, Cat. No. 37,429. Station 221 : D-l (Gulf of Chiriqui, Panama, March 13, 1938, 35 fathoms), Cat. No. 38,659. Specimen 37,477. Dorsal surface iridescent purplish-gray with a wide blackish band down the middle of each side within which are numer- ous oblique yellow streaks. On the mid-dorsal line of each segment are a pair of longitudinal, slightly oblique yellow lines. Tentacles and dor- sal cirri greenish-black; setae bright orange. Gills greenish-brown, with a central yellow band on each stem. Eurythoe Kinberg. Eurythoe complanata Pallas. Eurythoe complanata Pallas, Teste McIntosh. McIntosh, 1885 (as E. pacifica Kinberg), pp. 27, 28; pi. 2, figs. 3, 4; pi. 3, fig. 3; pi. 2A, fig. 13; pi. 3A, figs 5-9. Collected at Station 195: D-15 and D-16 (Port Guatulco, Mexico, December 6, 1937, in 1J4 fathoms), Cat. No. 37,526. Sihuatanejo, Mexico, November 24, 1937, in coral, Cat. No. 37,266. Station 196: D-18 (Tangola-Tangola Bay, Mexico, December 13, 1937, 30 fathoms), Cat. No. 37,743. Aba,jo Rocks, Port Parker, Costa Rica, January 22, 1938, tide-pool, Cat. No. 381,094. Eurythoe oculata sp. nov. (Text-figures 1-3). A single specimen. The body is broken and more or less macerated near its middle so that accurate measurements are impossible but it is approximately 180 mm. long. The greatest width, 10 mm., is at about the end of the anterior third of the body length. The prostomial width is 1 mm. The prostomium (Text-fig. 1), is ovate with its broader end anterior, and has two pairs of promi- nent eyes. The anterior tentacles are slender and reach as far as to the anterior prostomial border. The median tentacle is relatively long, is attached at the level of the posterior eyes and extends to the anterior border of the anterior ones. The tentacles are colorless, as is the prostomium. The caruncle is very small and inconspicuous and is attached to the dorsal border of the first somite (Text-fig. 1). This prostomium has certain re- semblances to that of E. duhia Horst as redefined by Munro (1933, p. 5), but is longer in proportion to width than in his figure 1, the eyes are much more prominent and the tentacles more slender. Horst (1912, p. 37), and Munro describe the caruncle as fiexuous though this is more clearly shown in Horst’s figure than in Munro’s. In oculata it is hardly more than a rounded papilla. The parapodial lobes are widely separated. The dorsal cirrus is prominent and lies anterior to the tuft of gills. Two dorsal aciculae with rounded ends protrude to the surface at the base of the cirrus and three similar aciculae occur in the neuropodium (Text-fig. 2). The ventral cir- rus is a rounded cone. The notosetae make up a dense tuft, are all alike, long, slender and sharp- pointed. Sometimes it is possible to see on the shafts the denticles figured by Horst ( loc . cit., pi. 10, fig. 7), but generally they seem smooth. I am uncertain what he meant by “dimples” on the shafts, and can find nothing that would answer to this description. The neurosetae (Text-fig. 3), are heavy, with apices bluntly rounded and an also bluntly rounded, subapical tooth. The gills first appear on the third somite and when well developed consist of a tuft of 12 or more filaments arising from a common base just behind the insertion of the dorsal cirrus. A few filaments may be simple but most are at least 2-branched and some are irregularly dichoto- mously divided into as many as 6 branches. The type was collected at the dock in Balboa, Panama, July 25, 1933. It is a female with eggs, and is No. 3,319 in the collection of the Depart- ment of Tropical Research of the New York Zoological Society. Family Polynoidae. Lepidasthenia Malmgren. Lepidasthenia picta Treadwell. Lepidasthenia picta Treadwell, 1928, pp. 456, 457; fig. 177; figs. 10-13. A single specimen in this collection makes possible additions and corrections to the original diagnosis. The drawing (loc. cit., fig. 10), shows a dividing line between the lateral prolongation of the prostomium and the cirrophore of the tentacle. This is incorrect. The cirrophore is a prolongation of the prostomium. In the type the palps and elytra were missing. In the Zaca material the former show as stout structures shorter than the median tentacle and having acuminate tips. The anterior elytra are approxi- mately circular in outline and cover the tentacles which may be seen through them. The anterior dorsal margin of the elytron is faintly tinged with black and there is a small black spot ventral to the point of elytrophore attachment. Later elytra are more lateral in position, leaving a con- 1941] Treadwell: Polychaeious Annelids from Mexico 19 Text-figures 1-21. 1-3. Eurythoe oculata sp. nov. 1, head X 7; 2, parapodiuin X 0; 3, neuroseta X 62. 4-7 . Lepidas- thenia elegans sp. nov. 4, head X 12; 5, parapodium X 20; 6, notoseta X 180; 7, neuroseta X 180. 8, Elytron of Eupholoe nuda Treadwell X 25. 9-12. Polyodontes calif ornicus sp. nov. 9, head X 18; 10, parapodium X 5; 11, seta X 85; 12, seta X 180. 13-17. Eulalia magnapupula sp. nov. 13, dorsal surface of headX 10; 14, ventral surface of head X 10; 15, parapodium X 33; 16, seta X 250; 17, seta X 250. 18-21. Arabella pacijica sp. nov. 18, head X 15; 19, seta X 68; 20, maxilla X 23; 21, mandible X 23. siderable portion of the dorsum exposed. While anterior elytra are lightly pigmented, throughout the median and posterior regions the pigmenta- tion is dense except for a colorless spot over the elytrophore, this giving an ocellated appearance to the dorsal surface. The elytra have neither marginal fringes nor surface spines. The body coloration is somewhat different from that of the type, indicating some degree of variability. Collected at Station 203: D-9 (Port Parker, Costa Rica, January 22, 1938, in 1^2 to 4 fath- oms), Cat. No. 381,092. Lepidasthenia elegans sp. nov. (Text-figures 4-7). Two specimens are in the collection, the type being incomplete. This is 25 mm. long and 10 mm. wide at the tenth somite. In the other, which is entire, the posterior region is badly pre- served and it is impossible to determine the number of somites or of the elytra. The speci- mens are assigned to this genus subject to later correction. The prostomium (Text-fig. 4), is 1 mm. in diameter, is broader than long and has a 20 Zoological New York Zoological Society [XXVI: 6 broad but not deep, anterior indentation into which fits the heavy cirrophore of the median tentacle. From the base of this indentation a dorsal groove runs posteriorly along the pros- tomial surface. The eyes are all near the pos- terior border, the posterior ones the smaller, nearer together and more or less covered by the margin of the first somite. The median cirro- phore is heavy, its length rather more than half that of the prostomium. The style is about as long as the prostomium up to the point where it begins to taper and it has a slender terminal filament. Cirrophores of the lateral tentacles are short and stout, their styles shaped much like those of the median but only about two-thirds as long. The ventral tentacular cirrus is similar in form to the median tentacle, the dorsal one similar to the lateral tentacle but much more slender. The palps are slender and inconspicu- ous, hardly longer than the median tentacle. The first somite, carrying the elytrophore, is twice as wide as the prostomium, later ones in- crease in width up to the sixth. The coloration is unlike in the two specimens, the type having five or six dark transverse bands on its dorsum in each somite, while the other has two irregular ones in this locality. Except for a slight tint near the apex of the palp there is no color in the head region. The parapodium has a much reduced notopo- dium into which a stout acicula extends, the neuropodium is much heavier and rounded at the apex, anterior and posterior lobes equal (Text- fig. 5). The dorsal cirrus is on a heavy cirrophore whose diameter is more than half that of the parapodium, its style reaching considerably be- yond the parapodium apex, slightly swollen near the end and terminating in a heavy filament. The ventral cirrus is very short, conical and has a slender tip. All but one of the notopodia examined had no setae, the exception having a single one. This was rather heavy, its apex bluntly rounded (Text-fig. 6), and it has trans- verse rows of teeth along one border, these being heavier than those in the neuropodium. There are about ten setae in the neuropodial group, these differing from one another mainly in size. They have rather heavy stalks which are swollen near the ends and then narrow to blunt points. Very small teeth, difficult to see unless the seta is in the proper position, occur along one margin of the swollen portion (Text-fig. 7). The protruded pharynx is as long as the first seven somites. Dorsally and ventrally at its apex are marginal rows of nine papillae. Anteriorly the elytra cover the dorsum. I am unable to say what the condition is in later somites. All elytra are circular in outline, the surface dotted with pigment spots and devoid of marginal filaments. The specimens are recorded as commensal in a holothurian. They were collected at Station 126: D-14 (East of Cedros Island, Mexico in 45 fathoms), Cat. No. 3,773. The type is in the collection of the Department of Tropical Re- search of the New York Zoological Society. Halosydna Kinberg. Halosydna brevisetosa Kinberg. Halosydna brevisetosa Kinberg, 1865, p. 85. Collected at Station 126: D-19 (S. E. of Cedros Island, Mexico, November 10, 1937, in 25 fath- oms), Cat. No. 3,794. Family Sigalionidae. Eupholoe McIntosh. Eupholoe nuda Treadwell. (Text-figure 8). Eupholoe nuda Treadwell, 1936, pp. 53, 54; figs. 10-14. As Eupholoe nuda I described a specimen from Bermuda in which I was uncertain whether all elytra had been lost or whether they had ever been present. The present collection contains one of this species with a full complement of elytra. They are decidedly lateral in position, leaving a considerable portion of the dorsum un- covered, and are essentially of the same form throughout the body. They, as well as the body surface, are covered with sand grains of which the largest on the elytra are those of the antero- dorsal area, some of which protrude beyond the elytral border (Text-fig. 8). The smallest of the grains are at the posterodorsal region where there are also some long marginal filaments. On the posteroventral face are two branched filaments and small ones occur on the remainder of the border except for a small part of the dorsal region where there are neither filaments nor sand grains. At this region the elytral surface is dotted with small spines. My earlier material showed two kinds of com- pound setae ( loc . cit., figs. 13, 14), one much heavier than the other, the slender one with a subterminal tooth. The present material shows that the subterminal tooth is present on both kinds of setae but evidently is easily broken away. The setae differ, therefore, only in size. Collected at Station 126: D-19 (S. E. of Cedros Island, Mexico, November 10, 1937, in 25 fathoms), Cat. No. 3,794. i Polyodontes californicus sp. nov. (Text-figures 9-12). The type and only specimen is a fragment, measuring 25 mm. to the region of the 20th ely- tron, and has a width of 10 mm. The prostomial diameter is about 1 mm., each half continued into an ommatophore, the base of each ommatophore being at about half way of the total prostomial- eye length, the apex of the eye being a trifle wider than the width of the stalk. There is a lens at the end. The sessile eyes are a pair of small dark spots, one on either side (Text-fig. 9). The base of the median tentacle is at the level of these sessile eyes and its stalk very slender, its apex narrowing to a filamentous tip which does not quite reach the apex of the ommatophore. At base of median tentacle is a ring of brownish pig- 1941] 21 Treadwell: Polychaetous Annelids from Mexico ment and a median caruncle-like ridge runs pos- teriorly across the dorsal prostomial surface. The lateral tentacles are similar to the median in form and size but are not visible from the dorsal sur- face, being hidden beneath the ommatophores. Each has a pigmented ring at the base and anoth- er just proximal to the filamentous tip. The palps are slender, nearly three times as long as prosto- mium and ommatophores combined and are pig- mented near the apices. The basal portion of the tentacular cirri extends as far as the base of the ommatophores, the styles slender, extending as far as the apex of the eyes, their apices abruptly ending in slender filaments. There is a band of brown pigment near the apex of the cirrophore and the dorsalmost cirrus has pigment near its apex and another pigmented mass near its base. The protruded pharynx extends 10 mm. be- yond the prostomium and at its apex is 7 mm. wide. The dorsalmost terminal papilla of the pharynx is 4 mm. long and on either side of it a row of 7 broad, flattened papillae with acute apices. The outermost one of each row is much the smallest and in most of them is a dark spot. The ventral median papilla is short, similar to, but a little larger than, the lateral ones of which there are 6 on either side. There are two light brown fangs with a single row of teeth on either side of each pair. The protrusion of the pharynx produces enough distortion so that it is not possible to determine exactly the position of the anterior elytra. Evi- dently the first pair covers the prostomium and is colorless except for a few widely distributed black spots. The second is much smaller than the first, the third larger than the second but still smaller than the first. These have a few dark spots. The fourth is about as large as the third and diffusely pigmented. Beginning with the fifth, a considerable area of the dorsal surface is uncovered by elytra and these assume a very characteristic coloration and form. Each is ap- proximately oval in outline, its long axis vertical and the outer end wider than the inner. Near the dorsal end is a densely pigmented band, separated by a colorless band from the elytron margin. From one-half to one-third of the ely- tron surface shows more or less pigmentation, the densest portion being the dorsal area mentioned above. The combined effect is that of two dark brown bands along the dorso-Iateral surface of the body. A very faint brownish tint, due to numerous small transverse lines, marks the dorsal surface beneath the elytra. The remainder of the body is colorless. The first pair of ventral cirri are very large, following ones successively smaller, most reaching only to end of parapodium. The dorsal cirri are nearly of the same size throughout. A median parapodium has nearly equal an- terior and posterior setal lobes and a spoon- shaped lip underlying the neuropod (Text-fig. 10). The dorsal cirrus has a heavy cirrophore and a conical style reaching a short distance beyond the setal lobe. The ventral cirrus is much more slender, its acute tip just about reaching the end of the neuropod. There are three types of setae. The first lie in the median portion of the row, have heavy brown shafts slightly enlarged and curved toward ends, apex obliquely truncated (Text-fig. 11). To this truncated portion is at- tached a slender terminal joint having fine hairs on its surface. Above and below these are slender setae, their apices very slender lanceolate in outline, the terminal portion covered by fine hairs. Another and larger tuft of colorless setae lies ventral to the heavy ones. These (Text-fig. 12) widen toward the ends and are curved. At the curve are toothed plates and beyond these, fine hairs cover the entire surface. The gills are bladder-like structures on the dorsal parapodial surfaces and are most promi- nent on the elytra-bearing somites. This species is here classed in the genus Polyodontes, because of the presence of gills. If, however, Chamberlin’s synonymy is correct (1919, p. 86), Polyodontes is preoccupied and it should be listed as Acoetes. The type was collected at Station 182: D-4 (Chamela Bay, Mexico, November 17, 1937, in 16 fathoms), Cat. No. 37,152, and is in the col- lection of the Department of Tropical Research of the New York Zoological Society. Family Aphroditidae. Aphrodita Linnaeus. Specimens of Aphrodita, all too badly pre- served to admit of species identification, were collected at Station 189: D-2 (17 miles S. E. by E. of Acapulco, Mexico, November 29, 1937, in 20 fathoms), Cat. No. 37,742; Station 195: D-17 (Port Guatulco, Mexico, December 7, 1937, in 6 fathoms), Cat. No. 37,741; Station 214: D-3 and 4 (14 miles S. by E. of Judas Point, Costa Rica, March 1, 1938, in 50 and 61 fathoms), Cat. No. 38,409. Family Piiyllodocidae. Phyllodoce Savigny. Phyllodoce oculata Ehlers. Phyllodoce oculata Ehlers, 1887, pp. 135-140; pi. 40, figs. 4-6. Collected at Station 220: L-2 (Isla Parida, Gulf of Chiriqui, Panama, March 12, 1938, surface light), Cat. No. 38,650; Station 126: D-19 (S. E. of Cedros Island, Mexico, November 10, 1937, in 25 fathoms), Cat. No. 3,794. Phyllodoce groenlandica Oersted. Phyllodoce groenlandica Oersted, 1842-43. Teste McIntosh. McIntosh, 1908-10, pp. 86-88; pi. 58, fig. 5; pi. 68, figs. 4-6; pi. 78, fig. 7. Collected at Station 223: L-l (Bahia Hermosa, Coiba Island, Panama, March 19, 1938, at sur- face light), Cat. No. 38,760. Eulalia magnapupula sp. nov. (Text-ligures 13-17). Characterized by very large eyes and short 22 Zoologica: New York Zoological Society [XXVI: (3 tentacles and cirri. Two specimens, of which the type is the larger but is incomplete. The smaller is 57 mm. long and in the anterior body region is 4 mm. in diameter to the ends of the setae. The posterior end is very narrow and possibly regenerating. In the preserved material the anterior margin of the prostomium is rounded, the lateral margins straight and running directly into the eyes (Text-fig. 13). The paired tentacles are heavy cones hardly longer than wide, only the anterior ones visible from the dorsal surface. The median tentacle is inconspicuous and is situated at about the middle of the dorsal prostomial surface. The eyes are very large, either one being larger than the exposed portion of the prostomium. Their lenses are visible only from the sides. The ten- tacular cirri are one pair on the first somite, two pairs on the second and one pair on the third (Text-fig. 14). The cirrophores of dorsal one on second and third pairs are heavy and about as long as the styles. The style of the third pair is the longest but this barely reaches the third somite. The eyes are dark brown, the peristomial surface is marked by brown pigment, cirrophores of tentacular cirri dark brown, styles colorless. Anteriorly the body color is light brown with a tendency toward a transverse dark band near the anterior margin of the somite and a darker spot on the anterior margin of the parapodium. This latter spot becomes more prominent posteriorly, and posterior to about the twenty-fifth somite is carried on a definite rounded lobe. The flattened dorsal cirri are more or less spotted with brown. Anteriorly on the ventral surface the pigmenta- tion is diffuse, while posteriorly there is a dark spot at the base of each parapodium. On the midventral portion of each somite is an uncol- ored spot shaped like a thickened H which is prominent against the diffuse brown of the general surface. Neither specimen retains the anal cirri. The parapodium is conical (Text-fig. 15), the posterior lip the longer and terminating in a fila- mentous tip, the acicula protruding from the surface by the side of this filament. Between the two lips arises a fan-shaped seta tuft. The dorsal cirrus is broad-lanceolate on a heavy cirrophore and reaches about as far as the end of the fila- ment on the setal lobe. The ventral cirrus is shorter than the setal lobe, heavy, with recurved tip. There are two kinds of setae, the first simple, bluntly rounded and slightly curved at the tip (Text-fig. 16), the other compound, with a long basal joint, the terminal joint slender and sharp-pointed (Text-fig. 17). In the type the pharynx is protruded to a dis- tance equal (in preserved material), to the length of the first eight somites. The type was collected at Station 225: T-l (11 miles S. W. by S. of Jicaron Island, Panama, March 20, 1938, in 500 fathoms), Cat. No. 38,767. Another specimen was taken at Station 227 : T-l (20 miles S. W. of Morro de Puercos, Panama, March 21, 1938, in 500 fathoms), Cat. No. 38,789. The type is No. 38,767 in the collection of the Department of Tropical Research of the New York Zoological Society. Family Tomopteridae. Tomopteris Eschscholtz. Tomopteris opaca Treadwell. Tomopteris opaca Treadwell, 1928, pp. 463, 464, fig. 178: fig. 29. Collected at Station 210: T-10 (20 miles south of Cape Blanco, Costa Rica, February 27, 1938, in 500 fathoms), Cat. No. 38,391. Station 233: T-l (55 miles south of Cape Corrientes, Colombia, April 3, 1938, in 500 fathoms), Cat. No. 38,908. Family Typhoscolecidae. Travisiopsis Levinsen. Travisiopsis atlantica Treadwell. Travisiopsis atlantica Treadwell, 1936, pp. 62, 63; figs. 30-33. Collected at Station 227 : T-l (20 miles S. W. of Morro de Puercos, Panama, March 21, 1938, in 500 fathoms), Cat. No. 38,789. Family Leodicidae. Leodice Savigny. Leodice longisetis Webster. Leodice ( Eunice ) longisetis Webster, 1884, pp. 317, 318; pi. 10, figs. 46-49. Collected at Sihuatenejo, Mexico, November 24, 1937, in coral, Cat. No. 37,266. Leodice paloloides Moore. Leodice paloloides Moore, (?) 1919, pp. 246-249; pi. 7, figs. 5-7. Collected at Sihuatanejo, Mexico, November 24, 1937, in coral, Cat. No. 37,266A. Diopatra Aud. et M. Ed. Diopatra ornata Moore. Diopatra ornata Moore, 1911, pp. 273-277; pi. 18, figs. 77-85. Collected at Station 196: D-1S (Tangola- Tangola Bay, Mexico, December 13, 1937, in 30 fathoms), Cat. No. 37,645A. Hyalinoecia Malmgren. Hyalinoecia juvenalis Moore. Hyalinoecia juvenalis Moore, 1911, pp. 277- 280; pi. 18, figs. 86-95. Collected at Station 203: D-2 (Port Parker, Costa Rica, January 20, 1938, in 10 fathoms), Cat. No. 38,109A. 1941] Treadwell: Polychaetous Annelids from Mexico 23 Arabella Grube. Arabella pacifica sp. nov. (Text-figures 18-21). The body is broken about at its middle but the entire animal is preserved, the total length being 160 mm. The body width is 3 mm., the pro- stomial width 1.3 mm. The prostomium is rounded (Text-fig. 18) and no eyes are visible. The anterior parapodia are very small, later ones increase in size but the posterior ones are small. There are two anal cirri, these being short with rounded ends. On the parapodia are prominent dorsal lobes extending to the ends of the setae. Setae are of only one kind (Text-fig. 19), having slender stalks widened and geniculate toward the extremities with very sharp apices and a pair of toothed wings at the bend. In profile only one of the wings is visible and unless slightly tilted the marginal denticulations do not appear. In each parapodium are three aciculae protruding only very slightly from the surface. All mouth parts jet black. In the maxilla the terminal plate on either side has only one tooth (Text-fig. 20). No. 2 has on either side a 6- toothed plate; No. 3 on either side 7 teeth. The forceps are heavy and overlie the proximal plates so that without destroying the specimen it was not possible to determine the precise number of teeth in each of the latter, but they extend as far as the forceps base and are toothed for the greater part of their length. Apparently also the inner margin of the forceps is toothed. The basal rods are very long, less than half of them being shown in the figure The mandible is rectangular with a forked base (Text-fig. 21). The type was collected at Sihuatanejo, Mexico, November 24, 1937, in coral, Cat. No. 37,266, and is in the collection of the Department of Tropical Research of the New York Zoological Society. Family Amphictenidae. Pectinaria Lamarck. Pectinaria gouldii Verrill. Pectinaria gouldii Verrill, 1873, p. 612; figs 87, 87A. Collected off Ballenas Bay, Gulf of Nicoya, Costa Rica, February 26, 1938, in mangrove mud, Cat. No. 38,359A. Family Opheliidae. Ammotrypane Rathke. Ammotrypane bermudiertsis Treadwell. Ammotrypane bermudiensis Treadwell, 1936, pp. 60, 61; figs. 24-26. Collected at Station 208: D-3 (Piedra Blanca Bay, Costa Rica, February 5, 1938, in 4 fathoms), Cat. No. 38,187A. Family Capitellidae. Notomastus Sars. A single specimen of undetermined species was collected at Station 126: D-16 (East of Cedros Island, Mexico, November 10, 1937, 42 fathoms), Cat. No. 3,781. Family Glyceridae. Hemipodus Quatrefages. Hemipodus mexicanus Chamberlin. Hemipodus mexicanus Chamberlin, 1919, pp. 349-350; pi. 63, figs. 2-3. Collected at Station 126: D-16 (East of Cedros Island, Mexico, November 10, 1937, in 42 fathoms), Cat. No. 3,781A. Family Chlorhaemidae. Stylarioides Della Chiaje. Stylarioides sp.? Collected at Station 221 : D-5 (Gulf of Chiriqui, Panama, March 13, 1938, in 35-40 fathoms), Cat. No. 38,659A. Family Terebellidae. Terebella Linnaeus. Terebella gorgonae Munro. Terebella gorgonae Munro, 1933 A, pp. 1070, 1071, fig. 18. Collected at Station 203: D-9 (Port Parker, Costa Rica, January 22, 1938, in 2 fathoms), Cat. No. 38,111. Terebellides Sars. Terebellides stroemi Sars. Terebellides stroemiSavs, 1835. Teste McIntosh. McIntosh, 1922, pp. 209-215; pi. 120, fig. 3; pi. 127, fig. 5. Collected at Station 126: D-16 (East of Cedros Island, Mexico, November 10, 1937, in 42 fath* oms), Cat. No. 3,781A. A single specimen of an empty tube of a tere- bellid, strongly resembling that of Lanice figured by McIntosh (1885, pi. 49, fig. 4), was collected at Station 224: D-3 (Hannibal Bank, Panama, March 20, 1938, in 35 fathoms), Cat. No. 38,755. Family Sabellidae. Sabella Linnaeus. Sabella melanostigma Schmarda. Sabella melanostigma Schmarda, 1861, p. 36. Elders, 1887, p. 263. Collected at Sihuatanejo, Mexico, March 24, 1937, in coral, Cat. No. 37,266A. 24 Zoologica: New York Zoological Society [XXVI : G Bibliography. Chamberlin, R. V. 1919. The Annelida Polychaeta. Memoir Mu- seum of Comparative Zoology, Harvard College, vol. 48, pp. 1-514; pis. i-80. Ehlers, Ernst 1864-68. Die Borstenwurmer, pp. 1-722; pis. 1- 24. 1887. Florida Anneliden. Memoir Museum of Comparative Zoology, Harvard College, vol. 15, pp. 1-335; pis. 1-60. Fauvel, Pierre 1916. Resultats des Compagnes Scientifiques ac- complies sur son yacht par Albert ler, Prince Souverain de Monaco. Monaco, Fasc. 48, pp. 1-152; pis. 1-9. Grube, A. E. 1856. Annulata Oerstediana. Vitensk. Meddel. fra d. Naturh. Foren. Kjobenh, pp. 44-62. Horst, R. 1912. Polychaeta errantia. Siboga Exp. Mono- graphs, vol. 24A, 43 pp.; 10 pis., Pt. 1, Amphinomidae. Kinberg, J. G. H. 1857. Ofversigt K. Vetensk. Akad. Fork., pp. 11- 14. 1865. Ofversigt K. Vetensk. Akad. Forh., pp. 167-180. McIntosh, W. C. 1885. Challenger Expedition Reports, vol. 12, pp. i-xxxvi; 1-554; pis. 1-55; 1A-39A. 1908-10. Monograph of the British Annelids. Ray Society Publication, vol. 2, Nephthy- didae to Ariciidae; pp. 1-524; pis. 43-87. Moore, J. Percy 1 909. Polychaetous annelids from Monterey Bay and San Diego, California. Proc. Acad. Nat. Science, Philadelphia, vol. 61, pp. 235-295; pis. 7-9. 1911. Polychaetous annelids dredged by the U. S. S. Albatross off the coast of Southern California in 1904. III. Euphrosynidae- Goniadidae. Proc. Acad. Nat. Science, Philadelphia, vol. 63, pp. 234-318; pis. 15-21. Munro, C. C. A. 1933. Polychaeta errantia collected by Dr. C. Crossland at Colon, in the Panama region and the Galapagos Is. Proc. Zool. Soc. London, pt. 1, pp. 1-96; text-figs. 1-36. 1933A. Polychaeta sedentaria collected by Dr. C. Crossland at Colon in the Panama region and the Galapagos Is. Proc. Zool. Soc. London, pt. 4, pp. 1039-1092; text-figs. 1-31. Schmarda, Ludwig 1861. Neue wirbellose Thiere. Bd. 1, Hft. 2, pp. 1-164; pis. 16-37. Treadwell, A. L. 1906. Polychaetous annelids of the Hawaiian Is. Bull. U. S. Fish Commission for 1903, pt. 2, pp. 1145-1181, 81 figs. 1928. Polychaetous Annelids from the Arcturus Oceanographic Expedition. Zoologica, vol. 8, no. 8, pp. 449-485, figs. 177-179. 1936. Polychaetous annelids from the vicinity of Nonsuch Id. Bermuda. Zoologica, vol. 21, pt. 1, pp. 49-68; 3 pis. 1937. Polychaetous Annelids from the West Coast of Lower California, the Gulf of California and Clarion Island. Zoologica, vol. 22, pt. 2, pp. 139-160, 2 pis. Verrill, A. E. 1873. Invertebrate Animals of Vineyard Sound. Report U. S. Fish Commission for 1872. Annelids, pp. 580-622; pis. 10-17. Webster, H. E. 1884. Annelida from Bermuda. Bull. U. S. National Museum, no. 25, pp. 305-327, 12 pis. 1941] Treadwell: Polychaetous Annelid* from Bermuda Plankton 7. Plankton of the Bermuda Oceanographic Expeditions. X. Polychaetous Annelids from Bermuda Plankton, with Eight Shore Species, and Four from Haiti.1 Aaron L. Treadwell Department of Zoology, Vassar College. (Text-figures 1-9). Con . 25 . 25 . 26 . 26 . 26 . 26 . 27 . 27 . 28 . 2S . 28 . 28 . 28 TENTS. Family Leodicidae Leodice stigmatura Verrill 28 Leodice culebra Treadwell 28 Leodice mutilata Webster 28 Family Syllidae Autolytus bidens sp. nov 28 Iiaplosyllis gula Treadwell 29 Family Amphictenidae Pectinaria gouldii Verrill 29 Family Opheliidae Ammotrypane bermudiensis Treadwell 29 Family Glyceridae Telake epipolasis Chamberlin 29 Family Terebellidae Eupolymnia magnified Webster 29 Family Sabellidae Dasychonopsis conspersa Elilers 29 Family Serpulidae Spirobranchus tricornis (Morch) Elders . 29 Bibliography 29 Introduction Systematic Account. Fandly Ampldnomidae Hermodice caruncuiata Kinberg . . Family Polynoidae Lepidonotus pilosus Treadwell . . . Family Sigalionidae Acanthicolcpis longicirrata sp. nov. Family Phyllodocidao Eulalia megalops Verrill My slides gracilis sp. nov Lopadorhynchus uncinatus Fauvel Family Alciopidae Vanadis fuscapunclata Treadwell . Fandly Tomopteridae Tomopteris longiselis Treadwell . . Fandly Typldoscolecidae Travisiopsis atlantica Treadwell . Fandly Nereidae Nereis bairdii Webster Nereis mirabilis Kinberg Nereis agassizi Ehlers [This is the tenth of a series of papers dealing with the plankton content of a series of nets drawn through a cylinder of water off the coast of Bermuda on the Bermuda Oceanographic Expeditions of the Department of Tropical Research under the direction of Dr. William Beebe. Full details as to this circle of water, and of the dates, depths, etc., of the nets will be found in Zoologica, Vol. XIII, Nos. 1, 2 and 3, pp. 1-45, and Zoologica, Vol. XXI, No. 3, pp. 69-73.] Introduction. The following is an account of 23 species of polychaetous annelids collected primarily at Bermuda on the Bermuda Oceanographic Ex- 1 Contribution No. 614, Department of Tropical Re- search, New York Zoological Society. Contribution from the Bermuda Biological Station for Research, Inc. peditions of the Department of Tropical Research of the New York Zoological Society, plus materi- als relating to specimens collected on the Haitian Expedition of the Department in 1927. Three new species are described, the types of which are deposited in the collections of the Department of Tropical Research. Systematic Account. Family Amphinomidae. Hermodice Kinberg. Hermodice caruncuiata Kinberg. Hermodice caruncuiata Kinberg, 1857, p. 13. Collected at Cooper’s Island, Bermuda, in mud, Sept. 11, 1931, Cat. No. 312,029. Nonsuch Island, Bermuda in shallow water, July 23, 1929, Cat. No. 29,135. [XXVI: 7 26 Zoologica: New York Zoological Society Family Polynoidae. Lepidonotus Leach. Lepidonotus pilosus Treadwell. Lepidonotus pilosus Treadwell, 1937, pp. 141- 143, pi. 1, figs. 1-7. Collected at Bizoton, Port-au-Prince Bajq Haiti, March 22, 1927. Family Sigalionidae. Acatithicolepis Norman. Acanthicolepis longicirrata sp. nov. (Text-figures 1-4). Characterized by the long tentacles and cirri and by the large eyes. As seen from above the pigmented borders of the two eyes on either side are in contact (Text-fig. 1) and appear to run together, forming a heavy pigmented border to the prostomium. The lens of the posterior eye points laterally, thus seeming to bend the pig- mented band toward the dorsal surface. The lens of the anterior eye is not visible from above. Length 15 mm., width of prostomium 1 mm., greatest body width to parapodial bases 4 mm. in region of 9th somite. At first gradually, later more rapidly, the body tapers to about 0.5 mm. at the posterior end. Only one elytron remains and the preservation is such that it is impossible to be certain of the elytrophore count in later somites but the number is apparently 18. The prostomium is somewhat broader than long, each half ending on its anterior margin in a well marked peak. The cirrophore of the median tentacle is heavy, set into a shallow depression between the two prostomial halves. The style of the median tentacle is about the length of the palps, these 3 mm. long, colorless and curved, the apices diverging. Cirrophores of the lateral tentacles small, styles slender, reaching to about the middle of the palps. Only one tentacular cirrus and one dorsal cirrus remain, these being about as long as the palps. Each cirrus has a central opaque white core, the remainder being- translucent. Apparently the elytra originally covered the dorsal surface but only one remains. This is so transparent as to be easily overlooked. On its surface are numerous, evenly distributed, low spines visible under a 40 diameter magnifica- tion. Its outline is approximately circular. In a parapodium taken from in front of the middle of the body (Text-fig. 2) the notopodium is shorter than the neuropodium and the latter when seen from above has an inflated appearance, its longitudinal diameter being greater than that of the notopodium. The notopodial acicula is shorter but slightly heavier than the neuropodial, the former being covered nearly to its apex by a symmetrical tissue outgrowth. A similar out- growth covers the neuroacicula but there is in addition a papilla lying parallel to the acicula end (Text-fig. 2). The ventral cirrus is slender and elongated, extending for nearly half its length beyond the apex of the setal lobe. The notosetae are heavier than the neurosetae and form a diverging tuft which extends almost vertically from the notopodium. They are of two kinds, one straight and sometimes very heavy, the others shorter and curved. Both have transverse rows of plates whose margins seem not to be particularly toothed. The neurosetae are more numerous and more slender. The dorsal ones in the tuft are slender, have very sharp points and alternately arranged rows of toothed plates (detail in Text-fig. 3). Ventral to these are fewer and larger ones which enlarge toward the ends and then narrow to acute apices. A row of toothed plates lies along one margin of the tapered portion (Text-fig. 4). The type was collected in Net 1,503, 600 fath- oms deep, off Bermuda, July 25, 1934. Others were taken in Net 1,316, 800 fathoms, off Ber- muda, Sept. 17, 1931, Cat. No. 312,135. It was also taken at Station 114: T-5, 500 fathoms, in the Hudson Gorge, 125 miles off New York City, July 8, 1928. The type is No. 3,465 in the col- lection of the Department of Tropical Research of the New York Zoological Society. Family* Phyllodocidae. Eulalia Savigny. Eulalia megalops Verrill. Eulalia megalops Verrill, 1900, p. 601. Collected in Castle Harbor, Bermuda, August 13, 1931, Cat. No. 311,247. Taken in a bivalve shell in coral. Mystides Theel. Mystides gracilis sp. nov. (Text-figures 5-7). A slender species represented only by an in- complete specimen. What remains is 62 mm. long and about 1.5 mm. in diameter in greatest width. The prostomium (Text-fig. 5) has the form of a cone with rounded apex, the sides nearly straight lines. The eyes are large and have prominent lenses. Behind the prostomial margin is a small tubercle. The tentacles are about one- third as long as the prostomium. There are four pairs of tentacular cirri arranged according to the generic formula. The third pair are the longest, reaching somite 7 ; the dorsal of the second pair reaches to somite 5. Except for faint transverse lines on the anterior dorsum, the body is colorless. In most of the anterior region the dorsal cirri have been lost but those that remain are very small, barely reaching to the end of the setal lobes. Posterior ones (Text-fig. 6) are much larger, are rectangular in form and carried on heavy bases. In the setal lobe there is a slightly bifid posterior lip and a rounded one coming to the surface between them. In each parapodium there are about ten setae having rather heavy basal joints, enlarged and spiny at the ends. The terminal joints are long and slender, curved, and toothed on the concave margins (Text-fig. 7). The ventral cirri are longer than the setal lobe, have ovate outlines and acute apices. 1941] Treadwell: Polychaelous Annelids from Bermuda Plankton 27 Text-figures 1-9. 1-4. Acanthicolepis longicirrata sp. nov. 1, head X 12; 2, parapodium X 23; 3, detail of stalk of neuroseta X 180; 4, neuroseta X 85. 5-7. Mysti- des gracilis sp. nov. 5, head X 15; 6, parapodium X 33; 7, seta X 250. 8, Nereis agassizi Ehlers. Parapodium X 23. 9, Autolytus bidens sp. nov. Head X 45. The type was collected in Net 1,179, off Bermuda at the surface, August 14, 1931, Cat. No. 311,288, and is in the collection of the Department of Tropical Research of the New York Zoological Society. Lopadorhynchus Grube. Lopadorhynchus uncinatus Fauvel. Lopadorhynchus uncinatus Fauvel, 1916, pp. 57-61, pi. 1, figs. 2-3; pi. 4, figs. 4-14. Collected at Bermuda in the following nets: Net 1,321, 50 fathoms, Sept. 18, 1931, Cat. No. 312,166, color, pinkish tan; Net 1,337, 600 fath- oms, Oct. 29, 1931, color orange. Family Alciopidae. Vanadis Claparede. Vanadis fuscapunctata Treadwell. Vanadis fuscapunctata Treadwell, 1906, pp. 1159-1160, figs. 29-31. Collected in Bermuda in the following: Net 10, 400 fathoms, April 9, 1929, Cat. No. 2932; Net 1,175, 600 fathoms, Aug. 14, 1931, Cat. No. 311,761, color transparent, eyes scarlet; Net 1,262, 700 fathoms, Sept. 4, 1931, Cat. No. 311,807; Net 1,264, 900 fathoms, Sept. 4, 1931, Cat. No. 311,821, color dirty tan in general, eyes coral red; Net 1,275, 1,000 fathoms, Sept. 7, 1931, Cat. No. 311,866, color white, eyes scarlet; Net 1,308, 100 fathoms, Sept. 16, 1931, Cat. No. 312,073, color olive, eyes red; Net 1,321, 50 fathoms, Sept. 18, 1931, Cat. No. 312,167, color lemon yellow, eyes coral red; Net 1,508, 100 fathoms, Aug. 14, 1934, Cat. No. 34,256. The following field-notes were made upon specimen No. 31,472 (Net 1,050, 25 fathoms, July 6, 1931), 23 mm. long: Color: Eyes coral red with gleaming, brighter centers which are at the base of a tiny projecting crystal dome. Otherwise the animal is entirely transparent except for a yellowish, reddish, brownish or blackish spot behind the base of each lateral appendage. These spots are lightest and brightest when the worm is active or when it has been exposed for some time to bright light. There is no noticeable change in their size, except possibly a slight contraction after the worm was placed in the darkness of a refrigerator. The eyes paled only slightly when the worm was kept in darkness. As the worm weakened a growing opaqueness was noticeable. Movements: Locomotion is principally by means of the short, three-leaved, shamrock-like paired appendages which, on alternate sides, are thrown abruptly forward and back so that the worm, seen from above, has a ridiculous sort of waddle. It seems to move either ahead or in reverse with equal facility and no apparent change of method. When greatly excited it shoots through the water with tremendous con- tortions, twistings and loopings of the entire body. The partitions between the segments, as trans- parent as the body wall, are pierced by holes which, more or less rhythmically, though not in unison, expand almost to full diameter of the segment and contract nearly to pinheads. The funnel-shaped proboscis is likewise contractile. This worm remained alive in the Bermuda laboratory for two days. It was kept in a qui- escent condition in the refrigerator most of the time, but was revived at intervals of a few hours in the warm air of the laboratory. About five to ten minutes were required for it to show signs of life. [XXVI: 7 Zoologica: New York Zoological Society 28 Family Tomopteeidae. Tomopteris Eschscholtz. Tomopteris longisetis Treadwell. Tomopteris longisetis Treadwell, 1936, pp. 58-59, figs. 18-21. Collected in Bermuda, Net 1,332, 600 fathoms, Oct. 28, 1931, Cat. No. 312,227, color white. Fragments and young of toinopterids were col- lected in other nets at Bermuda, but none were identifiable. Family' Typhoscolecidae. Travisiopsis Levinsen. Travisiopsis atlantica Treadwell. Travisiopsis atlantica Treadwell, 1936, pp. 62, 63, figs. 30-33. Collected in Bermuda in the following nets: Net 793, 700 fathoms, July 9, 1930, Cat. No. 30,477; Net 869, 1,000 fathoms, Sept. 10, 1930, Cat. No. 30,844, color orange yellow; Net 881, 600 fathoms, Sept. 12, 1930, Cat. No. 30,904, color orange yellow; Net 956, 1,000 fathoms, Sept. 28, 1930, Cat. No. 301,307; Net 1,258, 900 fathoms, Sept. 3, 1931, Cat. No. 311,778; Net 1,264, 900 fathoms, Sept. 4, 1931, Cat. No. 311,820, color transparent white. Family Nereidae. Nereis Cuvier. Nereis bairdii Webster. Nereis bairdii Webster, 1884, pp. 312-313, pi. 8, figs. 22-28. Collected at the surface at Bermuda as follows : Net 982, surface, May 19, 1931, Cat. No. 3,124; Net 1,179, surface, Aug. 14, 1931, Cat. No. 311,286; dip net at night-light, Aug. 27, 1937. Those from Net 982 are in the heteronereis stage. Nereis mirabilis Kinberg. Nereis mirabilis Kinberg 1865, p. 170. Collected in Bermuda as follows: Net 982, surface, May 19, 1931, Cat. No. 3,124A; Net 1,332, 600 fathoms, Oct. 28, 1931, Cat. No. 312,286. Nereis agassizi Ehlers. (Text-figure 8). Nereis agassizi Ehlers, 1868, pp. 542-546, pi. 23, fig. 1. ‘ A single specimen, doubtfully identified as this species. Its most characteristic features are the dark parapodial glands and the single heavy spine in each parapodium lying almost in contact with the notopodial acicula on its dorsal surface (Text-fig. 8) . This also conforms with reasonable accuracy to the brief description of N. kobiensis as given by McIntosh (1885, pp. 210-212, pi. 34, figs. 3-6; pi. 16a, figs. 2-4). I was unable to get a good view of the jaw apparatus which would have been confirmatory. Both of these species have some resemblances to N. dumerilii Aud. et M. Ed., but no diagnoses that I have seen of the latter species mention the heavy spine. Collected in Bermuda, Net 1,237, 700 fathoms, Aug. 29, 1931, Cat. No. 311,651, color orange- yellow. Family Leodicidae. Leodice Savigny. Leodice stigmatura Verrill. Leodice stigmatura Verrill, 1900, pp. 641-643. Collected in tidepool on Nonsuch Island, Bennuda, April 23, 1929, Cat. No. 2,945. Leodice culebra Treadwell. Leodice ( Eunice ) culebra Treadwell, 1901, p. 197, fig. 37. Dredged off Nonsuch Island, Bermuda, 2 fathoms, Nov. 4, 1931, Cat. No. 312,339. Leodice mutilata Webster. Leodice ( Eunice ) mutilata Webster, 1884, pp. 315-316, pi. 9, figs. 36-40. Collected in coral, Castle Harbor, Bermuda, May 1929. Family Syllidae. Autolytus Clrube. Autolytus bidens sp. nov. (Text-figure 9). In a bottle labeled “luminous worms, from Castle Harbor, Bermuda” were numerous frag- ments of an Autolytus that is evidently of a new species. No entire individuals wrere present and I have no information concerning absolute size or differences in body regions. The prostomium (Text-fig. 9), is broader than long, its posterior margin deeply indented and almost its entire dorsal surface taken up by the eyes of which the anterior pair is the larger. They are dark brown in color and very conspicuous. The median tentacle is from six to eight times as long as the prostomium, lateral tentacles much shorter and a little more slender. Tentacular cirri about as long as the median tentacle. The first few dorsal cirri are long, later ones shorter but all longer than body width. Width of prostomium about 2/3 mm. The pharynx extends through six somites, the brown oesophagus through two, the stomach through seven. In the type the first twenty-one somites have only compound setae while later ones have in addition a tuft of long and extremely slender simple ones. With the appearance of the longer setae the antero-postero diameter of the parapodia increases, giving them a swollen appearance. There was no indication of sex products. The compound setae are veiy small, the basal joint slightly widened and beveled at the end. 1941] Treadwell : Polychaetous Annelids from Bermuda Plankton 29 The terminal joint is extremely short and has a terminal tooth and a smaller one at about the middle of the concave surface. Collected at surface, Nonsuch Island, Bermu- da, Nov. 16, 1931. Type, No. 312,351, Depart- ment of Tropical Research, New York Zoological Society. Haplosyllis Langerhans. Haplosyllis gula Treadwell. Haplosyllis gala Treadwell, 1924, pp. 11, 12, figs. 19-23. Dredged in shallow water, Isle de Gonave, Haiti, June 1, 1927, Cat. No. 272. Family Amphictenidae. Pectinaria Lamarck. Pectinaria gouldii Verrill. Pectinaria gouldii Verrill, 1873, p. 612, figs. 87, 87a. Collected in Castle Harbor, Bermuda, Nov. 2, 1931, 2 fathoms, Cat. No. 312,342. Family Opheliidae. Ammotrypane Rathke. Ammotrypane bermudiensis Treadwell. Ammotrypane bermudiensis Treadwell, 1936, pp. 60, 61, figs. 24-26. Collected at surface off Nonsuch Island, Bermuda, May 12, 1931, Cat. No. 317. Speci- mens dredged at 2 fathoms, off Nonsuch Island, Bermuda, Aug. 25, 1931, Cat. No. 311,675, and Oct. 3, 1931, Cat. No. 312,234. Family Glyceiiidae. Telake Chamberlin. Telake epipolasis Chamberlin. Telake epipolasis Chamberlin, 1919, pp. 346- 348, pi. 63, figs. 4-8; pi. 64, fig. 1. Collected at Bermuda, Net 1,179, surface, Aug. 14, 1931, Cat. No. 311,287. Family Terebellidae. Eupolymtiia Verrill. Eupolymnia magnified Webster. Eupolymnia ( Terebella ) magnified Webster, 1884, p. 324, pi. 11, figs. 58-60. Collected in Bermuda as follows: Nonsuch Island, tidepools, Apr. 23, 1929, Cat. No. 2,945A; Nonsuch Island, tidepools, Aug. 22, 1929, Cat. No. 29,158; surface at night, near Nonsuch Island, Sept. 13, 1931, Cat. No. 311,996. Family Sabellidae. Dasychonopsis Bush. Dasychonopsis conspersa Ehlers. Dasychonopsis ( Dasychone ) conspersa Ehlers, 1887, pp. 226-270, pi. 54, figs 1-6. From bottom of old barge, Bizoton, Port-au- Prince Bay, Haiti, Cat. No. 27,212. Family Serpulidae. Spirobranchus Blainville. Spirobranchus tricornis (Morch) Ehlers. Spirobranchus tricornis Ehlers, 1887, pp. 292- 295; pi. 57, figs. 8-15. Collected at Lamentin Reef, Port-au-Prince Bay, Haiti, Apr. 26, 1927, Cat. Nos. 27,379, 27,385, color red with golden setae. Bibliography. Chamberlin, R. V. 1919. The Annelida Polychaeta. Memoir Mu- seum. Comparative Zoology, Harvard College, vol. 48, pp. 1-514, pis. 1-80. Ehlers, Ernst 1864-1868. Die Borstenwiirmer, pp. 1-722, pis. 1-24. 1887. Florida Anneliden. Memoir Museum Com- parative Zoology, Harvard College, vol. 15, pp. 1-335, pis. 1— 60. Fauvel, Pierre 1916. Residtats des Campagnes Scientifiques ac- complies sur son yacht par Albert ler, Prince Souverain de Monaco. Fasc. 48, pp. 1-152, pis. 1-9. Grube, A. E. 1856. Annulata Oerstediana. Vitensk. Meddel. fra d. Naturh. Foren. Kjobenh., pp. 44-62. Horst, R. 1912. Polychaeta errantia. Siboga Exp. Mono- graphs, vol. 24A, 43 pp., 10 pis., pt. 1, Amphinomidae. Kinberg, J. G. H. 1857. Ofversigt. K. Vetensk. Akad. Forh., pp. 11- 14. 1865. Ofversigt. K. Vetensk. Akad. Forh., pp. 167— 180. McIntosh, W. C. 1885. Challenger Expedition Reports, vol. 12, pp. i-xxxvi, 1-554, pis. 1-55, 1A-39A. 1908-1910. Monograph of the British Annelids, vol. 2, Ray Society Publication, Nephthy- didae to Ariciidae, pp. 1-524, pis. 43-87. 1922. Monograph of the British Annelids, vol. 4, Ray Society Publication, Hermellidae to Sabellidae, pp. 1-250, pis. 112-127. 30 Zoologica: New York Zoological Society [XXVI: 7 Moore, J. Percy 1909. Polychaetous annelids from Monterey Bay and San Diego, California. Proc. Acad. Nat. Sci., Philadelphia, vol. 61, pp. 235- 295, pis. 7-9. 1911. Polychaetous annelids dredged by U. S. S. Albatross off the coast of Southern Cali- fornia in 1904. Ill, Euphrosynidae- Goniadidae. Proc. Acad. Nat. Sci. Phila- delphia, vol. 63, pp. 234-318, pis. 15-21. Munro, C. C. A. 1933. Polychaeta errantia collected by Dr. C. Crossland at Colon, in the Panama region and the Galapagos Islands. Proc. Zool. Soc. London, pt. 1, pp. 1-96, figs. 1-36. 1933 A. Polychaeta sedentaria collected by Dr. C. Crossland at Colon, the Panama region and the Galapagos Islands. Proc. Zool. Soc. London, pt. 4, pp. 1039-1092, figs. 1-31. Schmarda, Ludwig 1861. Neue wirbellose Thiere. Bd. 1, Hft. 2, pp. 1-164, pis. 16-37. Treadwell, A. L. 1901. Polychaetous annelids of Porto Rico. Bull. U. S. Fish Comm, for 1900, vol. 2, pp. 181-210, 81 figs. 1906. Polychaetous annelids of the Hawaiian Islands. Bull. U. S. Fish Comm, for 1903 pt. 3, pp. 1145-1181, 81 figs. 1924. Polychaetous annelids collected by the Barbados- Antigua Expedition from the University of Iowa in 1918. Univ. of Iowa Studies in Natural History, vol. 10, no. 4, pp. 1-23, 2 pis. 1928. Polychaetous annelids from the Arcturus Oceanographic Expedition. Zoologica, vol. 8, no. 8, pp. 449-485, figs. 177-179. 1936. Polychaetous annelids from the vicinity of Nonsuch Island, Bermuda. Zoologica, vol. 21, pt. 1, pp. 49-68, pis. 1-3. 1937. Polychaetous annelids from the West Coast of Lower California, the Gulf of California and Clarion Island. Zoologica, vol. 22, pt. 2, pp. 139-160, 2 pis. Verrill, A. E. 1873. Invertebrate Animals from Vineyard Sound. Report U. S. Fish Comm, for 1872, Annelids, pp. 580-622, pis. 10-17. 1900. Additions to the Turbellaria, Nemertina and Annelida of the Bermudas. Trans. Conn. Acad. Sci., vol. 10, (annelids), pp. 598-669. Webster, H. E. 1884. Annelida from Bermuda. Bull. U. S. Nat. Mus., 25, pp. 305-327, 12 pis. 1941] Hollister: Caudal Skeleton of Bermuda Shallow Water Fishes 31 8. Caudal Skeleton of Bermuda Shallow Water Fishes. V. Order Percomorphi: Carangidae.1 Gloria Hollister Department of Tropical Research (Text-figures 1-20). Outline. Introduction PAGE . . 31 Key Opposite page 32 Bermuda Percomorphi Carangidae Decapterus macarellus 32 Decapterus punctalus 33 Trachinotus palometa . 34 Trachinotus goodei 35 Argyreiosus vomer 37 Trachurops crumenophthalma 38 Chloroscombrus chrysurus 39 Caranx ruber 42 Caranx bartholomaei 42 Caranx crysos 43 Caranx latus 43 Summary 44 Introduction. This is the fifth of a series of papers dealing with the caudal skeleton of Bermuda fishes. The Carangidae of Bermuda is represented by six genera and twelve species. No specimens are available of Caranx guara, considered uncommon in Bermuda, either in the collection of this de- partment or in the collections of several other institutions. Caranx guara, locally known as “Gwelly,” is seen by fishermen in shallow water but is said to be a shy, wary fish. This paper deals principally with the adult fishes, as is the case with the four preceding studies of this series, but when young stages were available these were included. The length of specimens in this paper is stand- ard length unless otherwise stated. For caudal fin terminology, general bibliog- raphy, and method of preparing specimens for this study, refer to Part I. The symbols used in the figures are EP, epural; HS, haemal spine; 1, 2, 3, 4, 5, hypurals; IHS, interhaemal spine; INS, interneural spine; NS, neural spine; UN, uroneural; UR, urostyle. 1 Contribution No. 615, Department of Tropical Re- search, New York Zoological Society, Contribution from the Bermuda Biological Station for Research, Inc. We are indebted to the American Museum of Natural History for three specimens of Argy- reiosus vomer. I take this opportunity to thank Dr. William Beebe, Director of this Department, and Mr. John Tee-Van, General Associate, for their cooperation. The drawings are by Miss Janet B. Wilson, who kindly volunteered to do this work. In Bermuda Carangidae several precaudal vertebrae are neither typical trunk nor typical caudal in structure. In all species these verte- brae have ribs (although in several specimens diminutive) , which is characteristic of the typical trunk vertebrae. Closed haemal arches are also present, a typically caudal character. These precaudals include several modified vertebrae which immediately precede the true caudal. The structure of the precaudal haemal arch in Trachurops differs radically from that of Chloro- scoitibrus and Caranx and is readily distinguished from these two latter species by the crescent- shaped arch. In Decapturus macarellus and Decapterus punctatus several precaudals and anterior true caudals are crescent-shaped and arise on the anterior half of their centra. Decapturus is the only species in this study which has the anterior true caudal haemal processes crescent-shaped. It is of interest to note that the Bermuda Synodontidae is also unlike the families studied in Parts I, II and IV in having several vertebrae in the precaudal region which deviate from the typical trunk and caudal forms. In Synodontidae there are present, between the trunk and caudal, vertebrae lacking ribs (ribs being characteristic of the trunk), and also lacking closed haemal arches with spines, this structure being charac- teristic of the caudal. In Bermuda Carangidae there is also this devi- ation from what is generally designated typical trunk and typical caudal vertebrae. This differ- ence is found in several precaudal haemals which 32 [XXVI: 8 Zoologica: New York Zoological Society have closed arches typical of the caudal region but lack the haemal spine which is typical of the trunk. Because of these differences the term precaudal centra is used as in Synodontidae. The key is based on characters of the adult. In this respect it is important to mention, as an example, that in the immature specimen of Chloroscombrus chrysurus, 6 mm., the difference in length of the three short posterior neural and haemal spines and the length of the preceding long pair of spines is not pronounced as it is in the adult. In the mature specimen both the difference in length and shape is conspicuous and is considered a key character. 1. Decapterus macarellus (Cuvier & Valenciennes). (Text-figure 7). Diagonostic Characters: 1 interneural and 1 interhaemal arising over the 4th vertebra anterior to the urostyle. The horizontal length of this posterior inter- neural and interhaemal about one and a half times longer than the anterior oblique part. Haemal base of first anterior caudal lunar- shaped. Caudal interneurals and interhaemals more slender than in the other genera. Third stout posterior neural and haemal spines dagger-shaped and extend obliquely over three quarters of the following vertebra. Material Studied. The following description is taken from one specimen caught in Bermuda, Cat. No. 9,139, KOH Cat. No. 604, length 140 mm. Caudal Osteology. Urostyle: The urostyle is conical in shape. The posterior end appears to be consolidated on the dorsal surface with the base of the uroneurals and the fifth hypural and on the ventral with part of the base of the first and second hypurals. The anterior part of the urostyle is identical in form and size with the adjacent posterior part of the penultimate centrum. Uroneurals: There is one pair of uroneurals. These paired bones extend from the anterior margin of the urostyle to the distal margin of the hypural complex. On dissecting away the caudal rays the dorsal tip appears distinct from the tip end of the fifth hypural with which the uroneurals are fused. Text-figure 9 of a 10 mm. Decapterus punctatus shows the small fifth hypural in relation to the undeveloped uroneurals. Both these bones are only slightly ossified. Where young speci- mens of Carangidae were available for study this same relation of the two bones is found. Text- figures 13, 16, 17 and 18. Hypurals: In this adult specimen there appear to be four hypural bones; two below and two above the median line. In the 10 mm. Decapterus punctatus, Text-figure 9, the large dorsal hypural is divided almost to the distal margin, indicating that in a slightly younger specimen this single bone of the adult is formed by two separate elements. This condition is found in several other young carangids. Text-figures 16, 17. The first and fifth hypurals are almost identical in size and are complimentary in their distal positions in the marginal complex. The two large median hypurals of the adult are completely separate to the anterior rounded ends which are partly covered by the base of the uroneurals and the projected lateral muscle attachment process of the first hypural. The fifth or dorsal-most hypural is recognized in the adult by the pointed tip adjacent to the end of the fourth hypural. For its entire length, with the exception of this tip, the fifth hypural is fused with the uroneurals, and appears as one with these bones. Younger specimens give the true explanation of the de- velopment and relationship of the fifth hypural and the uroneurals. In the 42 mm. specimen of Decapterus punctatus the fifth hypural is more easily distinguished by the prominent line of junction which is still present between the hy- pural and the uroneurals. Decapterus macarellus. Tail of 140 mm. specimen showing especially the prolonged pos- terior interneural and interhaemal spines. X 3.52. 1941] Hollister: Caudal Skeleton of Bermuda Shallou > Water Fishes 33 Epurals: There are two epurals. The posterior bone is slender and rod-like and slants obliquely forward and downward from the distal caudal margin to the dorsal edge of the uroneurals. The ventral tip is inserted between the two lateral uroneural bones. The anterior epural is wider than the ventral epural throughout the entire length. Anteriorly, a long finger-like projection extends forward and fills the area above the re- duced neural process of the last centrum. Similar to the smaller epural, part of the ventral tip is inserted between the uroneural bones. Additional Characters Worthy of Note: This is the only genus of Bermuda Carangidae which has the extremely long and slender posterior inter- neural and interhaemal spines. The anterior and median ends arise near the anterior zygapophyses of the fourth vertebra from the urostyle and the bones extend back and toward the centra almost to the posterior margin of the next centrum. The interneurals and interhaemals, preceding the prolonged posterior pair, are more slender than in the other species studied in this paper. The precaudal haemal processes and also the first caudal haemals are lunar-shaped and placed on the anterior part of the centra. This develop- ment is similar to Decapterus punctatus. 2. Decapterus punctatus (Agassiz). (Text-figures 8, 9). Diagnostic Characters: 2 interneurals and 2 interhaemals arising over the 4th vertebra anterior to the urostyle. The horizontal length of this posterior pair only a little longer than the anterior oblique part. Haemal base of first anterior caudal lunar- shaped. Caudal interneurals and interhaemals not as slender as those of Decapterus macarellus. Dorsal and ventral fin rays not as slender as those of Decapterus macarellus. Material Studied. The following description is taken from three specimens caught in Bermuda, Cat. No. 8,968, KOH Cat. No. 2316, length 10 mm.; KOH Cat. No. 340, length 18 mm.— taken from under a jellyfish; KOH Cat. No. 2315, length 42 mm. Caudal Osteology. Urostyle: In the 42 mm. specimen the urostyle is conical in shape and similar to Decapterus macarellus. No definite posterior tip can be seen. In the 10 mm. specimen the upturned posterior end of the urostyle extends beyond the basal end of the fifth hypural. The anterior margin is iden- tical in form with that of the preceding centrum but posteriorly the shape changes radically into a long slender upturned structure. The urostyle is slightly ossified as far as the base of the fifth hypural. Here, there is a heavy line across the urostyle and beyond this there is little ossifica- tion. Uroneurals: As in Decapterus macarellus, there is a single pair of uroneurals and in the 42 mm. specimen these paired bones are similar in shape and position to those of the above species. In Decapterus punctatus the line of junction between the fifth hypural and the uroneurals is distinct but this has disappeared in the older specimen of the other species. In the 10 mm. Decapterus punctatus the uroneurals are long, slender, curved bones which extend from the hypural margin to the anterior part of the urostyle. Throughout their entire length the uroneurals are separate from the urostyle and the fifth hypural. Text-figure 8. Decapterus punctatus. Tail of 42 mm. specimen showing the difference between this species and Decapterus macarellus in the arrangement and proportions of the interneural and interhaemal spines of the posterior groups. X 9.65. 34 Zoologica: New York Zoological Society [XXVI: 8 Decapterus punctatus. Tail of 10 mm. specimen. This stage shows the development of the 3rd and 4th hypurals which are fused in the 42 mm. specimen. The fifth hypural and the uro- neurals are separate bones in this stage. X 45.4. Hypurals: In the 42 mm. specimen there ap- pears to be four hypurals; two below and two above the median line. In the 10 mm. specimen there are three hypural bones above the median line, which indicates that the large dorsal hypural of mature specimens is a combination of two bones. Also this young specimen has a small slit in the base of the second hypural which probably indicates a one-time ancestral division of this other large hypural bone. In all Bermuda Isospondyli and Iniomi there are at least three hypurals below the median line. In general shape and proportions the hypural complex of the adult Decapterus punctatus resembles that of Decapterus macarellus, with the exception of the identity of the fifth hypural and uroneurals. Epurals : There are two epurals in the adult, as in Decapterus macerellus, and the shape and pro- portions are similar. In this species the ventral ends are not inserted between the uroneural bones, but this may be an age character. There is no trace of the epurals in the 10 mm. specimen. Additional Characters Worthy of Note: The posterior pair of interneural and interhaemal spines are conspicuously longer than the preced- ing spines and the oblique part extends posterior- ly for almost the length of the following centrum. This pair is not as long nor as slender as the corresponding pair of Decapterus macarellus. The precaudal haemal processes and also the first caudal haemals are lunar-shaped and placed on the anterior part of the centra. In the 10 mm. specimen the caudal centra are only moderately spool-shaped and the zyga- pophyses and the neural and haemal processes are undeveloped. The neural and haemal proc- esses are simple, slender bones and there is little differentiation in the development of the four anterior pair of spines. This difference is con- spicuous in the_adult. 3. Trachinotus palometa Regan. (Text-figure 10). Diagnostic Characters: 1 long and slender interneural and 1 inter- haemal arising over the 4th vertebra anterior to the urostyle. 17 interneurals in the caudal region. Bases of haemal processes arising at the center of the centra. No short or stout posterior neural and haemal spines. Material Studied. The following description is taken from one specimen caught in Bermuda, KOH Cat. No. 601, length 119 mm. Caudal Osteology. Urostyle: The anterior part of the urostyle is identical in size and shape to the adjacent part of the preceding penultimate centrum. The pos- terior end cannot be seen because it is consoli- dated with the bases of the uroneurals and hypurals. The dorsal surface is covered by the uroneurals and the ventral surface by the hypural bases. Uroneurals: There appears to be a single pair of uroneurals, the bones of which extend from the anterior dorsal edge of the urostyle to the distal margin of the hypurals. For part of the distal length the line of junction between the fifth hypural and the uroneurals is distinct. In the smaller specimen of Trachinotus goodei these bones appear separate. Hypurals: In this adult specimen there appear to be four hypural bones; two below and two above the median line. The second large hypural below the median line is solid with the exception Hollister: Caudal Skeleton of Bermuda Shalloiv Water Fishes 35 1941] of a small oblong hole near the base. In Tra- chinotus goodei of 54 mm. in length, the base is divided and a definite line can be seen which extends for half the length of the hypural bone. In the large dorsal hypural of Trachinotus palometa there is also indication that this hypural was at one time divided because a distinct line extends anteriorly from the distal margin for more than three-quarters of the length of the hypural. In the other species the base of this large dorsal hypural is divided and a narrow slit extends for about one-quarter the length of the bone and does not quite join with the slit extend- ing from the distal margin. In the species under discussion the fifth or dorsal-most hypural is closely associated with the uroneurals but each can be identified at their dorsal extremity by distinct and separate pointed tips and a line which extends anteriorly for almost half the length. From here the two are fused. In the other species this fifth hypural is more distinct. Hypural numbers have been omitted from this illustration because of the inclusion of the caudal fin rays; for hypural reference, see Text-figure 11 of Trachinotus goodei. Epurals: There are three epurals in this species and genus. The anterior or first epural is the largest, which is true of the adults of all the other species treated in this paper. Anteriorly, there is a broad thin growth which ahnost fills the entire area dorsal to the penultimate centrum and the urostyle. The second and third epurals are long, slender bones and very closely associated. At both the proximal and distal ends the tips can be distinguished as separate bones. The ventral ends are not inserted between the uroneurals as is the case in Trachinotus goodei. Often this appears as a growth character but the specimen of this species is less than half the size of Tra- chinotus palometa and the tips of the former are inserted. Additional Characters Worthy of Note: The il- lustration of this species is the only one which includes the caudal fin rays because the variation in the count is slight in the other species of Carangidae and therefore of little importance. Here the count is : dorsal raylets 6 plus 1 1 dorsal rays, and ventral raylets 6 plus 9 ventral rays, making a total of 17 dorsal plus 15 ventral. 4. Trachinotus goodei Jordan & Evermann. (Text-figure 11). Diagnostic Characters: 2 broad interneural and interhaemal spines arising over the 5th vertebra anterior to the urostyle. 15 interneurals in the caudal region. Bases of the haemal processes arising on the anterior half of the centra. No short or stout posterior neural and haemal spines. Material Studied. The following description is taken from one Text-figure 10. Trachinotus palometa. Tail of 119 mm. specimen with three epurals. One interneural and one interhaemal spine is associated with the 4th vertebra from the urostyle. X 3.6. 36 [XXVI: 8 Zoologica: New York Zoological Society 4 3 2 Text-figure 11. Trachinotus goodei. Tail of 54 mm. specimen with three epurals and evidence of two pairs of uroneurals. Two interneurals and two interhaemals associated with the 5th vertebra from the urostyle. X 8.5. specimen caught in Bermuda, KOH Cat. No. 438, length 54 mm. Caudal Osteology. Urostyle: The anterior part of the urostyle is cone-shaped and its margin similar to the pos- terior part of the penultimate centrum. The reduced posterior tip of the urostyle is hidden and consolidated with the base of the uroneurals. The uroneurals protect the dorsal and posterior part of the urostyle and the bases of the hypurals abut its ventral and posterior surface. Uroneurals: These paired bones are unlike those of the other species studied in this paper. For a complete interpretation a series of younger specimens is needed. In this 54 mm. fish there are two pointed tips which form part of the hypural margin. Each represents paired lateral bones, the bases of which partly overlap the urostyle and the fifth hypural. The anterior dorsal uroneural extends from above the anterior margin of the urostyle where it is deep and wing- shaped and where bases of the epurals are inserted between its lateral bones. It extends posteriorly and dorsally and diminishes into a slender bone with a pointed distal tip. The second uroneural is similar distally but it arises immediately on the anterior part of the urostyle centrum. It overlaps the base and dorsal surface of the fifth hypural for all but the distal tip end. A broad, thin, paired bone covers part of the anterior half of the uroneurals and the fifth hypural. In our specimen of Trachinotus palometa of 119 mm. in length, only one pair of uroneurals can be found if the second pointed tip is to be interpreted as the fifth hypural. A series of young specimens is necessary in the case of both species to inter- pret adequately the development of this part of the hypural complex. Hypurals: There are two hypurals ventral and two dorsal to the median line. In the second large hypural there is definite evidence of a divi- sion of this bone, for there remains a suggestion of a double base with a line extending from here for half the length of the hypural. In none of the Bermuda carangids has this second hypural ap- peared as two separate bones, even in the very young of several species. This is the reason for consistently designating this particular hypural as a single bone. Its counterpart, the large hypural dorsal to the median line, has shown in several young specimens that it is formed by the fusion of two separate bones, which gives a cri- terion for calling this hypural three and four. In this specimen hypurals three and four are distinct both anteriorly and posteriorly and only a small area in the center has fused beyond iden- tity of the two separate bones. The fifth and most dorsal hypural is similar in size at its distal extremity to the first hypural and balances this bone in its position in the hypural fan. Its length, however, is less and the median end di- minishes to a wedge-shaped tip. Epurals: There are three epurals, which is also true of Trachinotus palometa. But in the species of this description the epurals appear more separate and individual than in Trachinotus palometa, which is a larger and older specimen. In this 54 mm. Trachinotus goodei the first or anterior epural is the largest, the second inter- mediate and the third, or posterior, the smallest. This is true of both their lengths and widths. The ventral ends are inserted between the lateral bones of the uroneurals. This is not so in the other species. 1941] Hollister: Caudal Skeleton of Bermuda Shalloic Water Fishes 37 5. Argyreiosus vomer (Linnaeus). (Text-figures 12, 13). Diagnostic Characters: The caudal skeleton is conspicuously deeper and relatively shorter than any of the other species treated. 3 anterior caudal haemal spines with triangular projections on the anterior surface at mid- length. 3 short posterior neural and haemal spines stout and club-like. Material Studied. This description is from a study of three speci- mens presented by the American Museum of Natural History. Cat. No. 14,184 from Pom- pano, Florida; KOH Cat. No. 2321, length 18 mm. Cat. No. 12,505 from Mastic, L. I.; KOH Cat. No. 2322, length 45 mm. Cat. No. 13,583 from Rio de Janeiro; KOH Cat. No. 2323, length 140 mm. Caudal Osteology. Urostyle: In the adult specimen of 140 mm. in length the posterior end of the urostyle is con- solidated with the dorsal bones and cannot be distinguished. In our young specimen of 18 mm. the urostyle extends dorsally to the hypural margin. An- teriorly, the urostyle is almost a perfect half centrum but at the base of the second hypural it turns abruptly and extends in an oblique-dorsal direction and becomes a slender tapering rod. At the basal tip of the fifth hypural a mark ap- pears which indicates consolidation in this youth- ful stage of the urostyle. The bases of all of the hypurals are separate from the urostyle with the exception of the fifth hypural whose anterior surface is in conjunction with the urostyle. Uroneurals: There is a single pair of uroneurals which are more distinct in the 18 mm. specimen than in the two older stages. In the smallest fish the anterior part is already enlarged and fills the area dorsal to the urostyle but the spool- shaped urostyle is distinct. In the largest speci- men the urostyle and the uroneurals have become one and there is only a suggestion of the urostyle as such. Both the anterior and the posterior portion of the uroneurals have increased in pro- portions. Distally, the uroneurals and fifth hypural have fused and there is only a suggestion of the hypural which in the 18 mm. specimen was a distinct bone. Hypurals: In the adult specimen there are four hypural bones, two below and two above the median line. In the 18 mm. fish it is clear that the large hypural dorsal to the median line is the result of the fusion of two bones. Truly young specimens would undoubtedly show this as they have in several other species in this paper. The dorsal, or fifth hypural, is distinct in the 18 mm. specimen but in the 140 mm. fish it has be- IHS Text-figure 12. Argyreiosus vomer. Tail of 140 mm. specimen. The bones of this deep-bodied fish are con- spicuously stout and the centra less elongate than in the other species. X 3.23. 38 Zoologica: New York Zoological Society [XXVI: 8 4 3 2 Text-figure 13. Argyreiosus vomer. Tail of 18 mm. specimen showing the posterior end of the urostyle, the uroneurals and the fifth hypural. These are not distinct in the 140 mm. fish. There is a hint of the division of the 3rd and 4th hypurals. X 25.2. come one with the uroneurals. The first hypural is not as massive as the corresponding bone in the adult, and there is a conspicuous large hole in the base or arch area. This is present in most of the neural and haemal bases but in the largest specimen the holes have disappeared in the three posterior processes and become smaller in the preceding arch bases. Epurals: There are two epurals. In the 18 mm. specimen their shape in general is similar with the exception of the anterior projection on the first bone, which is a growth character. In no respect do these two bones resemble one another in the 140 mm. specimen. The anterior projection from the base of the first epural has become solid and massive and fills the area above the reduced neural process of the penultimate centrum. The rod-like form of the young specimen has disap- peared entirely. The second or posterior epural has changed very little from that of the young stage. The ventral ends of both epurals in all three stages are inserted between the lateral bones of the uroneurals. Additional Characters Worthy of Note: As is typical of this deep bodied species, the inter- neural and interhaemal spines are massive in the adult specimen. In each there is a central shaft which is heavily ossified and has lateral projec- tions on the horizontal part. In the smallest specimen there is the beginning development of this projection and all other parts of these bones are only slightly ossified. Several anterior caudal haemal spines are unique in development as com- pared with the other Bermuda Carangidae. The first and second are usually stout and at mid- length their surfaces abut. The three following haemal spines are also stout and have on their anterior sides heavy triangular projections. These are present in all three stages. This is illustrated in the Key (Text-fig. 5). 6. Trachurops crumenophthaltna (Bloch). (Text-figure 14). Diagnostic Characters: The posterior group of two interneurals and two interhaemals associated with the fourth vertebra from the urostyle. The haemal processes of several precaudal centra are crescent-shaped and the spines arise on the anterior half of the centra. Material Studied. Three specimens were used for this description and all were caught in Bermuda. Cat. No. 9,140, KOH Cat. No. 615. Cat. No. 9,140, KOH Cat. No. 614, length 115 mm. Cat. No. 25,147, KOH Cat. No. 1135, length 108 mm. 1941] Hollister: Caudal Skeleton of Bermuda Shallow Water Fishes 39 Caudal Osteology. Urostyle: In all specimens studied the posterior end of the urostyle is embedded in the terminal bones. The bases of the uroneurals and the hypurals surround this area. The anterior part of the urostyle is almost a perfect half centrum in shape and size. The dorsal surface is covered by the heavy uroneurals and the ventral surface by the base of the first hypural . Uroneurals: There appears to be but a single pair of uroneurals. In our adult specimens these paired bones are massive and extend obliquely backward and dorsally from above the anterior part of the urostyle to the distal margin of the hypurals. In the anterior area a small portion of the ventral ends of the epurals are wedged be- tween the lateral bones of the uroneurals. The oblique distal part probably represents the fifth hypural fused with the uroneurals. Our material is too advanced in development to illustrate this fact, which is found in the young specimens of the closely related genera. Hypurals: Our material shows two hypurals ventral to and two dorsal to the median line. The central bones are large and fan-shaped, which is characteristic of this family. Young specimens would undoubtedly show that the large dorsal hypural was formed by the fusion of two bones and also that the fifth hypural was separate. In these adult specimens the fifth hypural is fused with the uroneurals and not distinguishable from them. On the basis of repeated evidence in the young of several closely related genera the hy- purals are considered as five, which is seen in the Text-figure 14. Epurals: The two epurals fill the area above the penultimate centrum and the urostyle. The long projection of the first epural extends an- teriorly above the reduced neural process of the penultimate centrum. The base is inserted be- tween the uroneurals. The second or posterior epural is long and slender and slightly enlarged at the ventral end, which lies between the uro- neurals. 7. Chloroscotnbrus chrysurus (Linnaeus). (Text-figures 15-18). Diagnostic Characters: 26 or 27 interhaemals in the caudal region. 6 interhaemals between the first and second anterior caudal haemal spines. Material Studied. The following description is based on seven specimens taken in Haiti. Cat. No. 6,842, KOH Cat. No. 2317, lengths 100 mm. and 80 mm. Cat. No. 7,183, KOH Cat. No. 2318, lengths 15, 10, 7.5, 6.5, 6 mm. Caudal Osteology. Urostyle: Four stages of development were found in specimens of 6 mm., 10 mm., 15 mm., and 100 mm. in length. In the largest fish, which is fully ossified, the posterior end of the urostyle is completely hidden by the overlapping uroneurals and hypurals. As in the other Bermuda Carangidae the anterior part of the urostyle is almost a complete half centrum and is similar in proportions to the adjacent part of the preceding centrum. In the 6 mm. specimen the urostyle is a long, slender, rod-like structure which extends from the posterior margin of the penultimate centrum to the distal margin of the hypurals. Anteriorly, the proportions are similar to the adjacent part of the preceding centrum. The urostyle extends posteriorly in an oblique and dorsal direction, tapering to about one-half the depth. The ossification is delicate through- out and ends midway between the bases of the fourth and fifth hypurals. A cartilaginous rod INS Text-figure 14. Trachurops crumenophthalma. Tail of 115 mm. specimen showing two interneurals and two interhaemals associated with the 4th and 5th vertebrae from the urostyle. X 11.65. 40 Zoologica: New York Zoological Society [XXVI: 8 INS Text-figure 15. ■ Chloroscombrus chrysurus. Tail of 100 mm. specimen showing the 3rd and 4th hypurals fused and the 5th hypural united with the uroneurals. X 33.3. extends from here and disappears in the bases of the caudal rays. In the 10 mm. specimen ossi- fication is more pronounced and the posterior dorsal end is seen at the center of the base of the fourth hypural. Anteriorly, the shape has begun to resemble that of the adult. The 15 mm. specimen shows still greater ossification and con- solidation. Uroneurals: There is a single pair of uroneurals, which in the adult is indicated only at the distal end. These bones have become fused with the fifth hypural in the largest specimen. The three younger stages show clearly the development of the uroneurals. In the 6 mm. specimen this paired bone extends dorsally above the urostyle from the region of the second hypural to the margin of the hypurals. It is long and slender and the anterior end is slightly larger than the posterior tip. It is of importance to note the relation of this bone to the others and its inde- pendent position, which is found only in the very young stages. Chloroscombrus chrysurus. Tail of 6 mm. specimen which is partly ossified. There are no interneurals or interhaemals or epurals. The 3rd, 4th and 5th hypurals and the uroneurals are separate bones. The urostyle extends into the hypural margin. X 104. 1941] Hollister: Caudal Skeleton of Bermuda Shallow Water Fishes 41 Text-figure 17. Chloroscombrus chrysurus. Tail of 10 mm. specimen showing increased ossification. The centra are spool-shaped and the epurals, uroneurals and base of the interneural and interhaemal spines have appeared. The 3rd and 4th hypurals have begun to fuse and the uroneurals abut the fifth hypural. X 83. In the 10 mm. specimen the uroneurals have increased anteriorly both in depth and length and almost reach the anterior margin of the uro- style. The shape of this end has changed from a blunt rounded tip to a deeper indented form. The posterior part is reduced and the distal end appears in conjunction with almost half the proximal length of the fifth hypural. In this region the uroneurals have grown over the uro- style. The anterior part is still separate from the urostyle. In the 15 mm. specimen the uroneurals re- semble the structure of the adult more than the 10 mm. specimen, and the shape and proportions are similar. The base of the uroneurals has fused with the dorsal surface of the urostyle and in only one place is there any indication of the youthful stage where all the under part of this bone was entirely separate. There still remains two perforations or unossified areas. The ante- rior part of the uroneurals abut the reduced neural process of the penultimate centrum and extend dorsally almost to the distal margin of the fifth hypural. In this stage the uroneurals have be- come entirely consolidated with the dorsal surface of this fifth hypural. Hypurals: In the adult 100 mm. specimen there are two hypurals below and two above the median line. The first and fifth hypurals are the same size distally and flank the ventral and dorsal part of the hypural fan. Two large hypurals form the central part of the hypural complex ; one below and one above the median line. From a study of the young specimens it is certain that the large dorsal hypural is formed by the fusion of two bones and that the fifth hypural has fused with the uroneurals. In the 6 mm. specimen, which is the youngest stage available for study of the Bermuda Caran- gidae, there are five distinct hypural bones. The first, or anterior bone, is long and slender and slightly enlarged at the median base. This base, which is the haemal arch, develops rapidly in the 10 mm. and 15 mm. stages. In the smaller fish a square, wing-like growth projects ante- riorly and abuts the haemal process of the pre- ceding centrum and there is a large hole or unossified area in the center of this growth. In this stage the lateral ridge, for the muscle attach- ment, is a small triangular projection from the base and overlaps slightly the base of the second hypural. In the 10 mm. specimen this ridge has more than doubled in size and proportion. The first hypural of the 15 mm. specimen resembles the corresponding bone of the adult fish with the exception of the density of ossifica- tion, and extent of development of the muscle attachment process, and proximity to the ventral surface of the urostyle. There is still considerable space between the haemal arch and the urostyle, 42 Zoologica: New York Zoological Society [XXVI: 8 which entirely disappears in the adult specimen. This is also true of the preceding haemal arches. The second hypural changes less in shape and proportions than any of the other hypural bones. In the 6 mm. specimen the base is divided into two by a long oval slit which extends for about one-quarter the length of the bone. In the 10 mm. specimen the two bases have fused but a small oval hole remains. In the 15 mm. fish there is no indication of either the two bases or the slit of the younger stages. The third and form part of the caudal margin. In the 15 mm. specimen the degree of ossification is identical with the other caudal bones. The shape of both epurals is similar to that of the adult, and the median ends are ir erted between the uroneurals, which is also an adult character. Additional Characters Worthy of Note: The illustrations show the increase in ossification with age and also the change in shape and pro- portions of all of the caudal bones. In the 6 mm. specimen there is no evidence of epurals, inter- Text-figure 18. Chloroscombrus chrysurus. Tail of 15 mm. specimen. The posterior end of the urostyle is reduced and the distal portion of the uroneurals flank the side of the 5th hypural. In the 100 mm. specimen these two bones have fused. X 42. fourth hypurals are separate bones in the young- est stage and in the 10 mm. specimen their distal margin has fused and the space between become less. The 15 mm. fish still lias two bases which represent the two separate bones of the younger stage but the slit between has diminished by more than half its former length. In the 100 mm. specimen there is no trace of a slit or double base. The fifth hypural appears as a very small wedge-shaped bone between the distal end of the urostyle and the fourth hypural . Its growth and development from this stage to the adult is remarkable and without several very young specimens this hypural would not be identified as an hypural bone. In the youngest stage this bone is entirely separate from the surrounding structures. In the 10 mm. fish the bone has in- creased in length and depth and about one-lialf of its anterior length abuts the uroneurals on the dorsal surface. In the 15 mm. specimen the length is greatly increased and all but a small part of the dorsal end is in conjunction with the uroneurals. The adult specimen of 100 mm. shows that the fifth hypural and the uroneurals have fused and all that remains of their separate identity is a light line at the distal extremity. Epurals: There are two epurals in this species. Neither of these bones can be found in the 6 mm. fish. In the 10 mm. specimen these epurals appear with delicate ossification and as long slender bones. The anterior epural is longer and broader than the second one and has the be- ginning of the anterior projection near the median end. Both bones are free and their distal tips neurals and interhaemals. The centra are rectangular in shape with considerable distance between their margins. In the 10 mm. specimen the spool-shape of the centra begins to change and to resemble the adult. The epurals are lightly ossified and also the interneurals and interhaemals. The zygapophyses are present on the dorsal and ventral surfaces of the centra. With the consolidation from a long rectangular structure to a shorter spool-shaped centra the neural and haemal processes appear to shift from their youthful anterior position to a central position. The three reduced posterior neural and haemal spines of the adult are simple and unmodified in the 6 mm. and 10 mm. specimens. 8. Caranx ruber (Bloch). (Text-figure 19). 9. Caranx bartholomaei Cuvier & Valenciennes Diagnostic Characters (both species) : Total of 24 interhaemals in the caudal region. Total of 22 interneurals in the caudal region. 4 interhaemals between the first and second anterior caudal haemal spines. 3 interneurals and 3 interhaemals in posterior group in three specimens. The oblique part more slender than in the two following species. Minor species differences cited in Key and in text. 1941] Hollister: Caudal Skeleton of Bermuda Shallow Water Fishes 43 Material Studied. Caranx ruber; one specimen taken in Haiti. Cat. No. 7,016, KOH Cat. No. 2320, length 160 mm. Caranx bartholomaei; two specimens caught in Bermuda. Cat. No. 25,057, KOH Cat. No. 1074, length 50 mm. Cat. No. 25,065, KOH Cat. No. 1073, length 52 mm. Caudal Osteology. Urostyle: The reduced terminal end is hidden in both species by the heavy uroneurals on the dorsal and the base of the hypurals on the ventral surface. No small specimens are available to show the development. The anterior part is cone-shaped and of the same proportions as the posterior half of the penultimate centrum. The dorsal surface is covered by the uroneurals and the ventral by the first hypural in Caranx ruber (Text-fig. 19). In Caranx bartholomaei, which is one-third the length, the base of the first hypural is still separate from the ventral surface of the urostyle. Uroneurals: In the 50 mm. specimen of Caranx bartholomaei there is indication of one pair of uroneurals. At the distal end the line of junction between the uroneurals and the fifth hypural still remains. There is no evidence of this in the 160 mm. specimen of Caranx ruber. In this species the two bones have fused. Hypurals: In both species there are two hy- purals which are ventral to and dorsal to the median line. Their positions and proportions are similar in the corresponding bones. Because of facts presented by several young stages in closely related genera in this family, the hypural bones of these species are numbered 1 to 5 in the illustration. Epurals: There are two epurals in both species and the corresponding bones are similar in shape and position. In the distal dorsal portion of the smaller Caranx bartholomaei the epurals are more slender. This has been found to be an age character in other genera. Additional Characters Worthy of Note: In Caranx ruber the posterior neural zygapophyses have hook-like projections on the dorsal surface. This is shown in the anterior vertebra in the illustration. The development is especially prominent in the central caudal area. In Caranx bartholomaei the dorsal surfaces of the corre- sponding zygapophyses are smooth. There is also a difference in the development of the anterior neural zygapophyses of the two species. In Caranx ruber the surfaces are jagged and notched, especially in the central caudal area, but this is not true in Caranx bartholomaei. In a detailed study of the two species many minor differences such as these distinguish each as a different species. 10. Caranx crysos (Mitchill). 11. Caranx latus Agassiz. (Text-figure 20). Diagnostic Characters (both species) : Total of 17 to 20 interhaemals in the caudal region. Total of 15 to 18 interneurals in the caudal region. 3 interhaemals between the first and second anterior caudal haemal spines. 2 interneurals and 2 or 3 interhaemals in the posterior group in the five specimens studied. The oblique part broader than in the two preceding species. Minor species differences are cited in the Key and in the text. Text-figure 19. Caranx ruber. Tail of 160 mm. specimen. This figure represents both this species and Caranx bartholomaei which in gross structure are similar. X 2.74. 44 Zoologica: New York Zoological Society [XXVI: 8 INS Text-figure 20. Caranx latus. Tail of 87 mm. specimen. This figure represents both this species and Caranx crysos, the gross structure being similar. X 15.87. Material Studied. Caranx crysos; one specimen taken in Haiti. KOH Cat. No. 2319, length 130 mm. Caranx latus; four specimens caught in Bermu- da. KOH Cat. No. 1070, length 44 mm. KOH Cat. No. 412, length 63 mm. KOH Cat. No. 649, length 80 mm. KOH Cat. No. 1040, length 87 mm. Caudal Osteology. Urostyle: The anterior part of the urostyle in both species is cone-shaped and almost identical in shape with the posterior half of the preceding centrum. The dorsal surface is fused with the bases of the uroneurals and the ventral surface with the base of the first hypural. The posterior end cannot be distinguished in specimens as old as those in this collection. Text-figure 20 repre- sents both species. Uroneurals: In the 87 mm. specimen of Caranx latus one pair of uroneurals is found where it has fused during growth with the fifth hypural. In Caranx crysos, which is 130 mm. in length, there is no evidence left and the fifth hypural and uroneurals appear as one structure. Hypurals: In both species there are two hy- purals below and two above the median line. They are similar in proportions and arrangement to the other species of Caranx. None of these specimens are young enough to show the develop- ment and eventual fusion of the hypurals but the figure is numbered to show the position of five hypurals which the young of closely related genera have demonstrated. Epurals: There are two epurals, as in all but one genus of Carangidae. In the distal portion of the smaller Caranx latus these bones are more slender than those of Caranx crysos. This de- velopment is correlated with age in other genera. Additional Characters Worthy of Note: In Caranx crysos there are 20 interhaemals in the caudal region and 18 interneurals. There are fewer in Caranx latus, which has 17 interhaemals and 15 or 16 interneurals in the caudal region. The shapes and proportions of the zygapophyses differ in the two species, also the presence and absence of neural and haemal arch perforations. In a detailed study of the minor differences in both species these can be distinguished as two separate species. Summary. The following facts correlate the salient simi- larities found in the study of Bermuda Caran- gidae. Vertebral count: 10 trunk plus 14 caudal in all species but Caranx crysos and Decapterus punc- tatus. These have 10 trunk and 15 caudal. In both, only one specimen was available for the count. Caudal fin count : There is only slight variation among the species. The illustration of Tra- chinotus palometa (Text-fig. 10) shows the ar- rangement. It is usually 6 raylets plus 11 rays dorsal; 6 raylets plus 9 rays ventral, making a total of 17 dorsal, plus 15 ventral. Epurals: There are two epurals in all but the genus Trachinolus, which has three. Hypurals: In the adults of all species there are four hypurals; two below and two above the median line. The two central bones are large and fan-shaped. The dorsal or fifth hypural is fused with the uroneurals in all adults. In the young stages it is separate. In young stages the large dorsal hypural is found to be formed by the fusion of two separate bones. There remains a hint in several of the young of a one-time division of the large ventral or second hypural. This is not complete in any specimens of this 1941] Hollister: Caudal Skeleton of Bermuda Shalloiv Water Fishes 45 family. In all Bermuda Isospondyli and Iniomi there are at least three hypurals below the median line. Uroneurals: In all species but Trachinotus goodei there is a single pair of uroneurals. Here there is a suggestion of a second pair, but young specimens are necessary to ascertain this fact. Elongate spines in caudal complex: In the adult of all species one long neural spine and two long stout haemal spines form the anterior part of the caudal complex. Anterior caudal haemals: Decapterus is the only genus in this series which has the bases of the first few anterior caudal haemals crescent- shaped and situated on the anterior half of the centra. The typical shape is triangular. Pre-caudal haemals : Decapterus and Trachurops have crescent-shaped precaudal haemal bases which are situated on the anterior half of the centra. Specialized neural process: Adult specimens of all species of Bermuda Carangidae have a reduced neural process on the penultimate centrum. The long anterior projection from the first epural extends into the area above this process. ISHES. GROUP I 3 posterior neural and mal caudal spines spicuously short or st 2 epurals. c A R A N G I D A Text-figure 1. Caranx ruber Posterior neural zygapophyses with hook- like projections on the dorsal surfaces, especially in the central caudal area. Anterior neural zygapophyses with notched dorsal surfaces especially in the central caudal region. -Group A interhaemals and 20 to 22 interneurals in caudal region. linterhaemals between the first and second ‘anterior caudal haemal spines. Caranx bartholomaei No hook-like projections on dorsal surfaces of posterior neural zygapophyses. Anterior neural zygapophyses with smooth dorsal surfaces. GROUP II No conspicuously short terior neural and h spines. 3 epurals. Caranx crysos 20 interhaemals and 18 interneurals in cau- dal region. -Group B to 20 interhaemals and 15 to 18 inter- neurals in caudal region, interhaemals between the first and second anterior caudal spines. Caranx latus 17 interhaemals and 15 or 16 interneurals in caudal region. KEY TO CAUDAL SKELETON OF BERMUDA SHALLOW WATER CARANGID FISHES. (Text-figs. 1-6). c A R A N G I D A E GROUP I 3 posterior neural and hae- mal caudal spines con- spicuously short or stout. 2 epurals. Text-figure 1. Decaptcrus macarellus 1 intemeural and 1 interhaemal arising over the 4th vertebra anterior to the urostyle. The horizontal part about one and a half times longer than the anterior oblique part. Sub-Group A Horizontal length of posterior inter- neural and inter- haemal conspicu- ously longer than the preceding spines. 1st anterior caudal haemal arch cres- cent-shaped and on anterior half of centrum. Text-figure 3. Decapterus punctatus 2 intemeurals and 2 interhaemals arising over the 4th vertebra anterior to the uro- style. The horizontal part only a little longer than the oblique part. Sub-Group B Horizontal length of posterior inter- neural and inter- haemal not con- spicuously long and the general shape the same as the spines im- mediately preced- ing. 1st anterior caudal haemal arch not crescent -shaped and on entire centrum. Division I Argyreiosus vomer 3 anterior caudal hae- mal spines with tri- angular projections at mid-length. Posterior vertebrae rel- atively shorter and more massive than in Division II. Sub-Division A Trachurops crumenophthalma Posterior group of intemeurals and inter- haemals associated with 4th vertebra from the urostyle. Precaudal haemal process crescent-shaped and on anterior half of centrum. Division II No triangular projec- tions on anterior caudal haemals. Posterior vertebrae rel- atively longer and more slender than in Division I. Group I Chloroscombrus chrysurus 26 or 27 interhaemals. 6 interhaemals between the first and second anterior caudal hae- mal spines. Sub-Division B Posterior group of intemeurals and inter- haemals associated with 5th vertebra from the urostyle. Precaudal haemal process triangular-shaped and covers all of centrum. GROUP II Trachinotus palometa No conspicuously short or stout pos- terior neural and haemal caudal spines. 3 epurals. 1 intemeural and interhaemal spine over the 4th vertebra from the urostyle. 17 intemeurals in caudal region. Trachinotus goodei 2 intemeural and interhaemal spines over the 5th vertebra from the urostyle. 15 intemeurals in caudal region. Group II Caranx Less than 26 interhaemals. 3 or 4 interhaemals between the first and second anterior caudal haemal spines. Caranx ruber Posterior neural zygapophyses with hook- like projections on the dorsal surfaces, especially in the central caudal area. Anterior neural zygapophyses with notched dorsal surfaces especially in the central caudal region. Sub-Group A 24 interhaemals and 20 to 22 intemeurals in caudal region. 4 interhaemals between the first and second anterior caudal haemal spines. Caranx bartholomaei No hook-like projections on dorsal surfaces of posterior neural zygapophyses. Anterior neural zygapophyses with smooth dorsal surfaces. Caranx crysos 20 interhaemals and 18 intemeurals in cau- dal region. Sub-Group B 17 to 20 interhaemals and 15 to 18 inter- neurals in caudal region. 3 interhaemals between the first and second anterior caudal spines. Caranx latus 17 interhaemals and 15 or 16 intemeurals in caudal region. Text-figure 2. 1941] Crandall: Description of an Egg of the Long-tailed Bird of Paradise 47 9. Description of an Egg of the Long-tailed Bird of Paradise. Lee S. Crandall. Plate I. On June 16, 1937, a pair of Long-tailed Birds of Paradise, Epimachus meyeri meyeri Finsch, collected by F. Shaw Mayer “near the head of the Waria,” were received at the Zoological Park. During following years, frequent displays of the male were noted but as this is a common practice in the group and there was no evidence of nesting activity, we expected no further re- sult. However, the female was noticed in coma- tose condition on the afternoon of February 28, 1941, and on examination was found to be egg- bound. She responded well to treatment and the egg was safely laid about two hours later. Since the egg of this species appears to be undescribed, the following notes are given. The egg measures 44.25 X 28.5 mm. Its weight was 18.9 grams or approximately .62 ounces. The ground color is buff (close to Cart- ridge Buff of Ridgway). It is heavily blotched and streaked at the large end, the markings decreasing toward the small end, which is almost clear. The under markings are grayish (close to Pale Mouse Gray of Ridgway) while those on the surface are reddish-brown (near Russet of Ridgway) . 48 Zoologica: New York Zoological Society [XXVI: 9 EXPLANATION OF THE PLATE. Plate I. Fig. 1. Egg of Epimachus meyeri meyeri Finsch. Actual size. CRANDALL. PLATE I FIG. 1. DESCRIPTION OF AN EGG OF THE LONG-T A I LED BIRD OF PARADISE 1941] Breder cfc Krumhols: On the Uterine Young of Dasyatis sahinus 49 10. On the Uterine Young of Dasyatis sabinus (Le Sueur) and Dasyatis hastatus (De Kay). C. M. Breder, Jr. & Louis A. Krumholz New York Aquarium and University of Illinois. (Text-figures 1 & 2). The small sting ray, Dasyatis ( Amphotistius ) sabinus (Le Sueur) is abundant along the west coast of Florida in the vicinity of the field station of the New York Aquarium on Palmetto Key. Here mature gravid females may be found at sizes not greatly exceeding those of the relatively new- born young of the much larger Dasyatis hastatus (De Kay). Apparently there is a dearth of material in the literature concerning the size of gravid fe- males, the number of young per individual and the sex ratios of the young. The catches of the stop-netters operating in Bull’s Bay (Cara Pelau on the northern side of Charlotte Harbor on the U. S. coastal charts) provided the data of the accompanying tables, except as noted. The 1939 material was collected by M. B. Bishop of Peabody Museum, Yale University. The sizes of some of the gravid females of D. sabinus did not greatly exceed those of the new-born young of D. hastatus taken on June 29. These latter ranged from 160 to 176 mm. in length, while one not quite mature male of the former was only 174 mm. long. Three full-term embryos taken from one female D. sabinus varied in length from 99 to 100 mm. It can be seen that there is a steady increase in size of the embryos of D. sabinus from a mean of 35.8 mm. on June 19 to a mean of 99.6 on July 11. Also, there was no overlapping of the means and extremes on the various dates. If this can be taken to indicate the growth rate, on so few data, it is remarkably rapid. If the length of the disc (from the tip of the snout to the posterior border of the disc) is com- pared with the width, it is seen that the two dimensions are nearly equal. Moreover, this relationship holds from the smallest embryos obtained through to the adult (Text-fig. 1), clearly showing it to be a straight line relation- ship. D. hastatus, which is a little wider than long, has a similar straight line relationship, but is not so close to a 45° angle. This is clearly a most unusually constant growth rate for which there is no superficial explanation at this time. It is not of importance in intra-uterine life nor can it be in post-uterine life when the extremely different relationships of various other species of rays, otherwise much alike, are taken into con- sideration. The growth relationship of these is so constant that it might be simulated to flap- jack batter spreading out on a pan. Except in the early embryos, this relationship is sufficient to separate these two species which so closely resemble each other. Text-fig. 1 also indicates the relative sizes of the adults and also the size of the young at delivery. A full-term young of each species is illustrated in Text-fig. 2 for purposes of comparison. The general shape and the presence of the keel on the tail of D. hastatus make separation simple. The folding in of the tail, at its origin from the body, in D. hastatus as compared with the flaring out in D. sabinus is another characteristic. It was not always possible to identify each young fish with its mother, because in capture some of the females aborted, resulting in a mix- ture of mothers and young. These young, even though the yolk sac was still hanging outside the body, were able to swim around actively when placed in an aquarium. To avoid possible mistakes and confusion, the young have been considered in groups as taken. The material did show a variation in numbers of young from one to three in D. sabinus and from two to five for D. hastatus. On June 19, 1940, there was one gravid D. sabinus which carried but one embryo. This embryo was nearly twice the size of the others taken on the same date. However, on June 28, 1940, we took another female which carried but one embryo and this was very nearly the same size as others taken on the following day. The sex ratios of all the embryos of D. sabinus combined (25) showed 56% to be male. How- ever, one mother carried three female embryos. 50 Zoologica: New York Zoological Society Table I. [XXVI: 10 Size and Sex of Dasyatis Embryos. Dasyatis sabinus. Disc in mm. Female Male All Date Length Width Sex Max. Mean Min. Max. Mean Min. Mean 4/9/39 55 55 m) Captiva Pass 49 48 f) 49 49 49 55 55 55 52 6/19/40 30 29 f 33 32 f 33 30 f 50 53 f 33 30 m 50 38.5 30 33 33 33 35.8 6/28/40 78 81 m — — ■ — ■ 78 78 78 78 6/29/40 71 72 f 74 77 f 82 81 f 83 87 f 68 69 m 69 67 m 70 71 m 70 73 m 70 74 m 71 72 m 71 * 74 m 74 75 m 75 76 m 77 79 m 79 81 m 79 81 m 83 77.7 71 79 72.9 68 75.2 7/11/40 99 99 f East side of Useppa Island 99 99 f 100 101 f 100 99.6 99 — — — 99.6 Dasyatis hastatus. 4/9/39 76 83 f Captiva Pass 75 83 f 76 75.5 75 75.5 4/10/39 97 102 m Captiva Pass 95 101 f 97 97 97 95 95 95 96 6/24/40 85 87 f 82 88 f 86 87 m 85 89 m 82 82 82 85 85 85 83.5 6/29/40 172 193 f 176 197 f 160 189 m 170 192 in 168 195 m 176 174 172 170 166 160 169.2 The relatively few young of D. hastatus (9) showed 55.5% to be male. Since there were in all 12 gravid D. sabinus and 25 embryos, they averaged 2.08 per mother. The four D. hastatus females averaged 3.25 per mother. These latter are much larger in relation to the size of the young they liberate; the difference in number seems to be purely mechanical. 1941] 51 Date 6/19/40 Length to notch 6/28/40 6/30/40 7/11/40 6/24/40 6/29/40 Unlabeled Breder & Krumholz: On the Uterine Young of Dasyatis sahinus Table II. Size and Sex of Dasyatis Adults. Dasyatis sabinus. Length Width Sex (201 — f not gravid 202 — f not gravid 214 — f not gravid 222 223 f f gravid gravid ■5 embryos 244 — f gravid 247 — • f not gravid 262 — f not gravid 190 — m 190 — m 199 — m 213 — m 218 — m (234 — m 238 245 f gravid 1 embryo 174 179 m imm. — 235 f gravid ■ — 241 f gravid 243 245 f gravid — 254 f gravid 16 embryos — 260 f gravid — 286 f gravid — 286 f gravid 220 224 m — 235 m 291 297 f gravid 3 full term embryos (all females) Dasyatis hastatus. 851 984 f gravid 4 embryos — Circa 914 f gravid 5 embryos 272 310 f not gravid LENGTH OF DISC. MILLIMETERS 52 Zoologica: New York Zoological Society [XXVI: 10 Text-figure 1. Comparison of length and width of disc in pre-natal and post-natal Dasyatis based on data in Tables I and II. 1941] Breder & Krumholz: On the Uterine Young of Dasyatis sabinus 53 Text-figure 2. Dorsal view of full-term embryos of two species of Dasyatis. 1941] Shlaifer: Respiratory Behavior in Small Tarpon 11. Additional Social and Physiological Aspects of Respiratory Behavior in Small Tarpon. Arthur Shlaifer New York Aquarium & U. S. Fish & Wildlife Service, Woods Hole, Mass. Introduction. Tarpon, Tarpon atlanticus (Cuv. & Val.), whose swimbladders contain considerable al- veolar tissue, periodically rise to the surface to breathe atmospheric oxygen despite their ex- tensive gill surfaces and the relatively high oxygen content of the waters they normally inhabit. The work reported here represents a continu- ation of the program of investigation with small tarpon begun by Shlaifer & Breder (1940). In this earlier work, which, incidentally, is the first experimental report on this fish, it was found, among other things, that the surface rise of a tarpon to gulp air may induce similar rises in others in a group. For instance, in a group of four in a 50-liter aquarium, in 70% of the cases a rise by one tarpon would induce a rise in one or more companions. It was shown that “coinci- dence” was not involved as a disturbing factor and that, apparently, a tarpon had to be in a “physiologically receptive” state by virtue of a depleted oxygen supply before such imitative rises could be induced. In preliminary experi- ments, some success was attained in inducing rises by means of a silver-painted carved wooden tarpon model manipulated so as to simulate the normal rise of this fish. In the present report “model” experiments were considerably ex- tended. Shlaifer & Breder (1940) found that while the locomotor activity of tarpon did not change when the oxygen content of the water was raised from 2.50 cc. per liter to 5.60 cc., the rate of surface (respiratory) rises to gulp air decreased markedly. As a continuation of this line of in- vestigation, experiments were performed to de- termine the effect of waters of distinctly low oxygen content on locomotor activity and surface rises; also, survival time was determined for tarpon whose access to the surface and hence to atmospheric oxygen was cut off. The writer wishes to express his appreciation to Dr. C. M. Breder, Jr., of the New York Aquarium for the experimental animals and to the U. S. Fish and Wildlife Service at whose Woods Hole laboratories the experiments were performed. The Induction of Imitative Respiratory Rises. As discussed above, it has been found (Shlaifer & Breder, 1940) that imitative surface rises in small tarpon may be induced by a ma- nipulated wooden model as well as by other tarpon. Models have been used with and with- out success in various lines of investigation on behavior in fishes. Thus, Lissmann (1932) found that the characteristic fighting reaction of the Siamese fighting fish, Betta splendens, is elicited by plasticine models or paintings of the fish. Noble (1934) working with the dwarf sunfish, Eupomotis gibbosus, and Breder (1936) with the sunfish, Lepomis auritus, found that males would attempt to mate with a variety of objects irrespective of their general appearance provided they were manipulated so as to re- semble somewhat the actions of a female ready to spawn. On the other hand, Spooner (1931) found that the bass, Morone labrax, was not attracted by various rough models of itself. Breder & Coates (1935), investigating the sex recognition of the guppy, Lebistcs reticulatus, found that no reaction was given to carefully made models of female Lebistes with or without movement. Finally, Shlaifer (1940) failed to obtain a reaction to sunfish models by goldfishes. Granted that a properly manipulated wooden model may induce imitative rises in tarpon, will other objects properly manipulated but differing appreciably from a tarpon in shape and color also induce rises and if so to what extent? Will a tarpon model or other objects be effective if manipulated so as to rise in a manner appreciably different from the normal one? In an attempt to answer these questions the experiments de- scribed below were performed. Materials and methods. The animals used in these and other experiments described in this paper were 10-12 cm., presumably young, speci- mens sent from the collection at the New York 56 Zoologica: New York Zoological Society Aquarium but originally obtained along the Florida coast. They were kept in sea water and were fed chopped herring three times a week. The experimental group in the model tests was composed of six tarpon placed in 40 liters of standing sea water in a 48-liter rectangular as- sembled aquarium with transparent glass sides, whose dimensions were 55 cm. by 35 cm. and 25 cm. deep. The water temperature range was 19-22 degrees C. and the average oxygen content was 2.50 cc. per liter. Four different types of object, all 10 cm. in length, were used as “models” in the attempts to induce imitative rises in the experimental animals. These were: 1. A silver-painted carved wooden model of a tarpon having no fins or mouth but possessing an “eye” fashioned out of a thumbtack. 2. A white spatula. 3. A piece of red rubber tubing 1 cm. in diameter. 4. A solid glass rod 4 mm. in diameter. Two pieces of string were attached to these objects at opposite ends. The object to be tested was suspended horizontally in the experimental aquarium con- taining the six tarpon, next to the long side of the aquarium nearest the observer, at a distance of 3.5 cm. from the bottom, the level at which the tarpon in that situation generally swam. The observer, seated about two feet away from the aquarium and facing one long side, manipu- lated the object by means of the two strings. The normal air-gulping rise of a tarpon is a rapid movement lasting a little less than a second from the start of the rise to the return to the normal position. Usually the animal rises to the surface almost vertically. In A of Table I this type of rise was simulated by properly manipulating the objects. In B of Table I the object or model was manipulated so as to produce a distinctly “ab- normal” rise. The object was raised to the sur- face parallel with it rather than at a right angle and the rise from beginning to end lasted two seconds. The data in Table I are listed in terms of successful attempts to induce rises in the tarpon by means of the manipulated objects. Each time the object is raised to the surface is considered an “attempt.” The attempt is con- sidered successful if within one second or less following the start of the rise of the model one or more tarpon rise to the surface. Rises by the fishes two or three seconds after the object had risen might also have been imitative delayed reactions but were not considered. Indeed, such rises very rarely occurred. Either the fish would respond immediately or it would not rise for a minute or more. No effort was made to differ- entiate between induced imitative rises that in- volved varying numbers of tarpon. In fact, such differentiation is very difficult if not impossible since, except in rare cases in which two or more individuals arose at exactly the same tune, one cannot determine whether the rise of the tarpon following very closely that of the first animal to rise is a response to the object or to the first tarpon. Observations were made within a ten-minute experimental period. Every second, fourth, [XXVI: 11 sixth, eighth, and tenth minute on the minute the model was raised to the surface and the reaction noted. As a control, every first, third, fifth, seventh, and ninth minute on the minute an observation was made for rises of the tarpon without the manipulation of an object. Each observation, control and experimental, at the two minute interval lasted only one second. In general, observations were made during two con- secutive hours, or 12 experimental periods each day. Results. The data in A of Table I indicate that all four types of objects induce imitative rises whose statistical significance compared with the controls is very high. The average rise of a tarpon in the experimental aquarium was twice in ten minutes. In a group of six animals there would be at most twelve seconds of rising in 10 X 60 or 600 seconds; thus, according to the laws of chance, rises of tarpon directly following those of models cannot very well be considered coinci- dental. It is seen that, of the four objects used, the wooden tarpon model is significantly most ef- fective and the red rubber tubing is second. The degree of success in inducing imitative rises is, however, sharply limited by the physiological state of the animal. As Shlaifer& Breder (1940) demonstrated, a tarpon will not respond imita- tively to the rise of another object, even if it be an- other normal tarpon, unless, by virtue of the fact that the oxygen obtained at the last gulp is dis- tinctly depleted, it is in a “physiologically recep- tive ” state which permits a rising response. The induction of an imitative rise, then, depends on the reaching of a respiratory threshold by the reacting animal. This explains why even the wooden tarpon model enjoys only 48.3% success in inducing rises. The experimental animals may rise either singly or imitatively in groups just before the model is raised to the surface. If so, even raising the object consecutively three times or more will produce no response. The data in B of Table I demonstrate that objects manipulated so as to rise in an “abnor- mal” way induce few i mitative rises . Comparison with the controls indicates no statistical signifi- cance . Surface Rises in Blinded Tarpon. That the induction of imitative rises in small tarpon is a visual affair was demonstrated by Shlaifer & Breder (1940). It was found that the rise of a tarpon on one side of a transparent glass plate bisecting an aquarium could induce a rise in others on the opposite side. Carrying this a step further, four tarpon were blinded by piercing the cornea. Blindness was ascertained by appropriate tests. Several days were allowed for recovery from the operation. Subsequently, the animals were grouped together in 40 liters of standing sea water (oxygen con- tent— 2.50 cc. per liter) in a 48-liter aquarium whose dimensions have been listed above. In the course of several hours of observation no imitative rises occurred. The tarpon rose 57 1941] Shlaifer: Respiratory Behavior in Small Tarpon Table I. The Induction of Imitative Surface Rises in Small Tarpon by Various Objects. A. Object manipulated so as to simulate a normal surface rise. I Silver II Ill IV V Type of Model Wooden White Red Rubber Glass Control2 Tarpon Spatula Tubing Rod Total Number of Attempts1 176 181 176 170 703 Number of Successful Attempts 85 52 62 35 12 % of Successful Attempts 48.3 28.7 35.2 20.6 1.7 1 Object is raised to surface for one second every two minutes in a 10-minute observation period in an attempt to induce a rise in the experimental tarpon in the aquarium. 2 Observation made for one second every two minutes in the 10-minute period for rises by tarpon without manipu- lation of objects. Statistical significances (number of successful attempts).* I vs. II — .0096 I vs. Ill — .0210 I vs. IV — .0013 II vs. Ill — .4370 II vs. IV— .2898 III vs. IV— .0210 I, II, III, IV vs. V— .0000 * Upper limit of statistical significance is set at 0.05. This is three times the probable error. 0.01 indicates good significance while a value of 0.100 or more indicates little significance. (“Student,” 1925.) B. Manipulation of object does not simulate a normal rise.f Type of Model Total Number of Attempts1 Number of Successful Attempts % of Successful Attempts I Silver Wooden Tarpon 65 5 7.7 II III White Red Rubber Spatula Tubing 65 65 5 1 7.7 1.5 t Object is raised to surface slowly and parallel with the surface. 1 See 1 above. 2 See 2 above. Statistical significances (number of successful attempts). I vs. II — None II vs. Ill— .3834 I vs. Ill — .3342 II vs. IV— .4370 I vs. IV — .4954 III vs. IV — .5580 I vs. V— .1832 III vs. V— .5580 IV Glass Rod 65 2 3.1 V Control2 260 6 2.3 II vs. V — .2146 IV vs. V— .5580 I (A) vs. I (B)— .0000 II (A) vs. II (B)— .0000 III (A) vs. Ill (B)— .0000 IV (A) vs. IV (B)— .0000 V (A) vs. V (B) — .4954 individually at the same rate as did the normal animals, once every five minutes. Thus, it is apparent that in the absence of vision no imi- tative rise will occur; possible response to pres- sure stimuli does not obtain. The normal rate and type of surface rises in these blinded forms indicates that this pattern of behavior is not dependent on vision for its proper execution though imitative response by one fish to another is. Locomotok and Respiratory Activity in Waters of Low Oxygen Content. In previous experiments (Shlaifer & Breder, 1940) it was found that when the oxygen content of the water was increased from 2.50 cc. to 5.60 cc. per liter, the rate of respiratory activity (surface rises) decreased significantly though the locomotor activity remained the same. It be- comes desirable, therefore, to determine the respiratory activity and, to a minor extent, the locomotor activity at very low oxygen levels. Small tarpon, incidentally, are relatively hardy forms and are adaptable to a variety of situations. They endure handling in the laboratory remark- ably well and would be excellent laboratory material for many lines of research were they to be found more frequently and in greater abund- ance. They may be found both in the sea and in land-locked pools which are brackish or fresh- water. Some of these pools have relatively little dissolved oxygen but the tarpon survive. Materials and methods. Tarpon were placed in isolation in 7 liters of standing sea water in 8-liter rectangular battery jars whose dimensions were 20.5 cm. by 18 cm. and 21 cm. deep at the 7-liter mark. The water temperature range was 19-22 degrees C. The rates of surface rises and locomotor activity were determined for 15-minute periods. The rate of locomotor activity was determined by an observational technique which is fully described elsewhere (Shlaifer, 1938). Briefly, 3-crn. squares were ruled in red India ink on all four vertical sides of the rectangular battery jar. Using the eye of the tarpon as an anatomical landmark, the number of squares, actually cubes when projected in space, traversed in a 15-minute period was multiplied by 3 to give the results in centimeters. Since this tech- nique is not as accurate as the simple observation of surface rises and since the experimental vessels, considering the size of the tarpon, were quite small and thus limited activity, the rate of surface rises in this set of experiments may be considered the more important feature. Oxygen content was determined by means of the permanganate modification of the Winkler method. Results. Conparing the data in Table II with the results of Shlaifer & Breder (1940), it is seen that at the lower oxygen content the rate 58 [XXVI: 11 Zuologica: New York Zoological Society of respiratory rises increases significantly and the centimeters traversed per rise decreases. The significantly lower rate of locomotor activity at the lower oxygen level may be due primarily to the distinctly smaller vessel in which the tarpon were kept in these experiments rather than to the low oxygen content. As previously mentioned, Shlaifer & Breder (1940) found that while raising the oxygen content from 2.50 to 5.60 cc. per liter in the same aquarium reduced the rate of respiratory rises it did not affect activity. At any rate, the number of surface rises increases at the lower oxygen content de- spite the decreased activity. ously listed. In the first four experiments standing sea water was used; in the fifth, sixth, and seventh, running sea water. In all cases the battery jars were filled to capacity. In the standing water series, a fine wire screen cut so as to insure a tight fit was placed inside the battery jar at the 7-liter mark and in Expts. 5-7 a similarly tight-fitting wooden screen was placed at the same level. In each test the experimental tarpon was allowed a 48-hour period of acclima- tization in the battery jar with access to the surface. At the end of that time the screen was put in place and the time and oxygen content noted. Each battery jar contained only one Table II. The Locomotor Activity and Respiratory Rises of Small Tarpon in Waters of Low Oxygen Content. A. This set of Experiments B. Shlaifer & Breder (1940)3 Mean Oxygen Mean Locomotor Content cc. /liter Activity1 Mean Respira- tory Rises2 Mean cm. per Rise 1.02 304.0 3.6 84.4 2.40 480.0 2.9 165.5 1 Expressed in centimeters traversed per fish per 15-minute period of observation. Figure given is the mean of 80 such periods. 2 Expressed in surface rises per fish per 15-minute period of observation. Figure given is the mean of 80 such periods. 3 Based on Table I (AI), Shlaifer & Breder (1940). Figures given are for an isolated tarpon in 48 liters of sea water in a 50-liter aquarium. Figures for locomotor activity and respiratory rises represent the mean of 48 15-minute observation periods, Statistical Significances (A vs. B) Locomotor Activity: .0003 Respiratery Rises: .007 Finally, it may be noted that if access to the surface is provided, tarpon will survive well in waters of low oxygen content and will utilize relatively more atmospheric oxygen. Survival with No Access to the Surface. The air-breathing habit in tarpon is peculiar in view of the fact that they normally inhabit fairly highly oxygenated sea water and have rather extensive gill surfaces. The value of the air-breathing habit in land-locked pools of low- oxygen content in which tarpon may be found is obvious. However, its persistence in the w-aters of the open sea may possibly be an indication of imperfectly functioning gills, despite their size. It becomes of interest, therefore, to determine the survival time of tarpon when cut off from access to the surface and thus forced to rely on dissolved oxygen. Hora (1933) states that various air-breathing swamp fishes can be “drowned” if prevented from reaching the water surface. Das (1935) states that the air-breathing loach, Lepidocephalus guntea, if prevented from reaching the surface, will be asphyxiated in a little more than an hour. These fishes, however, inhabit waters in which the oxygen content is habitually low in contrast to the tarpon of the open sea. Materials and Methods. In Expts. 1-7 of Table III, the experimental vessels were 8-liter battery jars whose dimensions have been previ- individual at a time. A record was kept of the number of hours between the start of the experi- mental period and the end, which in every case was at the death of the tarpon. Oxygen de- terminations were made each day and at the conclusion of the experiment. The locomotor activity and surface rising of the fish were also noted but only in a general way. As controls, individual tarpon were kept in battery jars in running and standing sea water without being screened from the surface. In Expts. 8-13, the experimental vessel was a 56-liter rectangular assembled aquarium with transparent glass sides whose dimensions w-ere 60 cm. by 35 cm. and 27 cm. deep. This was kept constantly filled to capacity with running sea water. The bottom of the aquarium was covered with sand. Into this aquarium was placed a tight-fitting wire mesh screen containing three squares to the inch, at a level of 6 cm. above the bottom sand. The general procedure followed was the same as that described for the battery jar tests except that no observations were made for activity or surface rises. Results. The data in Table III demonstrate that compelling a tarpon to rely on dissolved oxygen by screening it from the surface will re- sult in asphyxiation in from 7 to 128 hours. Death may occur in a relatively short time even in running sea water (Expt. 5). In the standing water tests the oxygen content at the death of 1941] 59 Shlaifer: Respiratory Behavior in Small Tarpon the animal is high enough to maintain a tarpon indefinitely if it is permitted to respire atmos- pheric oxygen. There seems to be considerable individual variation in survival time. This may be a reflection of individual variations in meta- bolic rate. In the battery jar tests the tarpon would often make desperate and repeated futile attempts to reach the surface and, failing, would sink exhausted to the bottom, there to remain quiescent for long periods of time. In the aquari- um tests, much more area was allowed for normal swimming by the fishes; also, the position of the screen, 6 cm. above the bottom, prevented exhausting and futile attempts to reach the sur- face. Tarpon in this aquarium, even when not screened from the surface, normally swim near the bottom so that the experimental situation in this case is, except for the screen, quite normal. In the control tests, tarpon in battery jars with access to the surface survived indefinitely even when the standing water oxygen content was quite low. Discussion. It is interesting to note in the model experi- ments that imitative rises may be induced by very crude objects differing appreciably in form and color. At the same time, a more finished wooden model is more successful than are the crude ones. The results in B of Table I indicate that movement rather form or color is the im- portant factor for, if the model is so manipulated that the normal rise of a tarpon is not simulated, there is no significant induced imitation, even with the wooden model. Nevertheless, a properly ma- nipulated silver wooden model is more successful than are properly manipulated objects that are distinctly cruder models. Apparently, then, form and/or color may also be involved but only in a minor supplementary way. Possibly at a distance of several inches or more from the re- acting tarpon the type of wooden model used was more visible than were the other objects. Using this imitative rising reaction, it may be Table III. Effect of Prevention of Surface Rises on the Survival of Small Tarpon. Remarks Periods of great activity and attempts to reach surface followed by periods of quiescence. U Tarpon very active — many attempts to reach surface. Tarpon very active. Activity and rises of fish increase markedly one day before death. Experi- ment No. Water Oxygen Content* Survival in Hours I1 Standing 3.39-2.20-1.60 53 21 Standing 3.47-2.98-2.70-2.52 71 31 Standing 4.15-3.20-1.68 51 41 Standing 4.07-1.92 11 5l Running 4.86 7 61 Running 5.58 72 71 Running 5.58 128 8= Running 5.58 32 92 Running 5.58 115 102 Running 5.58 125 ll2 Running 5.58 24 122 Running 5.58 34 132 Running 5.58 46 * In the standing water experiments, the oxygen content represents the oxygen content of the medium at the death of the experimental animal. In the running water series, the figures given represent the average oxygen content in cc. per liter day by day. 1 Experimental vessel is an 8-liter battery jar. 2 Experimental vessel is a 56-liter aquarium. In several cases attempts were made to revive tarpon which, after a long period of time, were near asphyxiation and lay on their sides. They were permitted to come in contact with the surface but evidently were too far gone, for they soon died. In one case, however, three blinded tarpon, which when screened from the surface in standing water in a battery jar soon approached exhaustion and asphyxiation, did recover when the screen was removed after one hour. In that time, unlike normal tarpon, they kept rising al- most continually in attempts to penetrate a screen which they could feel but could not see and within 40 minutes lay on their sides. Das (1935) reports that the air-breathing loach will recover from partial asphyxiation if permitted to rise to the surface after being kept under a screen. possible, by employing a great variety of objects differing in form and color but manipulated normally, to obtain data which may be of sig- nificance in obtaining a measure of the visual acuity of tarpon. The data already obtained, showing as they do only a slight though real difference between a good and a crude model in their relative success at inducing rises, tend to indicate that such experiments would require considerable data to have any significance. It has been seen that preventing a tarpon from utilizing atmospheric oxygen will eventually be fatal, even in highly oxygenated water. Hence, there is no question of the survival value of the surface rise pattern of behavior in these forms. In contrast to the work of Hora (1933) and Das 60 Zoologica: New York Zoological Society (1935), the asphyxiation of tarpon under screens is a very slow process. Undoubtedly there is interesting material here for exhaustive physi- ological studies. It is quite difficult to evaluate the pattern of imitation in surface rising in a group of tarpon. Shlaifer & Breder (1940) have shown that imi- tation will not occur, apparently, until a respira- tory threshold is reached. An isolated tarpon will rise periodically in response to physiological need and over a period of time not much more frequently if it is in a group. It is difficult, at least at this time, to attribute any adaptive value to this imitative behavior which might be termed “social respiratory facilitation.” Summary. 1. Imitative air-gulping surface rises in small [XXVI: 11 tarpon may be induced by relatively crude models if they are manipulated so as to simulate the normal surface rise of this form. A relatively more life-like wooden tarpon model is more successful than are the cruder objects. No signifi- cant success is attained with any model if its ma- nipulated rise differs appreciably from the normal. 2. The imitative pattern is based on visual stimuli rather than on differential pressure stimuli. Blinded tarpon never rise imitatively. 3. At distinctly low oxygen levels, the rate of respiratory rises increases but if access to the surface and hence to atmospheric oxygen is maintained the tarpon survive. 4. Tarpon when prevented from utilizing atmospheric oxygen succumb in 7 to 128 hours even in highly oxygenated running sea water. Literature Cited. Breder, C. M., Jr. 1936. The reproductive habits of the North American sunfishes, Centrarchidae. Zoo- logica, 21: 1-48. Breder, C. M., Jr., & Coates, C. W. 1935. Sex recognition in the guppy, Lebistes reticulatus Peters. Zoologica, 19: 187-207. Das, B. 1935. On ecology and bionomics of an air- breathing loach, Lepidocephalus guntea (Ham. Buch.), with a review on air- breathing fishes. Proc. Xllth Intern. Congr. Zool., Lisbon, 2: 866-920. Hora, S. L. 1933. Respiration in fishes. Jour. Bombay Nat. Hist. Soc., 36: 538-560. Lissmann, H. W. 1932. Die Umwelt des Kampffisches ( Betta splendens Regan). Zeitschr. Vergleich. Physiol., 18: 65-111. Noble, G. K. 1934. Sex recognition in the sunfish, Eupomotis gibbosus (Linn6). Copeia, (4): 151-155. Shlaifer, A. 1938. Studies in mass physiology: effect of numbers upon the oxygen consumption and locomotor activity of Carassius auratus. Physiol. Zool., 11: 408-424. 1940. The locomotor activity of the goldfish, Carassius auratus L., under various con- ditions of homotypic and heterotypic grouping. Ecology, 21: 488-500. Shlaifer, A. & Breder, C. M., Jr. 1940. Social and respiratory behavior of small tarpon. Zoologica, 25 : 493-512. Spooner, G. M. 1931. Some observations on schooling in fish. Jour. Marine Biol. Assoc. United Kingdom, 17: 421-448. 1941] Smith: Notes on Mexican Snakes of the Genus Trimeresurus 61 12. Notes on Mexican Snakes of the Genus Trimeresurus. Hobart M. Smith. A review of the species of Trimeresurus now known from Mexico has revealed some confusion in the literature, to which I personally have contributed to some extent. In addition to correction of some of these errors, the present paper presents two forms new to the fauna of Mexico, names a population previously associ- ated with lansbergii, and diagnoses in a key the eleven forms definitely known to occur in the country. It should be noted that schlegelii, treated by Terron (Anal. Inst. Biol. Mex., vol. 1, 1930, pp. 196-7, fig. 7) as a member of the Mexi- can fauna (without locality), has never been reported nearer Mexico than the “Guatemala” record of Boulenger (Cat. Snakes, vol. 3, 1896, p. 567). Its existence in the country is prob- lematical. I am much indebted to Dr. E. R. Dunn, who has very generously checked the types of brachy- stoma and made a number of valuable additions to the paper. The study has been completed, and a portion of the material on which it is based has been collected, during tenure of a Walter Rathbone Bacon Traveling Scholarship. Trimeresurus godmani (Gunther). Bothreichis godmanni Gunther, Ann. Mag. Nat. Hist., ser. 3, vol. 12, 1863, pp. 364-365, pi. 6, fig. G (Totonicapam, Guatemala). Bothrops godmani Martin del Campo, Anal. Inst. Biol. Mex., vol. 9, 1938, p. 229, fig. 2. One specimen from Santa Rosa, near Comitiin, Chiapas, was reported by Martin del Campo. It has 142 ventrals, 33 caudals, dorsals in 21 rows, 9-10 supralabials. I have examined six specimens, all from Guatemala and Costa Rica. Trimeresurus barbouri (Dunn). Lachesis barbouri Dunn, Proc. Biol. Soc. Wash., vol. 32, 1919, pp. 213-214 (Omilteme, Guerrero) . Published by permission of the Secretary of the Smith- sonian Institution. Agkistrodon browni Shreve, Copeia, 1938, p. 9 (Omilteme, Guerrero). This diminutive species is known only from three specimens collected at the type locality, of which I have examined one, the type of barbouri (U. S. N. M. 46347). I cannot see that browni is different from barbouri. The apparent difference in ventral counts of the types (134, browni; 154, barbouri) is not real, since the browni type is a male, and the barbouri type a female, the ventrals of which should have been written 145 (I count 144 to 147, according to the number of smaller anterior scales counted as ventrals) . The only remaining apparent difference is in the presence of larger head shields, with distinct prefrontals, frontal and parietal in browni. I believe the size of the dorsal head scales varies in this species, as it certainly does in godmani, to which the Omilteme species is related. In six specimens of godmani several intermediate stages between more or less uniform, small scales and the presence of distinct frontals and parietals are shown. Since the type of barbouri has small but distinguishable prefrontals, and since the presence or absence (and size when present) of the frontal and parie- tals is variable in a closely related species, I be- lieve there is no reasonable doubt that browni and barbouri are synonymous. Trimeresurus bicolor (Bocourt). Bothrops bicolor Bocourt, Ann. Sci. Nat., ser. 5, vol. 10, 1868, p. 201 (San Agustin, Guatemala). The only specimen examined is one in the National Museum (No. 46511), from Chicharras, Chiapas. It is a small female with 168 ventrals, 48+ caudals (tail tip missing), 10-11 supralabi- als, 12-12 infralabials, and 10 scales between the narrow supraoculars ; second supralabial enters pit on one side, does not on other. The species is closely related to lateralis Peters (1852), of Costa Rica, which has larger head scales (7 or 8 between supraoculars), a yellow line along outer row of scales and sometimes short black-and-yellow cross-bars (fide Boulen- 62 Zoologica: New York Zoological Society [XXVI: 12 ger). There is no difference between the two in number of labials (the single specimen examined of lateralis has ten), nor in exclusion of the second supralabial from the pit (see above). Trimeresurus nigroviridis aurifer (Salvin). Thamnocenchris aurifer Salvin, Proc. Zool. Soc. London, 1860, p. 459, pi. 32, fig. 1 (Cobdn, Guatemala). Bothrops nigroviridis aurifer a Barbour and Loveridge, Bull. Antiv. Inst., vol. 3, 1929, pp. 1-3; Martin del Campo, Anal. Inst. Biol. Mex., vol. 9, 1938, pp. 228-229, fig. 1. This species was first reported from Mexico by Martin del Campo, who cites a specimen from Santa Rosa, near Comitdn, Chiapas. It has 159 ventrals, 44 caudals (lower than other aurifer reported in the literature), 9-10 supralabials, scales in 19 rows. I have seen no specimens. Trimeresurus atrox (Linnaeus). Coluber atrox Linnaeus, Syst. Nat., 1758, p. 22 (Asia, erroneous). Trimeresurus atrox Schmidt and Andrews, Zool. Ser. Field Mus. Nat. Hist., vol. 20, 1936, p. 182. Specimens of this species have been taken as far north as Valles, San Luis Potosi (Martin del Campo, loc. cit., records it from Tamaulipas) ; it occurs throughout the Atlantic coastal region south of this point, and also along the Pacific coast in southeastern Chiapas. Boulenger records it from farther north on the Pacific coast, from Atoyac, Guerrero (Cat. Snakes, vol. 3, 1896, p. 536). The nineteen Mexican specimens examined differ from South American specimens in the character of the carinae of the median scales, as pointed out by Boulenger (op. cit., pp. 535-539). I can find no other difference. Trimeresurus nummifer nummifer (Ruppell). Alropos nummifer Ruppell, Verz. Senck. Mus., Amph., 1845, p. 21 (Mexico). Trimeresurus nummifer nummifer Dunn, Proc. Biol. Soc. Wash., vol. 52, 1939, pp. 165-166. Specimens are known from central Veracruz along Atlantic slopes into Central America, and on Pacific slopes in southeastern Chiapas. The species is rarer than atrox, and is apparently confined to hilly regions, while atrox is more widespread and ranges from hills to plains. I have examined thirteen Mexican specimens. Trimeresurus nasutus (Bocourt). Bothrops nasutus Bocourt, Ann. Sci. Nat., ser. 5, vol. 10, 1868, p. 202 (Panzos, Rfo Polochic, Guatemala). One specimen (U. S. N. M. 110415) is known from the Mexico-Guatemala border, bearing the locality data Piedras Negras, Guatemala. Dr. E. R. Dunn tells me that there is one in the Museum of the Philadelphia Academy of Natural Sciences, labelled Veracruz (No. 4873, collected by Rev. II. F. Heyde). This is quite different from dunni and yucatanicus, all of which are related; nasutus has a much higher rostral and a different arrangement of the pre- ocular and subocular scales. Two specimens have been examined, the other from Panamd. Trimeresurus dunni Hartweg & Oliver. Trimeresurus dunni Hartweg & Oliver, Occ. Papers Mus. Zool. Univ. Mich., no. 390, 1938, pp. 6-7, pi. 1 (Tehuantepec, Oaxaca). Known from the semi-arid Pacific slopes of the Isthmus of Tehuantepec (seventeen specimens examined). The definition of this species by Hartweg & Oliver has greatly clarified the situation with respect to lansbergii, with which the Tehuantepec as well as Yucatdn specimens were previously associated. Prior to their action, three separate populations of lansbergii were apparent: one in northern South America and in Panamd; one in Oaxaca; and the third in Yucatan. Between Mexico and Panama no specimens related to lansbergii are known. All of this group from that intermediate area are referable either to ophryo- megas or to nasutus (c. f. Amaral, Bull. Antiv. Inst., vol. 3, 1929, pp. 19-27). 1 The isolated group in Yucatan is not the same as dunni, nor can it be referred to typical lans- bergii of South America. I propose the name Trimeresurus yucatanicus sp. nov. Ilolotype. U. S. National Museum 46571, fe- male, from Chichen Itza, Yucatan, collected by Nelson & Goldman, February, 1901. Paratypes. Field Museum of Natural History No. 504, Yucatan, and No. 20621, Chichen Itza, Yucatdn. Diagnosis. Snout turned up in front; rostral no more than one and one-half times as high as wide; scale rows 21 posteriorly; two lower pre- oculars subequal, both excluded from orbit; two large scutes on top of head bordering internasals and canthals, nearly meeting medially; loreal square; caudals 32 to 41 in females; bands on body single. Description of Holotype. Rostral very high, its length (2.6 mm.) somewhat greater than greatest width (2.2 mm.), twice width of upper part (1.3 mm.); internasals elongate, elevated, in con- tact medially; one canthal; preocular encroach- ing on dorsal surface between canthal and supra- ocular; five or six scales between supraoculars; dorsal head scales keeled ; a large scale, as large as canthal, bordering internasal medially, in contact with canthal, separated from its mate by one scale; ten supralabials; first labial in contact with anterior section of nasal, separated from posterior section by a very small scale and a small, wedge- 1 Trimeresurus lansbergii annectens Schmidt is certainly referable to ophryomegas, of which it may be a subspecies. 1941] Smith: Notes on Mexican Snakes of the Genus Trimeresurus 63 shaped protrusion from anterior section; scale between second labial and posterior section of nasal quite small; one row of small scales between labials and border of pit ; two tiny scales between nasal and border of pit; loreal moderately large, nearly square; one moderately large scale be- tween lower preocular and third labial; one row of scales between subocular and labials ; subocular single, very elongate; one very large upper pre- ocular, about twice as large as loreal, somewhat smaller than canthal ; two lower preoculars, both separated from labial border, both small; median preocular partially fused with an elongate scale bordering edge of pit; edge of supraocular thin; edge of canthal somewhat raised, rather sharp anteriorly; edge of preocular not keeled, rounded. Eleven or twelve infralabials; first infralabials in contact with each other behind mental; three labials in contact with chinshields; one large anterior pair of chinshields, about as long as combined width; this followed by two small pairs of scales. Dorsals in 27-27-21 rows; 148 ventrals; 32 caudals, entire; anal entire. Total length 254 mm.; tail 26 mm. Specimen badly faded. Ground color light gray, stippled; seventeen somewhat staggered, darker cross bands, split on middorsal line; blotches covering about six scale lengths, sepa- rated by light areas covering about four scale lengths; anterior and posterior edges of cross bands black, narrowly light-edged ; some blotches faintly divided by a very slightly lighter, trans- verse area; the dark borders terminating about four scale rows lateral to vertebral row ; blotches faintly interrupted at this point, below which are two rounded spots of same color as blotches, separated from each other, and more or less fused with the blotches; belly heavily stippled; a few light marks in labial region. Remarks. So far as now known, the only speci- mens of Trimeresurus from areas north of Panama having a rostral similar to that of lansbergii are those from the Tehuantepec area which have been described by Hartweg & Oliver as dunni, and the present three specimens from YucaUin. The latter three are not the same as dunni, since the bands are not paired; the scale rows usually more numerous in front of anus (21 at anus in one out of 27 dunni, 21 in all three yucatanicus) ; two lower preoculars small, sub- equal, excluded from orbit (middle larger, entering orbit in dunni). From lansbergii itself the present species ap- parently differs in having 21 scale rows posteri- orly, usually more numerous subcaudals (28 to 33 in females, 31 to 36 in males of lansbergii; 32 to 41 in females of yucatanicus)-, two relatively large scutes bordering internasals and canthals, nearly meeting each other medially (no such scales in lansbergii, apparently); loreal square (usually narrow, about twice as long as broad in lansbergii). In 1859 Cope described a form called brachy- stoma ( Proc . Acad. Nat. Sci. Phila., 1859, p. 339). The series on which this name was based is now in the collections of the Philadelphia Academy of Natural Sciences, and consists of four specimens (of which No. 7043 is the type), all collected by Mr. Cuming at unknown localities. All four have 19 scale rows posteriorly, elongate loreals and bands single, and accordingly agree with the characters of lansbergii. They probably were secured in South America, as Dr. Dunn states that numerous other specimens collected by Cuming are in the Academy collections, all apparently from South America. Trimeresurus undulatus (Jan). Trigonocephalus ( Atropos ) undulatus Jan, Rev. Mag. Zool., 1859, p. 157 (Mexico). Specimens in the National Museum are from Orizaba, Veracruz (No. 6319); Omilteme, Guer- rero (Nos. 46345-6, 46348) ; and Oaxaca, Oaxaca (No. 46466). The species has also been reported from the state of Hidalgo (Martin del Campo, op. cit., p. 14). Trimeresurus tnelanurus Muller. Trimeresurus melanurus Muller, Mitt. Zool. Mus. Berlin, vol. 11, pt. 1, 1923, pp. 92-93 (Mexico). Trimeresurus garciai Smith, Proc. Biol. Soc. Wash., vol. 53, 1940, pp. 62-64, fig. 2 (Cacaloa- pam, Puebla). This species is known only from the desert region about Tehuacan, Puebla. Its divided caudals and the horn-like supraocular distin- guish it from all other species in Mexico. Three specimens have been examined. Ivet to Mexican Trimeresurus 1. Supraocular produced as a soft, horn-like scute 2 Supraocular flat 3 2. Subcaudals entire melanurus Subcaudals divided undulatus 3. Snout produced, turned up 4 Snout not noticeably turned up 6 4. Rostral about twice as high as wide. . . .nasutus Rostral no more than one and one-half times as high as wide 5 5. Bands on body distinctly paired; two lower preoculars small, subequal, excluded from border of orbit; scale rows 21 in front of anus yucatanicus Bands on body paired, each half well defined and usually separated medially from its mate; middle preocular considerably larger than lower, broadly in contact with border of orbit; scale rows usually 19 in front of anus dunni 6. Subcaudals double atrox Subcaudals single 7 7. Scales in 23 rows or more; size large nummifer nummifer Scales in 21 rows or less; size smaller 8 8. Width of a supraocular about half distance between supraoculars; subcaudals 22 to 34; brown; terrestrial 9 Width of a supraocular one third distance between supraoculars, or less; subcaudals 44 to 67; green; arboreal 10 64 Zoologica: New York Zoological Society 9. Scales in 21 rows; upper preocular large, forming part of canthal ridge; loreal small, excluded from canthal ridge, not in contact with supraocular godmani Scales in 19 or 17 rows; upper preocular small, excluded from canthal ridge; loreal large, forming part of with supraocular. 10. Head scales smooth; Head scales keeled; [XXVI: 12 canthal ridge, in contact barbouri ; scales in 19 rows nigrovirulis aurifer scales in 21 rows . . tricolor ZOOLOGIGA SCIENTIFIC CONTRIBUTIONS OF THE NEW YORK ZOOLOGICAL SOCIETY VOLUME XXVI Part 2 Numbers 13-18 PUBLISHED BY THE SOCIETY THE ZOOLOGICAL PARK, NEW YORK September 8, 1941 CONTENTS PAGE 13. The Life History and Bionomics of the Trematode, Zygocotyle lunata (Paramphistomidae). By Charles H. Willey. (Plates I-IV) 65 14. Eastern Pacific Expeditions of the New York Zoological Society. XXIV. Fishes from the Tropical Eastern Pacific. [From Cedros Island, Lower California, South to the Galapagos Islands and Northern Peru.] Part 1. Lancelets and Hag-fishes. By William Beebe & John Tee-Van. (Text-figures 1 & 2) . . 89 15. Eastern Pacific Expeditions of the New York Zoological Society. XXV. Fishes from the Tropical Eastern Pacific. [From Cedros Island, Lower California, South to the Galapagos Islands and Northern Peru.] Part 2. Sharks. By William Beebe & John Tee-Van. (Plates I & II; Text-figures 1-34) 93 16. Correlations Between Structural Eye Defects and Behavior in the Mexican Blind Characin. By Charles M. Breder, Jr., & Edward B. Gresser. (Plates I-IV ; Text-figures 1 & 2) 123 17. On the Species of Otus scops. By Jean Delacour 133 18. A New Genus of Kaleege Pheasants. By Marquess Hachisuka. 143 1941] Willey: The Life History and Bionomics of Zygocotyle luiiata 65 13. The Life History and Bionomics of the Trematode, Zygocotyle lunata (Paramphistomidae). Charles H. Willey Department of Biology, University College of Arts and Pure Science, New York University (Plates I-IV). Zygocotyle lunata belongs in the family Param- phistomidae Fischoeder, 1901. Diesing (1836) described the species from material collected by Natterer, in Brazil, from the ceca of several species of water birds and from the cecum of the deer, Cervus dichotomus, and named it Amphi- stoma lunatum. Fischoeder (1903) redescribed the species from 4 whole mounts of Diesing’s original material in the Vienna Museum. Stunkard (1917) erected the genus Zygocotyle, to contain A. lunatum and a new species, Z. ceratosa, described by him from the intestine of a duck, Anas platy- rhynchos, from Nebraska. Price (1928) showed that Z. ceratosa was specifically identical with Z. lunata. The host relationships of the genus Zygocotyle are of considerable interest. Diesing (1836) de- scribed it from the cecum of a deer, Cervus dichotomus, and also from the ceca of the birds, Anas melanotus, A. epecutiri and Himantopus wilsonii. Fischoeder (1903), after an examina- tion of 4 of Diesing’s specimens, concluded that the record of the ruminant Cervus dichotomus as a host of this form was probably due to an error in labelling. As pointed out by Price (1928), Dujardin (1845) and Diesing (i850) had earlier arrived at the same conclusion because of the wide difference in hosts. The suspicion of error persisted until the occurrence of Zygocotyle in both birds and mammals was definitely shown by Price (1928) who studied specimens of an amphi- stome reported by Hall (1927) from the cecum of a cow, Bos taurus, and identified them as Zygoco- tyle lunata, thus confirming and validating Diesing’s original record. In the present study, infections with Z. lunata have been experimen- tally produced in the sheep, Ovis aries, and in the rat, Mus norvegicus, as well as in ducks. Thus a rodent as well as another ruminant is here added to the list of mammalian hosts. Ml the rats exposed '(more than 60) became infected. Several years ago the writer (1930, 1936) described a new species of cercaria, C. poconensis, from Helisoma antrosum collected near Henry- ville, Pennsylvania. Life history studies with the material in an attempt to identify the adult stage failed, and the investigation had to be abandoned because the infestation was no longer present in the snails from that region. In Septem- ber, 1937, one specimen of Helisoma antrosum from Prospect Park Lake in New York showed an infestation with Cercaria poconensis, and subse- quent collection of 450 snails from the same lake during October yielded 3 additional infections. Cercariae from these 4 naturally infested snails encysted in dishes in the laboratory and furn- ished the material for feeding experiments. Metacercariae fed to rats, ducks and sheep devel- oped into adults of Zygocotyle lunata in the ceca of these hosts. Eggs, collected from feces, developed in the laboratory and young labora- tory-raised snails of the species Helisoma antro- sum, on exposure to the hatched miracidia, be- came infected, yielding the sporocyst, redial and cercarial stages, thus demonstrating experimen- tally the complete life cycle. Preliminary reports were presented in abstract form (Willey, 1937, 1938). Gower (1938) in a study of trematodes infesting wild ducks in Michigan obtained an experimental infestation in a half-grown mallard duck of 46 immature specimens of Zygocotyle lunata with metacercariae from a naturally infected snail identified as Helisoma trivolvis. Much progress has been made in recent years in the demonstration of North American trema- tode life cycles. Most of the advances have been made, however, in groups other than the amphi- stomes. Life cycle studies in this group, with the exception of that of Bennett (1936) on Cotylo- phoron cotylophorum, are mostly incomplete, since they lack one or more stages or fail to pro- vide experimental proof in support of conclusions based on morphological similarity. The demonstration of a complete life cycle requires that all stages be described and obtained experimentally. The studies may begin with any step in the sequence. In trematode studies either of two modes of attack are commonly em- ployed. One may start with the cercarial stage, 66 Zoologica: New York Zoological Society [XXVI: 13 obtained from naturally infected snails, and expose various possible hosts to it or to the metacercaria, if the cercaria encysts in the open. Such experimental hosts must be known to be free from previous infection which might be confused with the experimental one. Frequently, the structure and activities of the cercariae to- gether with information on the fauna occurring in the same environment will give some indication of the next host. Similarly, one might start with the adult stage of the trematode in the final host. Eggs are then collected and permitted to develop and various species of snails are then exposed to the hatched miracidia. Successful infestation provides the asexual stages in the snail. In either method, the cycle is followed back to the starting point. In the present study the investigation began with 4 naturally infected snails and, as described later, infection was obtained in the final hosts. The rats, all of which became infected, had been laboratory-raised for many generations and the ducks were known to be previously uninfected, because none in a control series of ducks, being used for other experiments, ever showed any eggs or specimens of Zygocotyle lunata, while all the experimentally-fed ducks became infected with the species. The complete life cycle was passed through several times in the laboratory. No further collections of naturally infected snails were made since ample material was available in the laboratory throughout the investigation. The life cycle as here presented begins with a descrip- tion of the eggs obtained from experimentally infected hosts, after which the succeeding generations are considered in the order of their occurrence. The materials and methods em- ployed with the various stages are described separately for each stage. Egg. The eggs of Zygocotyle limata examined alive from feces or from living unstained worms are colorless. The shell is very delicate and as studied in optical sections under high magnifica- tion, it does not exceed 1.5 p in thickness. If pushed about with a blunt needle the eggs are easily dented and burst readily with rough handling (Fig. 1). They are broadly ovoid and show little variation in shape (Fig. 2). The slightly attenuated end bears an operculum measuring from 28 p to 31 p in diameter. The edges of the operculum are irregularly notched and these notches interdigitate closely with similar irregular notches on the edge of the shell. A slight excrescence, from 10 to 15 p in diameter, occurs on the shells of all the eggs at the broad end. The freshly deposited egg contains a large number of granular yolk masses suspended in a transparent fluid; and the ovum, measuring from 20 m to 25 m in diameter, still unsegmented, lies embedded between the vitelline masses in the opercular half of the egg (Fig. 1). Eggs of Z. lunata vary much in size. Several hundred eggs collected from feces of ducks and rats infected with newly matured worms, as well as eggs from infestations of more than a year’s duration, were measured. Eggs from feces of a duck and of a rat, each of which had been infested from a single feeding of metacercariae 13 months previously, show no significant difference in size. The eggs from the duck varied in length from 132 p to 152 n and those from the rat from 132 p to 158 p. Width varied in eggs of the duck para- sites from 92 p to 102 p and in those from the rat from 89 p to 102 p. The average size of 38 eggs deposited by worms in the duck was 142 p by 98 p, while that of the 38 eggs from the rat feces measured 141 p by 96 p. The variation in size of eggs from these 13-mont.h-old worms was 26 p in length and 13 p in width. Bennett (1936) showed a similar size range (30 p in length and 10 p in width) in eggs of Cotrjlophoron cotylo- phorum. Fifty-six eggs of young, newly mature Z. lunata collected in a fecal sample from a duck experimentally infected 48 days before, varied in length from 132 p to 152 p and in width from 99 p to 102 p, the average measurements for the group being 145 p by 99.5 p. Similarly, 54 eggs of young, newly mature worms collected in a fecal sample from a rat infected 49 days before, varied in length from 132 p to 158 p and in width from 92 p to 105 p, the average measurements for this group being 141 p by 98.8 p. As with eggs from older infestations, no significant difference in size ap- pears between eggs of worms in the 2 hosts. Further, it may be seen that the size range is the same for eggs from new and old infestations. While in the eggs measured in these series, the average length and width of eggs from young infestations slightly exceed those of older infesta- tions, the difference is not significant. Thirty eggs measured in stained whole mounts of newly matured worms varied from 132 by 86 p to 152 by 99 p, the average being 143 by 91 p. While the average width (91 p) of these eggs in stained and mounted young specimens is about 5 p less than that of living eggs from older worms, the extremes in both dimensions are practically the same in eggs from fixed and stained specimens of all ages as in eggs from feces. The average obtained for all eggs of Z. lunata measured in the present investigation is 142 p in length and 96 p in width. Price (1928) tabulated measurements of Z. lunata from different hosts and showed a re- markably wide variation in egg size. The eggs measured by Price and others from 7 host species vary from 124 to 153 p in length and from 72 to 96 p in width, the extreme range of variation being 29 p in length and 24 p in width. This range is somewhat wider than in the present study. All of the material used by the present author for egg measurements was obtained from experimental infestations with metacercariae from 2 naturally infected snails collected from the same small pond. The hosts yielding the specimens of Z. lunata for egg measurements as tabulated by Price were natural infections studied under different conditions by different workers and collected from widely separated parts of 1941] Willey: The Life History and Bionomics of Zygocotyle lunata 67 North and South America. The differences in egg size may possibly be explained by this difference in source of material. With regard to egg size in amphistomes, it is difficult to draw any conclusion of value other than that egg size is of little importance in classification within a family. Miracidium. Development. Studies on development of the miracidium of Zygocotyle lunata were carried out on eggs obtained from feces of experimentally infected ducks and rats. All eggs observed were in the one-cell stage at the time of deposition. Much difficulty, due to bacteria and molds, was experienced in obtaining successful development of the ovum. The best results were obtained by keeping the eggs in petri dishes and changing the water twice daily. Immediately after collection from feces, the eggs were washed through 8 or 10 changes of tap water. Other methods involving (1) the use of sterile water, (2) gentle mechanical agitation and (3) a method by which drops of water fell continuously into dishes of water containing eggs were all found to be unsatis- factory. Eggs allowed to develop at a constant temperature of 25° C. in an incubator gave some- what better results than those exposed to varying temperatures in the laboratory. By transferring the eggs twice daily with a micropipette to dishes of clean water, about 50 to 60 per cent of the embryos hatched and even then it was necessary to clean the mold from some of them with dissecting needles before they would hatch. In no case was it possible to obtain complete development without some mold forming on the eggs. Bennett (1936) described in considerable detail the developmental stages of the miracidia of Cotylophoron cotylophorum and found a re- markable similarity in the sequence of organ development between that species and develop- ment in other species of trematodes as described by Thomas (1883), Looss (1892, 1896), Ortmann (1908), Johnson (1920), Stunkard (1923), Barlow (1925), Ishii (1934) and Suzuki (1931). Bennett (1936) says, “The development of the miracidium as described here coincides in practically every detail with the development of the miracidia described by the workers mentioned earlier in this discussion. This result points to the con- clusion that the chronological sequence of organ development in trematode miracidia is essentially the same.” In view of Bennett’s complete account of development in C. cotylophorum, a closely related amphistome, and the similarity manifested in the developmental history of miracidia in trema- todes generally, detailed observations on the sequence of organ development are not recorded here for Zygocotyle lunata. However, all stages were observed in the course of the present study, using living material in hanging drops, and the process in Z. lunata follows very closely that described for C. cotylophorum. The time required for development and hatch- ing varies, the minimum time being 19 days and the maximum 40 days in the material of Z. lunata studied under laboratory conditions. Between these extremes the rate of development varied with the season of the year, more time being required during the colder than in the warmer months. Most of the eggs were embryo- nated at room temperature, and during the winter the room was much cooler during both day and night than in the summer. Temperature is probably the factor controlling speed of devel- opment, but controlled experiments using differ- ent temperatures at the same season of the year were not conducted. Table 1 shows the relation between time required for development and season of the year. In each case the time noted is that time elapsed from deposition of the eggs until emergence of the first miracidium in the given batch of eggs. In all the cultures, hatching was spread out over periods of 4 days or longer, indicating individual differences either in ma- turity of the eggs at deposition or in develop- mental rate. In a culture of developing miricidia of Z. lunata, most of the embryos are at about the same stage of organ formation but some develop more slowly than others in the same dish. Table 1. Seasonal variation in developmental rate of miracidia of Zygocotyle lunata. Date of deposition Date of hatching Days of eggs of first miracidia elapsed Dec. 2, 1937 Jan. 10, 1938 39 Dec. 4, 1937 Jan. 13, 1938 40 Dec. 16, 1937 Jan. 18, 1938 33 Jan. 22, 1938 Feb. 16, 1938 25 Feb. 2, 1938 Feb. 27, 1938 25 Feb. 23, 1938 March 21, 1938 26 March 5, 1938 March 27, 1938 22 March 14, 1938 April 7, 1938 24 April 15, 1938 May 8, 1938 23 May 25, 1938 June 15, 1938 21 June 9, 1938 June 30, 1938 21 June 10, 1939 June 29, 1939 19 July 22, 1939 Aug. 11, 1939 20 July 25, 1939 Aug. 16, 1939 22 Aug. 12, 1938 Aug. 31, 1938 19 Sept. 1, 1938 Sept. 22, 1938 21 Sept. 4, 1939 Oct. 4, 1939 30 Oct. 15, 1938 Nov. 9, 1938 25 Oct. 29, 1938 Dec. 1, 1938 33 Nov. 19, 1938 Dec. 19, 1938 30 Figures 3, 4 and 5 are photomicrographs of embryos of Z. lunata at 13, 18 and 21 days of development respectively, from a culture in which the first miracidium hatched in 21 days. In the 13-day stage the embryo is 135m to 140m long and is slipper-shaped, occupying the entire length of the egg. The yolk masses have begun to increase in size and decrease in number by a process which suggests a coalescence of the con- tents of adjacent masses. Very little movement of the embryo could be observed at this stage. Cilia are present but were not seen to move. The anlage of the “gut” is well established and flame cells are only occasionally seen. They have large nuclei at the base of a small tuft of cilia which measures approximately 6m X 3m. The 68 [XXVI: 13 Zoologica: New York Zoological Society nuclei are visible only after the embryonated egg is somewhat compressed and only after the em- bryo reaches a moribund state. No vitelline membrane could be discerned and from this stage on, the embryo occupies the entire length of the egg. No space is left at either end and the granular mass called a “mucoid plug” by Barlow (1925) and described in detail by Bennett (1936) for Cotylophoron cotylophorum, is entirely absent at all stages in eggs of Zygocotyle lunata. In Z. lunata numerous refractive, spherical masses are usually to be seen lying outside the embryo (Figs. 3, 4). They are left behind when the miracidium hatches and may be excretory in nature. An 18-day old embryo (Fig. 4) shows practi- cally all the body organs well developed and exhibits much movement, shifting backward and forward and constricting the body transversely at the junctions of the rows of epidermal cells. The yolk masses are now few in number and very large. Increase in length and width has necessi- tated a folding forward of the posterior fourth of the body, which becomes J-shaped. The nerve mass is well developed as a concentrated, some- what spindle-shaped band of small cells near the posterior limit of the anterior third of the embryo. The cilia are active, beating spasmodically every few minutes. The flame cells are large (12 n by 5.5 n) and actively beating, but no collecting tubules were observed at this stage. Germ cells are well-defined in the posterior third of the body. The embryo of 21 days development shown in Fig. 5 would not be ready to hatch before 24 to 48 hours although some others in the same culture dish were hatching. Several yolk masses are still present and appear to be partitioning off the part of the egg not occupied by the embryo. These apparent divisions always disappear com- pletely before hatching occurs. At this stage there is much constricting of the body as a whole and of the “ gut ” which is half as long as the body. Granules suspended in a fluid in the “gut” shift back and forth due to apparent waves of con- traction in the wall of the “gut” (Fig. 5). The cilia beat slowly from time to time. Anteriorly, the apical papilla presses closely against the operculum. The embryo maintains a J-shaped form and the position shown in Fig. 5 until the time of hatching. Hatching. With the disappearance of all the yolk masses the content of the egg is a continuous fluid in which the miracidium lies free. It shifts back and forth rapidly and the cilia begin to beat violently. One gets the impression at first that the apical papilla batters against the oper- culum to open it by mechanical pressure. More gentle movements of the same nature take place during the last 2 days before hatching. In the majority of cases the miracidium emerges without having turned around in the egg and this appears to be the normal procedure. Shortly before hatching, many large refractive granules from 4 m to 10 m in diameter, and clusters of smaller spherules, remain more or less stationary within the egg shell, but when the miracidium is ready to emerge, the violent ciliary action causes the granules to be swirled rapidly about with the fluid. Normally the operculum opens within one hour after this violent activity starts and emer- gence of the larva requires from a few seconds to fifteen minutes. Those which require the longer time to get out of the egg shell are apparently abnormal in some way or are injured in emerging, since they usually do not swim far and soon die. The diameter of the operculum is only about half the width of the miracidium which must constrict appreciably as it passes through. During the entire process of emergence the cilia are beating very rapidly. Some miracidia turn around again and again in the shell at hatching time. In such eggs, opening of the operculum appears to be delayed. Some of these larvae emerge normally when the operculum does finally open, but in other eggs the operculum opens only after the miracidium has exhausted itself by its activity and may emerge only half way and die in that position. Some swim around incessantly within the egg shell for as long as 7 hours and die there. It seems significant that in many cases where the larva fails to emerge readily, it turns about actively in the egg and applies the apical papilla against the two ends of the shell with equal fre- quency, apparently trying to find an opening at first one end and then the other. In two such eggs observed in hanging drops, the operculum opened while the apical papilla was directed to- ward the opposite end of the egg. Within a few seconds the larva turned and emerged rather slowly through the opening but swam away vigorously when free from the shell. This obser- vation seems to indicate that glandular secretions are produced at hatching time which effect the opening of the operculum, and that mechanical pressure by the apical papilla is probably not the direct cause for its removal. Normal larvae develop in a position adapted to straight-forward emergence. No embryos were observed develop- ing in a reversed position. A remarkable periodicity exists in the time of day that miracidia of Z. lunata hatch. In ob- servations on hundreds of mature miracidia in many different culture dishes at all hours of the day and night, very few were observed to hatch or to have hatched before 5:00 P.M. and no miracidia were ever found swimming in the cul- tures in the morning after 9:00 A.M. The ma- jority hatched between 10:00 P.M. and 2:0Q A.M. In many cases, mature miracidia were observed in eggs until about midnight or later in order to study the hatching procedure. Often none would hatch, but almost invariably some of them would be found either moving very feebly or dead in the dish at 9 :00 in the morning, having hatched during the early morning hours. Those which had not hatched during the night usually remained in the egg shell until the fol- lowing night, when more would emerge. Darkness does not seem to be the controlling factor, for just as many larvae hatch in a culture dish placed under the strong beam of a microscope 1941] Willey: The Life History and Bionomics of Zygocotyle lunata 69 lamp as in dishes covered with black paper. Attempts to induce consistent hatching of mi- racidia for study during the day, by the stimuli of light and darkness, were entirely unsuccessful. Agitation with dissecting needles of eggs con- taining mature miracidia sometimes causes the operculum to open, probably by mechanical in- jury to the egg shell, but most if not all individu- als induced to emerge in this fashion are abnormal and die without getting very far away from the opened shell. The miracidium swims very rapidly, usually in straight lines for considerable distances before darting off suddenly in a new direction. In Syracuse watch glasses they may be seen either swimming back and forth across the dish or fol- lowing around the edge in either direction. When no snail is in the dish the movements of the larva seem to be entirely at random. No phototropism was observed and they do not show any tendency to concentrate in any one place or in any way to influence each other. They swim incessantly for varying periods up to 7 hours, after which they slow down, sometimes swimming for a time in narrow circles, and die on the bottom of the dish in a somewhat bloated condition and contracted in length. As described in a later section of the present paper, they usually enter the snail host within two hours. Morphology of the Mature Miracidium. Studies on the morphology of mature miracidia of Zygocotyle lunata were conducted, using un- stained living and moribund individuals in hang- ing drops, silver impregnated specimens prepared according to the method described by Lynch (1933), and specimens stained intra vitally with neutral red. When swimming, the miracidium is elongated with sides almost parallel. Anterior- ly it is cone-shaped and terminates in a pro- trusible cap called by some authors the apical papilla. The posterior end is bluntly rounded. Moribund specimens and those which may be momentarily quiescent contract appreciably in length and become broader. The anterior, coni- form region and apical papilla may be retracted and the specimen becomes much expanded in the anterior half. Some variation occurs in size of miracidia of Z. lunata as in other species, but most of it is due to the varying amounts of con- traction and swelling which occur on natural death or fixation. Ten specimens killed by hot 0.5 per cent, silver nitrate for the silver impreg- nation method showed proportions most closely approaching those of the living miracidium. In these 10 specimens length varied from 184 n to 211 /z and width varied from 53 n to 59 n, the average length and width being 194 n and 55 n respectively (Figs. 7, 8) . Ten moribund specimens in hanging drops varied from 170 n to 231 n in length and from 68 n to 75 u in width, the average measurements of this group being 187 n by 69 n- The entire surface except the apical papilla and the very narrow spaces between rows of epithelial cells is ciliated. The cilia (Fig. 6) are about 10 n long except immediately behind the apical papilla where they are about 3 or 4 n long. These short cilia beat as do all the others. The cuticular non-ciliated cap or apical papilla is about 10 n across and a slight constriction of the body wall occurs just behind it. The apical papilla, called by various authors a “rostrum,” “oral cone,” “head papilla,” or a “terebrato- rium,” is said by some to be perforated by small glandular pores. Lynch (1933) described such pores in the apical papilla of Heronimus chelydrae. Coe (1896) found an opening which he refers to as the mouth in the “Ivopfpapille” of the mira- cidium of Fasciola hepaiica. No pores were identified on this structure in Z. lunata, and such pores have not been reported on the papillae of other amphistome miracidia. The external layer of the miracidium of Z. lunata is an epithelium consisting of 20 flattened, ciliated, epidermal cells. They are arranged in 4 tiers or rows, with 6 cells in the first (anterior) row, 8 in the second, 4 in the third and 2 in the last or posterior tier (Figs. 7, 8). These cells are well demonstrated by the silver impregnation technique used by Lynch (1933). Figure 7 is a photomicrograph of a silver impregnated mira- cidium taken with a 16 mm. objective to obtain sufficient depth of focus to show the entire thickness of the specimen. Figure 8 is a photo- micrograph of the same miracidium as in Fig. 7 but using an 8 mm. objective focussed on the upper surface, with consequent loss of depth of focus. The epidermal cells of the first tier covering the anterior fifth of the larval body are triangular in shape due to the tapering at the front end of the miracidium. They converge, almost meet, and seem to unite to form a ring in the base of the apical papilla. The cells of the second and third tiers are rectangular in surface view. Each of the 2 cells in the posterior tier covers one-half of the surface in the posterior fifth of the miracidium, the contiguous borders of the 2 cells lying some- what to the left and right of the median plane. The spaces between adjoining epidermal cells and between tiers of cells in Z. lunata are from 0.5 n to 1.5 n in width as seen in silver preparations. The spaces are homogeneous in appearance, without perforations and are bordered by wavy, irregular lines. No overlapping of plates was observed. Papillae and excretory pores open laterally through spaces between epidermal cells, as will be described later. The disposition of the epidermal plates with regard to the dorso-ventral axis of the miracidium is shown in Fig. 9, which is a dorsal view. Ad- joining borders of the dorsal (and ventral) epidermal cells of the second and third rows lie in the median plane, while those of the posterior row lie in a plane about 45 degrees removed from the median plane. The borders of the 6 anterior cells do not fall in line with any of those of the second row of 8 cells, and the medial edges of the 2 dorsal epidermal cells in the anterior row lie just a little to one side of the median plane. The miracidium shown in the photomicrographs (Figs. 7, 8) lies in a position of about 45 degrees 70 Zoologica: New York Zoological Society [XXVI: 13 of rotation to the right from a dorsal aspect, permitting favorable illustration of the lateral location of excretory pores and lateral papillae. Epidermal cells have been studied in miracidia of a few other species. A review of this work was presented in tabular form by Bennett (1936). The cell formula, 6;8;4;2 was reported by Bennett (1936) for Cotylophoron cotylophorum, and by Krull & Price (1932) for Diplodiscus temperatus. The present work shows that the miracidium of Zygocotyle lunata also possesses the formula 6;8;4;2, indicating that this formula is probably characteristic for the family Paramphistomidae. In studies on these cells in miracidia from other families, Thomas (1883), Ameel (1934) and Lynch (1933) reported some variation between different individuals within a species. In the majority of species, however, in which the formula has been described, no variation is re- ported. In the present work, 25 silver impregna- tion preparations of miracidia of Z. lunata showed no variation in the formula for the epi- dermal plates. As pointed out by Price (1931) and Bennett (1936) these structures are probably of importance in establishing natural relation- ships among the trematodes. Each epidermal cell contains a nucleus which may sometimes be seen in moribund or unstained dead miracidia, and may be best observed with the aid of intra-vitam stains. They do not show in silver impregnated preparations. In the pres- ent work they were studied from surface views of whole mounts, either unstained or stained with neutral red. The nuclei of the first tier of epi- dermal cells are much elongated, measure from 14 /j to 17 n in length and about 2 n in width, and lie very near the posterior border of the cells. Each of the 8 epidermal cells in the second tier contains an elongated but somewhat irregular- shaped nucleus near its posterior border (Fig. 9). These nuclei are 11 n to 13 n in length and 3 ^ in width. Those of the third tier of epidermal cells show the same position, shape and length but are only about 2 ^ in width. Each of the 2 posterior cells contains a centrally located nu- cleus about 14 n long and 5 m wide. As pointed out by Bennett (1936), the nuclei of the epidermal cells have been described for only a few miracidia. With slight variations in shape, size and position, these structures are apparently similar in the miracidia of all the Paramphistomidae in which they have been in- vestigated. In this group Sinitsin (1931) de- scribed them in the miracidia of Paramphistomum cervi, Krull & Price (1932) in Diplodiscus tem- peratus and Bennett (1936) in the miracidia of Cotylophoron cotylophorum. Beneath the surface layer of ciliated epidermal cells is a layer of transparent subepithelial cells. As indicated by their nuclei, these cells form a continuous layer around the internal structures of the miracidium. The nuclei are distributed over all portions of the subepithelial layer, but are more numerous in the anterior half of the body (Fig. 9). In an optical section of a whole mount viewed from the dorsal side, from 13 to 15 subepithelial cell nuclei may be seen along each outer edge of the miracidium. In surface view a group of from 12 to 15 may be seen overlying the “primitive gut” and smaller groups may be identified in the middle and posterior regions of the body. A few of these nuclei appear spherical but most are slightly elongated and vary from 6 n to 7.5 m in length and from 4 n to 5 m in width. Krull & Price (1932) showed that in the mira- cidium of Diplodiscus temperatus these nuclei are arranged in 3 definite rows, and Bennett (1936) found them to be distributed in 4 principal groups in the miracidium of Cotylophoron cotylophorum. Bennett pointed out, however, that not all the nuclei are to be found in these groups. In Zygocotyle lunata , the subepithelial nuclei are distributed irregularly through the subepithelial layer, and while they are more numerous in some areas than in others no definite arrangement into groups could be observed. Actually, examination of different specimens of Z. lunata showed con- siderable variation in number and distribution of the nuclei. Bennett (1936) observed mitosis in some of the subepithelial nuclei in C. cotylo- phorum and he points out the futility of attempt- ing to determine their number. The “primitive gut” is a saccate or flask- shaped structure occupying a considerable portion of the anterior region in the miracidium of Z. lunata. Its shape is exceedingly variable both in developmental stages within the egg (Figs. 4, 5) and after hatching of the larva. It may be elongated and narrow or it may be con- stricted transversely, or it may shorten and become very broad. When elongated it extends past the middle of the body and when shortened it occupies only the first third of the body. No lumen in the ordinary sense of the term is present and it is completely filled with a fluid containing a coarsely granular material which surges back and forth actively due to contraction of its walls and of the body of the miracidium. Anteriorly it tapers and seems to terminate blindly just behind the apical papilla. No opening to the outside could be found and nothing was observed to be either taken in or extruded from it. Its walls show no cell boundaries in the mature miracidium. Four large, somewhat ovoid nuclei, measuring about 8 n long by 5 m wide, are situ- ated near the posterior end, where they may be seen at all times since they remain attached to the wall and do not surge about with the granular contents of the sac (Figs. 5, 9). Most helminthologists have considered the “gut” of the miracidium as a primitive or ves- tigial intestine but more recently the work of Reisinger (1923), Manter (1926), Price (1931), Lynch (1933), and Bennett (1936) seems to suggest that this structure is a gland rather than a gut. Bennett (1936) says, “The development of the primitive gut at some distance from the anterior end of the body, the size of the cells and their nuclei, the early development of the granu- lar contents, the absence of a definite cell wall around each nucleus after the four-cell stage is reached, the concentration of cytoplasm around 1941] 71 Willey: The Life History and Bionomics of Zygocotyle lunata the nuclei at the posterior end of the gut, the absence of a mouth and a lumen, and the com- plete disappearance of the contents immediately after penetration of the miracidium into the snail host while the nuclei may still be identi- fied— all give evidence in favor of interpreting this structure as being a gland rather than a gut. ” The writer is inclined to agree with the opin- ions of the above-mentioned workers that the so-called “primitive gut” is probably glandular in its function, but more intensive studies on the embryonic origin and the fate of this structure will have to be conducted before final decision can be made as to its nature. If it functions in the penetration of the larva into the snail host, an opening must develop at that time, and the granular contents may consist of secretory granules formed in a structure which might be homologous with a gut in other forms. The excretory system of the miracidium of Z. lunata resembles that described for other mem- bers of the Paramphistomidae. Two large flame cells, one on each side, measuring from 13 u to 15 ju in length of flame and from 6 u to 7 u in width, are located just anterior to the middle of the body (Figs. 6, 9). Each possesses a large spherical nucleus at the base of the flame. A collecting tubule passes posteriorly in loose coils from each flame cell to a level behind the excre- tory pores, then loops forward to encircle the flame cell and again passes back to open at the excretory pore, which is located laterally just in front of the junction of the third and fourth tiers of epidermal cells (Fig. 8). A large spherical vesicle lies just anterior to each flame cell in the miracidium during late development in the egg as well as after hatching (Figs. 5, 6). However, no morphological association could be estab- lished between them and the excretory system. No accessory excretory cells as described for the miracidia of Heronimus chelydrae by Lynch (1933) and no duct nucleus as reported for that of Diplodiscus temperatus by Ivrull& Price (1932) could be found in the miracidium of Z. lunata. Krull & Price (1932) and Bennett (1936) re- ported 2 pairs of penetration gland cells in the miracidia of D. temperatus and C. cotylophorum respectively. They described these glands as 4 unicellular units extending from the base of the apical papilla backward for about one-fifth of the body length, with nuclei at their posterior ex- tremities. From the descriptions and figures, the 4 ducts which open anteriorly are each a part of one of the 4 gland cells. No indication is given concerning their role in penetration. In the present material of Z. lunata, 2 nuclei with clear areas around them were found on each side at a level near the junction of the first and second rows of epidermal cells. These nuclei, slightly ovoid in shape, were about 4 ^ in length. In spite of repeated observations on different miracidia at different stages of development, it was not possible to find any ducts leading an- teriad from these clear spaces around the nuclei. The greater portion of the posterior half of the miracidium is occupied by germinal tissue, which consists of about 40 germ cells as evidenced by their large spherical nuclei which measure from 5 m to 8 m in diameter. In all miracidia observed after hatching, one or more germ balls were present in addition to the germ cells. Most of the larvae contain one large germ ball measuring 25 n across and consisting of about 16 cells with a definite membranous covering. This germ ball lies in the anterior half of the body behind the “gut” and between the flame cells. In some specimens an additional smaller germ ball could be identified. The germ balls and the larger germ cells lying in the central cavity of the miracidium appear to be completely free and not attached to any other structure. The nervous system consists of nerve cells and fibers in association with a large ovoid mass lying dorsal to the “gut” and sending out pro- cesses laterally and posteriorly. Other processes probably extend forward but these were not observed. The dorsal nerve mass contains cells and fibers and measures about 25 m by 30 n- Large lateral processes could be observed in living miracidia while confined in the egg as well as after hatching. They could be traced to the body wall and were then lost. Very small cells, visible only with the aid of intra-vitam stains, lie scattered about outside of the central nerve mass. Two papillae protrude laterally through open- ings between the ciliated epidermal cells at the level of the posterior border of the first tier of epidermal plates (Figs. 7, 8). They have been observed in numerous other species and have been variously called lateral papillae, anterior papillae, lateral processes and anterior ducts by different authors. They are probably sensory in function since in some forms they are described as being associated with the central nerve mass. They are about 6 m in diameter and their position between the epidermal plates is clearly shown in silver impregnations by large round spaces. Smaller but similar spaces appear at other points of union between the first and second tiers of cells, but no structural units associated with them could be identified in Z. lunata. Lynch (1933) observed a number of small motionless bristles in this position in the miracidia of Heronimus chelydrae. Experimental Infestation of the Intermediate Host. The intermediate host in the life cycle of Zygocotyle lunata is the snail Helisoma antrosum. Naturally infected snails of that species provided the cercarial stage from which the adults were obtained and the complete life cycle demon- strated. Laboratory-raised snails were experi- mentally infected with miracidia developed in eggs from feces of experimentally infected ducks and rats. Helisoma antrosum reproduces readily in the laboratory at all seasons of the year and was available in stock tanks at all times. All snails used in the infection experiments were laboratory-bred. Miracidia were strongly at- tracted to the snails, to pieces of snail or to snail 72 Zoologica: New York Zoological Society [XXVI: 13 feces. To collect the rapidly-swimming mira- cidia from a dish, it was only necessary to place a snail in the dish and by the time one could place the dish on the stage of the microscope and focus on the snail, the miracidia would have gathered around it. They swim under and over and around the shell, occasionally attaching momentarily to the foot, edge of the mantle or to the shell, and then break away again and attach at some other point or even swim away to a different snail. If a snail is removed from the dish before penetration has occurred, the miraci- dia are attracted by the mucus left behind by the snail. No response to light could be observed. If no snail is present in the dish, the miracidia swim for periods up to 7 hours, settle to the bottom of the dish, and they may swim there in a narrow circle for a short time until death ensues. Miracidia were observed with snails for vary- ing lengths of time. In some cases miracidia disappeared under the shell within 15 minutes. Long and careful search failed to find them again and either they had penetrated or had been caught in the mucus secreted by the snail. This procedure was observed repeatedly, the time required for this apparent penetration varying from 15 minutes to 2 hours. In two cases, miracidia were seen to enter the space within the shell and attach to the base of the foot in a posi- tion perpendicular to it. Little progress, if any, toward penetration had occurred after one hour when observation was discontinued. In all the early attempts to infect Helisoma antrosum, medium sized snails from 8 to 10 mm. in diameter were exposed individually in separate finger bowls to 1 to 15 miracidia. Nineteen snails were thus exposed between January and June, 1938, but none of the snails became in- fected. After observing the apparent penetration of the larvae into the snails, successful infections were fully expected and no explanation could be offered for the negative results. On June 29, the writer, preparing to leave for Woods Hole, dumped about 40 eggs which were due to hatch 2 days later, into a battery jar containing some very young laboratory-raised snails. Another batch of 70 eggs of Z. lunata due to hatch July 8 was placed in a 2-gallon aquarium jar containing laboratory-raised snails of various sizes. Re- turning to the laboratory on August 17, the writer found encysted metacercariae on the glass in both jars. Isolation of the 42 snails showed 9 infestations, all but one of which were in snails 7 mm. or less in size. The single larger infected snail was 13 mm. in diameter. On August 27, 20 eggs of Z. lunata with miracidia ready to hatch were put into a finger bowl with 3 very young snails, each being about 2 mm. in diameter, and 35 days later 2 of the 3 snails began giving off cercariae. Young snails from 2 mm. to 6 mm. in size were used in all later experimental infections. Mass infections using large numbers of miracidia in small aquarium jars containing from 20 to 75 snails were much more successful than when a single individual was exposed to a few miracidia in a separate dish. In such mass infections, from 10 per cent, to 55 per cent, of the snails became infected. Not less than 80 successful experimental infesta- tions of Helisoma antrosum were obtained in which mature cercariae were produced. Many other experimentally infected snails were killed for ■ dissection or sectioning during the early stages of the infestation. The miracidium meta- morphoses in the snail into a sporocyst which in turn produces rediae. The redial generation gives rise to cercariae which emerge from the snail from 32 to 49 days after penetration of the miracidium. The snail host may carry an infestation with Z. lunata for long periods. The 4 naturally in- fected snails from which the life history studies began were collected in September and October, 1937, at which time they were giving off cercariae in numbers up to 100 per day. Of these, one died on December 29, 1937, 2 others died in April, 1938, and the 4th, which was recorded as a light infestation the previous September, lived until June 25, 1938. During this period they were kept isolated in finger bowls and fed lettuce leaves, and the water was changed about once a week. Many thousands of cercariae were pro- duced. Some of the experimental infections per- sisted equally long. From a group of 12 snails which began giving off cercariae on December 18, 1938, seven were still living in the laboratory and shedding cercariae after 9 months. From a size of not more than 5 mm. when infected they had grown to an average size of 15 mm. in diameter of shell. No snails were exposed to miracidia a second time after they were once infested with Z. lunata. None was observed to have lost the infestation and those which were crushed or died naturally after giving off cercariae for 9 months still contained large numbers of immature and mature rediae and cercariae. The larvae infest most heavily the liver and gonads of the snail host but rediae and immature cercariae are found in considerable numbers in the lymph spaces and practically everywhere in the snail except in the lumen of the intestine. None of the snails produced any eggs after becoming in- fected. As seen in sectioned snails, the gonads are reduced to a few shreds of tissue or cannot be identified at all. Sporocyst. The sporocyst of Z. lunata was observed only in the mature condition. Attempts to find early sporocysts by dissection of snails within a few days after exposure to miracidia were unsuccess- ful. Mature sporocysts were found in snails sectioned from 22 to 28 days after penetration of the miracidium. At this stage many rediae and immature cercariae are already free in the tissues of the snail. Usually no sporocysts can be found in an infected snail which has begun to shed cer- cariae, but one snail dissected 47 days after infestation, yielded a sporocyst measuring 297 n in length and 195 n in width which contained a single well-developed redia and nothing else. On manipulation of the coverslip the wall was ruptured and the young redia emerged. All other 1941] Willey: The Life History and Bionomics of Zygocotyle lunata 73 snails dissected later than 28 days after infesta- tion failed to yield sporocysts. No sporocysts of Z. lunata were found in the naturally infected snails studied. The mature sporocyst varies in shape from ovoid to elongate and is broader anteriorly than at the posterior end. It is a simple saccate structure with a body wall consisting of a cuticle, a membranous sheet and muscle fibers. Within the central cavity are germ balls and young rediae in various stages of development. The excretory system is that which is carried over from the miracidium and consists of a pair of flame cells with collecting ducts which open laterally a short distance behind the middle region of the body. A terminal bladder was ob- served on each side. A mature sporocyst found in sections of a snail killed 28 days after penetration of the miracidium contained 2 germ balls at the pos- terior end and 5 young rediae which occupied the central cavity. A reconstruction of this specimen is shown in Fig. 10. Each of the contained rediae shows a well-developed pharynx and an intestine which occupies most of the space within the young larva and extends nearly to its posterior border. The pharynx varied in diameter from 23 n in the 2 individuals in the posterior region to 27 m in the young redia near the anterior end. Figure 14 is a photomicrograph of one of the sections of this sporocyst and shows parts of 4 of the 5 rediae. The pliarynges shown are those of the two posteriorly placed rediae. Redia. The redial generation of Zygocotyle lunata was described by Willey (1936) in a paper on Cercaria poconensis. Since Cercaria poconensis Willey, 1930, is the larva of Z. lunata, the mature redia need not be again described here. However, since only naturally infected snails were available for the previous study, no information was ob- tained on the early stages of infestation. With an abundance of material from experimentally infected snails it has been possible in the present investigation to make more complete observa- tions which reveal the presence of daughter rediae. In the earlier description (Willey, 1936) of the stages found in the snail, it was stated that, “All the snails examined showed very many rediae of all sizes, but no sporocysts or mother rediae were found.” Many very small rediae as well as larger ones are always to be found in crushed infected snails, even in those killed 9 months or longer after experimental or natural infestation. This fact indicated strongly the existence of two generations of rediae, since the sporocyst disap- pears early in the course of the infestation. Ac- cordingly, in the present study, experimentally infected snails were dissected or sectioned at different periods following penetration of the miracidium of Z. lunata. Most of the observa- tions were made on living material from crushed snails, but sections of infected snails also show all the more important stages. In some of the snails crushed during early stages of the experimental infestations, mother rediae were found which contained only a single daughter redia and a few undeveloped germ balls near the posterior end. A specimen of this type shown in Fig. 11 was 330 m long and its pharynx measured 33.4 p in diameter. The daughter redia was 198 p in length and its pharynx also measured 33.4 p in diameter. The young redia showed considerable movement within the parent and its intestine extended back more than three- fourths of the length of the body. In this early stage of infestation no cercariae were as yet to be found free in the tissues of the snail, but rediae were also present which contained only devel- oping cercariae (Fig. 15). In a group of snails crushed 20 days after experimental infestation, numerous rediae were present which contained a single daughter redia and 3 to 6 germ balls which were not as yet differentiated. In no case could more than a single daughter redia be identified within a parent redia. Infected snails, killed 25 days or later after penetration of the miracidium, contained mother rediae with one daughter redia and numerous immature cercariae. Few such mother rediae were present in any one snail, but all snails examined between 25 and 47 days following in- festation contained this stage in addition to many rediae in which only cercariae could be observed. The identification of the different larval stages within the same mother redia was unmistakable. The single daughter redia seen in Fig. 13 showed much movement, squirming and turning around actively. The several cer- cariae present were larger and showed no move- ment. The well-defined pharynx in the daughter redia measured about 30 p in diameter, while the oral sucker of a developing cercaria is much larger, when it first becomes definable, than the pharynx of a large mature redia. Further, in some cases the eye-spots of the cercariae had begun to develop. The cones of cilia in the flame cells of daughter rediae are from 11 to 13 p in length, whereas those of cercariae of this species are only from 5 to 7 p long. Mother rediae containing both a daughter redia and cercariae were found in laboratory- infected snails studied during the period from August to March, but it seems probable that rediae are produced singly by mother rediae throughout the course of the infection. This is indicated by the fact that infected snails at all stages of infestation always contain great num- bers of mature and immature rediae of various sizes free in the tissues. They must be produced more or less continuously in the snail host and emerge from the parental generation at a rela- tively early stage. All daughter rediae observed were located in the anterior region of the mother rediae and they are always produced singly. Rediae containing only a daughter redia and no cercariae (Fig. 11) are small and are found only in the early stages of the infestation. Rediae containing both a single daughter redia and cercariae were all older and larger individuals 74 Zoologica: New York Zoological Society [XXVI: 13 measuring from .7 mm. to .9 mm. in length. The redia shown in Fig. 13 was .792 mm. long and contained a daughter redia measuring .335 mm. by .066 mm., while that in Fig. 12 was .860 mm. in length and contained a much younger daughter redia which measured .172 mm. in length and .063 mm. in width. These facts sug- gest the probability that each redia produces one daughter redia and proceeds from that time on to produce cercariae only. This would explain the presence of rediae of all sizes during all stages of the long-term infestation in this species. The rediae move sluggishly when freed from the snail, and in a watch glass usually remain in one place with only slight changes in body shape due to muscular contraction. No locomotor ap- pendages are present at am’ stage. Twelve flame cells are present in the mature redia of Z. lunata while only 6 appear in younger individuals. Most amphistome rediae described show a total of only 6 flame cells in the excretory system of the mature redia. Looss (1892) mentions 4 flame cells on one side in the redia of Diplodiscus sub- clavatus and Krull & Price (1932) figure a total of 7 in that of Diplodiscus temper atus. Looss (1896) described 5 pairs of flame cells in the mature redia of Paramphistomum cervi. A daughter generation of rediae has been described for several species of amphistomes. Looss (1896) reported more than a single gener- ation of rediae in Gastrodiscus aegyptiacus and in Paramphistomum cervi. In life history studies on P. cervi, neither Takahashi (1928) nor Szidat (1936) were able to find any evidence for a second generation of rediae. Beaver (1929) described 2 generations of rediae in Allassostoma parvum. Le Roux (1930) indicated that mother rediae occurred in the life cycle of Cotylophoron cotylo- phorum and Bennett (1936) found one specimen of a mother redia in that species. Krull & Price (1932) reported only one redial generation for Diplodiscus temperatus, but Herber (1938) demonstrated both mother and daughter rediae in that species. The development of rediae and cercariae within the same mother redia is known for only one other species of amphistome, and the present work constitutes the second report of this condition in amphistomes. Looss (1896) described a redial generation of this type in Gastrodiscus aegyptiacus. He figures a redia con- taining 3 daughter rediae and 3 cercariae and states that rediae containing only cercariae are rarely encountered in that species. Some authors have postulated that in certain forms a definite number of rediae and cercariae are produced. In the absence of experiments with known numbers of miracidia infecting the snails, no definite conclusions can be reached in this regard in any species of trematodes. In Zygocotyle, the finding of only small numbers of mother rediae containing single daughter rediae and cercariae together indicates that the produc- tion of rediae is a continuous process and is not limited to a definite number. Variations may occur in the redial generations in the snail under natural conditions accompanying seasonal changes. Experimentally infected snails kept under laboratory conditions do not always afford sufficient information on which to base conclu- sions concerning the prevalence and numbers of the various larval stages. The course of events probably varies with changing conditions. Much work still remains to be done on these problems of development. Cercaria. The cercaria of Zygocotyle lunata was named Cercaria poconensis in an abstract by Willey (1930), and in 1936 he described the redial and cercarial generations together with a report of infestation experiments with the metacercarial stage. The cercaria need not be redescribed here and only additional information not available from the earlier studies will be presented. The material from which C. poconensis was described and that from which the life history of Z. lunata is demonstrated in the present study are unmis- takably identical. The size and the shape and distribution of the body organs in the two groups of specimens agree perfectly. The size measure- ments vary somewhat with different degrees of flattening. For example, the oral sucker in fixed and unflattened cercariae shows an average measurement of .064 mm. in diameter, while in living specimens, flattened under medium pres- sure, it may be as large as .106 mm. across. The oral sucker and oral evaginations together measure .165 mm. in length in fixed, unflattened individuals. Nasmark (1937), in a revision of the Param- phistomidae, has attempted to show that the sucker at the anterior end in amphistomes is a pharynx rather than an oral sucker. On the basis of histological studies, he believes that this structure is homologous with the pharynx of the monogenetic trematodes and the rhabdocoel turbellarians, and states that contrary to the opinion of Looss (1902) and others it should be designated a pharynx. However, due to the incomplete nature of Nasmark’s evidence, the conclusions are not accepted by the present author and the sucker is here referred to as an oral sucker. Fischoeder (1903) referred to the anterior sucker of Z. lunata as a pharynx, causing some misinterpretation in the literature, but all the more recent authors have preferred to call it an oral sucker. Until final conclusive evidence is forthcoming, the terminology currently used seems more desirable. The posterior overhanging lip of the acetabu- lum with its 2 lateral conelike projections (Figs. 21, 22), which are so characteristic of the genus Zygocotyle in the adult stage, were not observed in the earlier studies on the cercarial stage. But after it became known that this cercaria was the larval stage of Zygocotyle, the lip and conical projections were observed from the ventral sur- face of cercariae when under only slight pressure. This modification of the acetabulum is present on the cercaria in an immature condition. Figure 16 is a photomicrograph of a living cer- caria under medium pressure and shows the 1941] Willey: The Life History and Bionomics of Zygocotyle lunata 75 relationships of the various organ systems. The characteristic branching of the main excretory ducts is clearly outlined by the presence of ex- cretory concretions. Drawings from both living and fixed material of redial and cercarial genera- tions appear in the earlier description (Willey, 1936). The cercaria leaves the redia while still in a very immature condition. The intestine, oral sucker, acetabulum, eye-spots, tail rudiment and the excretory system, already laid down while the cercaria is in the redia, continue their devel- opment after emergence. Some variation occurs in the size and degree of differentiation of the cercariae at the time of emergence. Young cercariae are frequently observed free in the tissues of the snail which have not developed as far as other cercariae which are still within nearby rediae. In experimental infestations, young cercariae were first found free in the tissues of snails crushed 16 days after exposure to miracidia. In other snails, crushed 20 days after exposure to miracidia, none had as yet emerged from the rediae. The rate of development is known to vary with temperature conditions. The time required for the complete development of the cercaria, from penetration of the miracidium until shedding of the first mature cercaria from the snail, varied in the laboratory from 32 days during the warmer months to 49 days during the winter. After emergence, cercariae swim about vigor- ously in the water at the side of the container toward the light for 30 minutes to 2 hours. They respond very rapidly to changed lighting condi- tions and will follow a beam of light moved about from one side of the dish to the other. Encyst- ment occurs on the side of the dish toward the light, on the shell of the snail, or occasionally on vegetation. If handled or otherwise irritated they encyst almost immediately, often attaching and encysting within a dropper when being transferred to a slide for examination. If placed on a slide in a small amount of water, a coverglass must be added immediately to prevent encyst- ment. The shell of a snail producing cercariae is usually crowded with encysted metacercariae. Cercariae emerge in greatest numbers between 10:00 A. M. and 2:00 P. M. and usually only on bright days. On dull, rainy days very few or none escape, while on bright, sunny days as many as 100 may emerge from a single infected snail at the peak of the infestation. Cercariae may be produced from an infected snail for 9 months and longer, but the number of cercariae escaping each day decreases as the snail remains longer in the laboratory. Little difference could be ob- served between the numbers of cercariae pro- duced in naturally and experimentally infected snails. Metacercaria. The cercaria encysts and passes into the meta- cercarial stage in which it awaits ingestion by the final host. The process of encystment is very rapid. The cercaria attaches itself by means of its suckers, the tail vibrates from side to side somewhat more slowly than in swimming, and the body appears to undergo rapid squirming movements. Cystogenous material then oozes out rapidly over the surface of the body from the elliptical cystogenous granules which occupy most of the dorsal half of the cercaria. The cyst wall forms rapidly and the tail is left attached to the outside of the cyst, where it lashes violently for an hour or more and then drops off, sinks to the bottom and may continue lashing about for several hours. The body of the cercaria twists and turns about during the process as though molding the inner wall of the cyst. Finally the cyst wall hardens, the metacercaria coils about in the cyst and after several hours becomes rela- tively motionless with suckers apposed, periodi- cally undergoing slight twitchings and contrac- tions in various regions of the body. The cysts are large dome-shaped hemispheres with thick resistant walls which are brown to black in color when seen with the naked eye. The base of the cyst flares out slightly from the margin of the cyst proper (Fig. 18). When formed on glass they are flat on the bottom and the greatest diameter shows an average measure- ment for 10 metacercariae of .368 mm. The diameter at the base of the dome varies in 10 individual cysts from .277 mm. to .343 mm., the average being .289 mm. Figure 19 is a photo- micrograph of a 10-day old metacercaria which has been dissected from its cyst. When removed thus in water or salt solution, they contract and undergo little or no movement. Much black pigment is present in the body wall and this tends to obscure the internal structures, which show no advance in development over the condition found in the cercaria. The black pigment of the eye-spots and body persists throughout meta- cercaria] life and scattered granules of eye-spot pigment are still present after 3 weeks of devel- opment in the final host (Figs. 21-23). Infestation Experiments with Metacercariae. Early attempts to infest the final host with the metacercaria of Z. lunata were conducted, using cold-blooded hosts. As outlined by the writer in an earlier paper (1936), encysted metacercariae were fed to tadpoles but all the experiments were negative. In 1937, following unsuccessful at- tempts to infest turtles with the larvae, some of the metacercariae were dissected from their cysts and placed in cold Ringer’s fluid at 20° C. All were dead after 3 hours. Artificial digestion of some of the metacercariae from their cysts with pepsin and pancreatin solutions at 37° C. indi- cated that the final host was a warm-blooded animal. Under this treatment the cyst walls became soft and movement of the metacercaria was observed after 4 hours. After 10 hours and up to 20 hours of artificial digestion, young worms were still alive in the cysts while others showed much activity on being released from the cyst. Consequently on October 8, 1937, 2 young laboratory-raised rats were each fed about 65 76 Zoological New York Zoological Society [XXVI: 13 metacercariae obtained from the naturally in- fected snails as described earlier in the paper. Five days later one of the rats was killed and yielded 59 young worms in the cecum with none above or below this level in the intestine. De- velopment had proceeded far enough to permit identification of the worms as Zygocoiyle. The small cone-like projections on the posterior edge of the acetabulum are well developed at this stage (Fig. 20). The worms, still immature, show average measurements for 10 of them of .940 mm. in length and .460 mm. in width. The common natural hosts of Zygocoiyle are various species of ducks. Thirteen young ducks were subsequently fed varying numbers of cysts and all became infected. Similarly, in the course of the investigation, more than 60 rats and one ram were experimentally infested. Attempts to infect pigeons and rabbits gave negative results. In order to determine the time required for the young worms to mature in the final host, fecal examinations were conducted almost daily on some of the experimental hosts. Two ducks, (nos. 1 and 3) fed metacercariae on October 14, 1937, began giving off eggs of Zygocoiyle in fecal material on November 24, or on the 41st day of infestation. Duck number 3 was killed 5 days later and contained 7 mature specimens of Zygocoiyle lunata, 5 in one cecum and 2 in the other. Similarly 4 rats (nos. 3, 4, 5 and 6) were fed from 30 to 40 encysted metacercariae each on October 14, 1937. On November 24, when the 2 ducks described above showed eggs in the feces, the rats were still negative. One rat, number 5, killed on November 28, the 45th day, contained 23 Zygocoiyle, several of which showed a few eggs in the uterus near the ovary, but none was fully mature as yet. Eggs of Z. lunata from this series of rats first appeared in the feces of rat number 6 on the 46th day after infestation. This rat was killed on December 4, the 51st day after the experimental feeding, and contained 8 fully mature Zygocoiyle lunata in the cecum. The more rapid development in the duck may pos- sibly be due to the higher body temperature maintained by that host. This difference was confirmed in other series of infected rats and ducks, the data for which will be described later. Thus, beginning with cercariae from naturally infected snails, all stages in the development of the adults of Zygocoiyle lunata were obtained in rats and ducks. As described in the earlier sections of the present paper, eggs from feces of the experimental hosts were embryonated to produce miracidia which were experimentally introduced into laboratory-raised snails. In the snail, Helisoma antrosum, the sporocyst and the redial and cercarial stages were obtained and the complete cycle was carried out several times in the laboratory. Experiments on Injectivity of Metacercariae. After encystment the metacercaria remains quie- scent and awaits ingestion by the final host. No development occurs within the cyst. Metacer- cariae dissected out or freed from their cysts by artificial digestion after one day of encystment are in no way distinguishable from those similarly freed after 20 or 30 days. This observation was confirmed by feeding experiments with rats. Cysts in a few drops of water were placed on small pieces of bread and given to the rats after withholding all food for 24 hours. In an experi- ment conducted on December 30, 1937, rat 22 was fed 100 metacercariae which were 14 days old and rat 25 was fed 100 metacercariae which were 85 days old. Both rats were killed 15 hours after ingesting the larvae and each rat yielded more than 30 excysted worms in the cecum. Some larvae were still within their cysts and none could be found anywhere in the digestive tract other than in the cecum. No significant differ- ence could be detected between the two groups of young worms. All were equally active and the number of worms found was approximately the same in the two hosts. In an experiment de- signed to detect any changes which might occur immediately after encystment of the metacer- caria, rat 12 was fed 50 newly encysted larvae, some with the cercarial tail still attached and the cyst wall still soft, and rat 11 was fed 70 meta- cercariae which were 47 days old. Both rats were killed 3 days later and yielded young worms in the ceca which were practically indistinguish- able morphologically in the two hosts. Thirty- five worms were collected from rat 12 which re- ceived 50 newly formed cysts and 60 worms were obtained from rat 11 which received 70 cysts 47 days old. A few cysts are probably destroyed by the rat in the chewing of the bread. In a third experiment, one rat (no. 23) was fed 15 metacercariae which were 2 days old and 15 which were 32 days old. This rat was killed 6 days later and the 20 young worms collected from the cecum showed no significant differences in size or degree of differentiation. The rate of development of Zygocotyle lunata to sexual maturity is not correlated with the age of the metacercaria at the time of its ingestion. These experiments show conclusively that no develop- ment occurs in the cyst of Z. lunata and also that the encysted metacercariae are infective im- mediately after encystment. Longevity. In one of the foregoing experi- ments, metacercariae, encysted for 85 days, were viable when fed to rat 25. In a series of longevity experiments, it was determined that the meta- cercaria will live much longer than that. En- cysted metacercariae were kept at room tem- perature in the laboratory in covered finger bowls attached to the glass wall where first deposited, and the water was not changed except to add water occasionally to make up for evapo- ration. Metacercariae not older than 3 months were viable and when fed to rats produced in- festations with Z. lunata. Encysted larvae older than 4 months gave variable results (Table 2). The oldest metacercariae which successfully produced an infection in rats had been in the laboratory for 138 days. In this experiment rats 48 and 49 were each fed 50 metacercariae from cercariae emerged from snail number 2 on Janu- ary 30, 1938. Movements of the larvae within 1941] 77 Willey: The Life History and, Bionomics of Zygocotyle lunata Table 2. Longevity of encysted metacercai’iae of Zygocotyle lunata. Host and date Age of Number of Elapsed time Number of of feeding metacercariae metacercariae before killing worms metacercariae in days ingested host (Days) recovered Remarks Rat 34 June 2, 1938 86 10 11 3 Rat 35 June 2, 1938 86 10 11 8 Rat 30 Feb. 3, 1938 121 100 46 0 Movement observed in cyst before feeding to rat Rat 29 Feb. 3, 1938 131 100 46 0 Movement observed in cyst before feeding to rat Rat 46 June 9, 1938 130 50 9 23 Movement observed in cyst before feeding to rat Rat 47 June 9, 1938 130 50 9 28 Movement observed in cyst before feeding to rat Rat 48 June 17, 1938 138 50 169 1 Eggs in feces after 74th day Rat 49 June 17, 1938 138 50 169 2 Eggs in feces after 74th day Rat 63 Feb. 10, 1940 152 50 14 0 Cysts shrunken Movement in 10% Rat 44 June 4, 1938 166 150 9 0 Cysts not examined Rat 45 June 4, 1938 166 150 9 0 Cysts not examined Rats 64, 65, 66, 67 Feb. 29, 1940 171 25 each 5-10 0 Cysts shrunken No movement observed the cyst were observed before feeding. Seventy- four days later many eggs of Z. lunata were col- lected from feces of both rats and when these 2 rats were killed after 6 months of infestation, large, adult specimens of Z. lunata were re- covered. In a similar experiment with rats 46 and 47 (Table 2), metacercariae 130 days old were fed and when the rats were killed 9 days later, 23 young worms were recovered from the cecum of rat 46 and 28 from rat 47. As indicated in Table 2, cysts 152, 166 and 171 days old re- spectively were fed to rats but no infestations resulted. Some variation in longevity occurs. In the experiment involving rats 29 and 30, no infection was obtained from metacercariae aged 131 and 121 days respectively, although movement was detected within the cyst. Minor variations in other factors would readily account for such slight variations in longevity and infectivity. That these are not due to individual differences in the host reaction is indicated by the fact that in most of the experiments dealing with longevity of metacercariae, 2 host animals were fed similar numbers of larvae produced from the same snail host on the same dates, and in each case the re- sults were always the same in the 2 hosts con- cerned. No evidence is available as to what constitutes optimum conditions for long survival of encysted metacercariae. Bennett (1936) reports that under optimum conditions the metacercaria of Cotylophoron cotylophorum probably lives for several months. In his experiments, he pre- sumably kept them at room temperature and after 3 months (June 5 to September 5) 33 per cent, were still alive. At this point observations were discontinued. Krull (1934) kept metacer- cariae of the same species alive under the same conditions for as long as 5 months from July 2 until December 2. Both of these workers used the criterion of movement of the metacercaria to determine its life span. As indicated by the data on rats 29, 30 and 63 (Table 2), movement in the cyst shows the metacercaria of Z. lunata to be alive as long as 5 months, but such larvae, when fed to the final host, were in these three attempts unable to survive conditions within the final host. It seems probable that motility is not a valid criterion of infectivity. Experimental Studies on Viability of Meta- cercariae. Experimental feedings conducted with rats show that at least 85 per cent, of the meta- cercariae of Z. lunata excyst and survive for a time in the final host. The percentage of viable metacercariae is probably higher since some of the cysts may be destroyed in the chewing proc- ess by the rat host. The number of meta- cercariae which have excysted and are recover- able as young worms is seen to be very high if the host is killed within the first few days after ingestion of the cysts, because, as will be shown later, the host may begin to lose some of the worms as early as the second or third day. Table 3 shows the number and percentage of worms obtained in 10 experiments in which the hosts were killed within the first 11 days. The high percentage of viability of metacercariae is shown in feeding experiments with ducks as well as rats. For example, duck 3a, fed 150 encysted larvae and killed 11 days later, yielded 144 young worms in the ceca. 78 Zoologica: New York Zoological Society [XXVI: 13 Table 3. Viability of encysted metacercariae of Zygocotyle lunata. Metacercariae 73 bfi O ingested by host co .2 & s O -3 'ri W PORT ANGELES"^ PORT GUATULCO- SANTA CRUZ B.'' TANGOLA-TANGOLA B. 1 GULF OF FnNAFrA-' GULF OF (alcat.ai ,sl. SAN JUAN DELSURr* PORT PARKER MU RC I EL AGO B. POTRERO GRANDE PORT CULEBRA BRAXILITO B. PIEDRA BLANCA B NICOYA GULF OF DULCE EASTERN PACI F1C E X P E D T 1 0 N S NEW YORK ZOOLOG ICAL SOCIETY S H 0 R E C OLLECTI NG STATIONS GA L APAGOS IS. Text-figure 1. Principal localities in the tropical eastern Pacific where collections were made by the Department of Tropical Research of the New York Zoological Society. 15. Dorsal fins each provided with a spine Squalidae, p. 120 Dorsal fins without spines Scymnorhinidae, p. 121 16. Gill-openings on sides of body; anterior exten- sion of the pectoral fins separated from the neck by a deep notch in which lie the gill- openings Squatinidae,p. 121 Gill-openings on under side of the body. (Rays and mantas; these will be treated in a suc- ceeding paper.) 96 [XXVI: 15 Zoologica: Neiv York Zoological Society Systematic Account. Family Chlamydoselachidae. Chlamydoselachus Garman, 1884. Chlamydoselachus anguineus Garman. Unrecorded from our region, although Garman4 mentions the following: “This shark is one that may confidently be expected to appear in future collections from the region about the Gahipagos.” Family Hexanchjdae. Hexanchus Rafinesque, 1810. Hexanchus griseus, the six-gilled shark, is one of a number of sharks unrecorded from the tropi- cal eastern Pacific, but reported from immedi- ately north and south of the region under con- sideration. On the Pacific coast the range is from San Diego, California, northward, and the species is known from Chile, if the synonymizing of the South American form with the northern is correct. Heptranchias Rafinesque, 1810. The seven-gilled sharks also have representa- tives both north and south of the region from Cedros Island to northern Peru. Family Orectolobidae. Gin gly mo stoma Muller & Henle, 1837. Ginglymostoma cirratum (Gmelin). Nurse Shark (Sand Shark; Gato). Range: In our province from the Gulf of Cali- fornia, south to Ecuador. (Mexico: Gulf of California, Mazatlan, Jalisco, Chamela Bay, Manzanillo; Costa Rica: Port Parker, Potrero Grande, Gulf of Nicoya; Panama: Bahia Honda, Panama Bay; Ecuador: St. Helene Bay.) Field Characters: A sluggish, blunt-headed shark with very small eyes; mouth transverse and near the tip of the snout, furnished with a pair of barbels; two large dorsal fins placed far back, the first over the pelvics; the fourth and fifth gill-slits very close together. Olive brown; young covered with small black spots which usually disappear in the adult. (Illustration after Garman; 365 mm.) 4 Garman, S., Mem. Mus Comp. Zool., 24, 1899: 41. No detailed comparison has been made at first hand between nurse sharks from the Atlantic and from the Pacific, and though probably there is specific differentiation, we can consider them at present as only a single species. Size: Reaches a length of 10 feet. Local Distribution: Feeding and usually living close to the bottom in shallow water near shore. Abundance: Cannot be considered a rare shark wherever it is found. Food: A 385 mm. shark taken from a tide pool in Costa Rica had eaten two carideans and five small fishes. One of the latter was a Rupiscartes atlanticus (85 mm.) and another an Anchoviella (43 mm.). An adult shark had eaten a six-inch fish. Enemies: Man was the only enemy of this shark which came to our notice. A dozen nurse sharks, six to eight feet in length, were found on the beach at a camp devoted to drying shark fins, near the tip of Lower California, on May 1, 1936. Unlike the other three species found at this camp, the nurse sharks had their fins intact, these hav- ing evidently no commercial value. Parasites: Two leeches, Pontobdella muricata, were taken from a nurse shark. Breeding: The largest specimen (2,800 mm.) showed no signs of active breeding; the smallest (295 mm.) was taken in a tide pool at Bahia Honda, Panama, on March 19, 1938. Study Material: We observed this shark seven times; three specimens were collected, the others were watched through helmet or water-glass. Mexico: Chamela Bay, 1, observed, ca. six feet, Nov. 10, 1937; Manzanillo Bay, 1 (26,093) 392 mm., Jan. 10, 1938, poisoned; Costa Rica: Port Parker, 1, ca. four feet, Jan. 13, 1938, observed; Potrero Grande, 1, ca. seven feet, Jan. 23, 1938, observed in 2 feet of water; Gulf of Nicoya, (26,170) 2,880 mm., Feb. 23, 1938, snagged through dorsal with hook from Zaca; Panama: Bahia Honda, 1, (26,206) 295 mm., March 19, 1938, poisoned in tidepool; Bahia Honda, 1, ca. 300 mm., March 19, 1938, specimen lost, from same pool as 26,206. References: Squalus cirratus, Bonnaterre, Tabl. encyc. meth., Ichtliy., 1788: 7 (original descrip- tion) . Ginglymostoma cirratum , Jordan, D. S. & Gilbert, C. H., Bull. U. S. Fish Comm., 2, 1882: 105 (Mazatlan, Mexico). Jordan, D. S. & Gilbert, C. H., Proc. U. S. Nat. Mus., 5, 1883: 371 (First record from Pacific coast, 2 ten-inch specimens collected by Xantus, Colima, Mexico). Jordan, D. S. & Gilbert, C. H., Proc. U. S. Nat. Mus., 5, 1883: 620 (Panama). Vaillant, L. L., Bull. Soc. Philom., Paris, (8) 6, 1894: (Gulf of California). Jordan, D. S., Fishes of Sinaloa, 1895: 381, (color, common name “Gata,” 3 specimens, 2 to 6 feet). Jordan, D. S. & Ever- mann, B. W., Fishes North and Middle America, 1, 1896: 26 (short description). Boulenger, G. A., Boll. Mus. Torino, 13, 1898: 1 (Ecuador). Gilbert, C. H. & Starks, E. C., Fishes Panama 1941] Beebe & Tee-Van: Sharks 97 Bay, 1904: 5 (color, Panama). Garman, S., Bull. Mus. Comp. Zool., 46, 1906: 229 (Panama). Regan, C. T., Proc. Zool. Soc. London , 1908: 350 (synonymy, description, Jalisco, Mexico). Garman, S., The Plagiostomia, 1913: 54, plate 7 (description, figures). Meek, S. E., & Hilde- brand, S. F., Marine Fishes of Panama, 1, 1923: 29 (Panama, description). Ginglymostoma fulvum, Jordan, D. S., Proc. U. S. Nat. Mus., 8, 1886: 363 (listed as a doubtful species) . Family Rhineodontidae. Rhincodort Smith, 1849. (Usually consid- ered as a misprint for Rhineodon.) Rhineodoti typus Smith. Whale Shark. (Plate 1, Figs. 1 & 2). Text-figure 3. Range: Warm waters of the Atlantic, Pacific and Indian Oceans. In the eastern tropical Pacific it is known from the following localities, Mexico: Gulf of California, Cape San Lucas, Acapulco; Panama: Panama Bay, Galapagos Islands; Peru: Callao. Field Characters: A giant whale-like shark with spotted head and with the body covered with longitudinal and transverse pale bands forming a checker-board ; each space in the checker-board with a pale yellowish spot. (Illustration after Norman, 1937.) Size: Grows to 45 feet long, with possible records ranging up to 70 feet. General Habits: Considering its size, this is a common fish in Gulf of California waters. Our observations on the species are summarized in “Zaca Venture,” pages 162-170. Study Materials: No specimens. Many obser- vations and photographs were made of this species in the Gulf of California. References: Rhineodon typus, Smith, A., Zool. Journal, London, 16, 1829: 443-444 (original description; type locality, Table Bay, Cape of Good Hope, South Africa). Micristodus punctatus, Gill, T. N., Proc. Acad. Nat. Sci., Phila., 17, 1865: 177. Gill, T. N., Science, 15, 1902: 824-826. Jordan, D. S. & Evermann, B. W., Fishes North and Middle America, 1, 1896: 52 (Quotes original description of teeth). Mowbray, L. L., Preliminary Report on the Taking of the Pacific Whale Shark during the Scientific Cruise of the Yacht “Nourmahal” in the North Pacific, 1933, under the direction of Vincent Astor, Privately printed, July, 1933, No. 1, 2 pp. (Galapagos Islands). Rhineodon typicus, Smith, R. W., Illustrations Zool. South Africa, 1849: Plate 26. Nation, W. S., “South Pacific Times,” Callao, Peru, Jan. 24, 1878. (Callao, Peru). Chierchia, G., Collezioni per Studi di Scienze Naturali Fatte nel Viaggio Intorno al Mondo della R. Corvetta “Vettor Pisani” . . . Anni 1882-83-84-85. Roma, 1885: 66-68 (Panama Bay). Grey, Z., Santa Catalina Islander, May 27, 1925: 10 (Cape San Lucas). Grey, Z., “Fishing Virgin Seas,” New York, 1925: 204-216 (Cape San Lucas). Beebe, W., “The Arcturus Adventure,” New York, 1926: 414 (Galapagos Islands). Gudger, E. W., Science, 65, 1927: 545 (Galapagos Islands). Gudger, E. W., Science, 65, 1927: 211-212 (Gulf of California). Gudger, E. W., Nature, 132, 1933: 569 (Galapagos Islands). Lonnberg, E., Fauna och Flora Upsala, 1933: 97-104 (Acapulco, Mexico). Gudger, E. W., Proc. Zool. Soc. London, 1934: 874-878 (Distribution in the eastern tropical Pacific). Gudger, E. W., Natural History, 36, 1935: 128-132 (Acapulco, Mexico). Gudger, E. W. & Smith, R. S., Bull. N. Y. Zool. Soc., 68—71 (six specimens at Acapul- co, Mexico, photographs). Beebe, W., Bull. N. Y. Zool. Soc., 39, 1936: 241-242 (notes, photo- graphs), (Gulf of California, Cape San Lucas). Gudger, E. W., Nature, 141, 1938: 516 (Panama). Gudger, E. W., California Fish and Game, 24, 1938: 420-421 (Lower California). Fowler, H. W., Acad. Nat. Sci. Phila., Monograph 2, 1928: 248 (Galapagos Islands, check-list). Beebe, W., “Zaca Venture,” New York, 1938: 162-170 (observations and harpooning, abundance, size, Gulf of California and Cape San Lucas). Family Alopiidae. Alopias Rafinesque, 1810. Alopias vulpinus (Bonnaterre). The thresher-shark, although supposedly world-wide in distribution, is unrecorded from our area, but it has been found in considerable numbers in southern California and in Chile. Walford, in writing of southern California Alopias vulpinus, states: “It has been recorded on our coasts as far north as Coos Bay, Oregon, and is said to extend at least as far south as the Isthmus of Panama.” This statement seems to be the only recorded reference to the species in coastal tropical eastern Pacific waters. Alopecias barrae Philippi6 and A. longimana Philippi,6 both reported from Chile, are con- sidered by Fowler7 as being the same as Alopias vulpinus. 6 Philippi, R. A., Anales Univers. Chile, 71, 1887: 553, pi. 5, fig. 2. « Philippi, R. A., Anales Univers. Chile, 109, 1901: 308. 7 Fowler, H. W , Proc. hth Pac. Sci. Congr., Java 1929, Biol. Pap. 3, 1930: 488. 98 [XXVI : 15 Zoologica: New York Zoological Society Whitley8 * points out that the assumption that Alopias vulpinus has a world-wide distribution may not be correct. A number of new forms of Alopias have been described, and the eastern tropical Pacific form may very well represent another species. Family Isuridae. Key to genera mentioned in this paper. la. Teeth slender and sharp, with entire edges. I surus lb. Teeth compressed and triangular, with serrated edges Carcharodon Isurus Rafinesque, 1810. la. Teeth with a basal cusp on each side nasus lb. Teeth without basal cusps, long, flexible and acute glaucus Isurus nasus (Bonnaterre) and Isurus glaucus (Muller & Henle). The genus Isurus and the species mentioned above are unreported from our province, al- though they are present north and south of the area if the Chilean Lamna hudobroii 9 is the same as Isurus glaucus as Fowler states,10 and if Lamna philippi Perez11 is the same as Isurus nasus, as Fowler mentions in the same publication. We have the jaws of an Isurus obtained from a native in the Gulf of Dulce, Costa Rica. There is no definite locality record for this jaw, and although it was probably taken from somewhere within our region, we have no positive assurance that this was so. This specimen would be referred to glaucus in the key. There are 13 teeth on each side in each jaw, the teeth becoming progressively smaller from front to back. The teeth are smooth and there is no indication of cusps. The longest teeth measure 31 mm. from base to tip. In the upper jaw the third tooth is considerably smaller than the following teeth, being approximately equal in size to the eighth. The space between the third and fourth teeth in the upper jaw is wider than the spaces between other teeth. Carcharodon Muller & Henle, 1838. Carcharodon car char ias (Linnaeus). There seems to be no authentic record of the great white shark in our region, although from the distribution of the species one would expect to find it. It is recorded from southern Cali- fornia and from Chile. 8 Whitley, G. P., Rec. Austr. Mus., 20 (1), 1937: 5. 3 Philippi, R. A., Anales Univers. Chile, 71, 1887: 551. 10 Fowler, H. W., Proc. hth Pac. Sci. Congr., Java 1929, Biol. Pap. 3, 1930: 488. ■' Perez, C. C., see Philippi, R. A., Anales Univers. Chile, 71, 1887: 549. Family Cetorhinidae. Cetorhinus Blainville, 1816. Cetorhinus maximus (Gunner). Basking Shark. Range: Arctic, Antarctic and temperate seas; in the eastern tropical Pacific it is known from Ecuador and Peru and from somewhere near the Galapagos Islands. On the northern limit of the area under discussion it is known from southern California, and there is a questioned record from near Cape San Lucas. Field Characters: A very large shark with long gill-slits extending from the back and almost meeting under the throat. Caudal fin lunate, with keels on each side near the base. Back leaden gray, lower sides white. Often found drifting at the surface. (Illustration after Norman, 1937.) Size: Grows to 45 feet. Study Material: None. References: Squalus maximus, Gunner, J. E., Trondjhem sels. selskr., 3, 1765: 33, tab. 5 (orig- inal description, coast of Norway). Cetorhinus, Stevenson, C. H., Rep’t. U. S. Fish Comm. 28, 1902 (1904) : 227-228 (Peru and Ecuador, use for oil). Gudger, E. W., Science, 42, 1915: 653-656 (Occurrence in the southern hemisphere) . Basking shark, Grey, Z., “Tales of Fishing in Virgin Seas,” New York, 1925: 185 (Questionable sight record, Cape San Lucas. “I saw the dark leathery fin of a basking shark stick up out of the water. It was immense. But before we could get near enough for a picture it sank”). Cetorhinus maximus, Gudger, E. W., Science, 65, 1927: 545 (locality “not far from the Galapa- gos”). Fowler, H. W., Proc. fth Pac. Sci. Congr., Java 1929 (1930): 489 (range; includes Cali- fornia, Ecuador, Peru). Gudger, Proc. Zool. Soc. London, 1934 (1935): 877 (Contains the following statement . . and not Ceto- rhinus, which latter shark is, however, found all along the western coast of South America, and even in the vicinity of the Galapagos Archi- pelago.”) Discussion: As Norman suggests ( Discovery Reports, 16, 1937: 7) the southern basking shark, as represented by Falkland Island and southern Beebe & Tee-Van: Sharks 99 1941] South American specimens, may prove to be distinct from the northern species. The speci- mens from Peru, Ecuador and the Galapagos Islands may represent a northward projection of the range of this southern race or species of basking shark. The record of basking shark near Cape San Lucas, made by Zane Grey in “Tales of Fishing in Virgin Seas,” page 185, must be questioned until further records of this shark are secured in Gulf of California waters. The record was merely the sight of a dorsal fin which disappeared before Grey could approach near enough to secure a photograph. Family Scylliorhinidae.12 Key to genera and species of the tropical eastern Pacific.13 la. Labial folds absent or rudimentary; belly capa- ble of inflation. (Color grayish-brown, tinged with yellowish above and below; back with black cross-bars; upper parts with large round black spots Cephaloscyllium uter lb. Labial folds present; belly not inflatable. 2a. First dorsal fin originating behind the origin of the pelvics; base of the second dorsal fin shorter and wholly above that of the anal fin, the latter overlapping it before and behind; belly pale Pristiurus xaniurus 2b. First dorsal fin originating well in front of the pelvics; second dorsal fin about equal to the anal in size, and terminating slightly behind it; belly same color as rest of the body Cephalurus cephalus 12 The genera in this family are not easy to differentiate and it is possible that many of them ought to be merged or at least recognized as subgenera. We have here followed Garman, 1913, except for the use of Scylliorhinus in place of Catulus. Halaelurus chilensis (Guichenot) of Chilean and south- ern South American waters, approaches the borders of the region under consideration. It has been recorded from Mollendo, Peru, by Evermann, B. W & Radclifle, L. (Bull. U. S. Nat. A/us., 95, 1917: 3). 13 Apristurus brunneus (Gilbert). The range of this species has been noted as including the Gulf of California. However, this is not true and the species is known, as far as the literature is concerned, only from north of the United States boundary. The history of the false Gulf of California record seems to be as follows: In the original description of Catulus brunneus ( Proc . U. S. Nat. Mus., 14, 1891 (1892): 542), which is included in a paper entitled “Descriptions of thirty-four new species of fishes collected in 1888 and 1889, principally among the Santa Barbara Islands and the Gulf of Cali- fornia,” no mention is made of the type locality of the species. Jordan & Evermann. in “Fishes of North and Middle America, ” Vol 1, page 24, supplied the deficiency by stating “Gulf of California,” and this notation has been copied in a number of papers. Unfortunately, this selection of a type locality is not in accord with the known distribution of brunneus and, in addition, a checking of the type. No. 51,708, U. S. National Museum, reveals that it was taken at Albatross station No. 2396, 32° 49' N. Lat., 117° 28' 30" W. Long., in 359 fathoms, on February 4, 1889. This locality is about 10 miles west of Point La Jolla, California, quite within the generally known range of the species, and a long distance from the Gulf of California. Apristurus brunneus may thus be removed from con- sideration as a Gulf of California species, and the type locality should be changed in the literature to coincide with the notes given above. Cephaloscyllium Gill, 1861. Cephaloscyllium uter (Jordan & Gilbert). Swell Shark. Range: Monterey Bay to northern Lower California, with a single record at Acapulco, Mexico. Field Characters: Small, blunt-headed sharks with first dorsal fin far back on the body, over the pectoral fins; teeth small, tricuspid, in about four series. Grayish-brown with bands and spots of black, sometimes with white spots on the body. Belly capable of great inflation. (Illustration after Jordan & Evermann, 1900: 790 mm.). Size: Grows to about three feet. Study Material: None. References: Catulus uter, Jordan, D. S., & Gilbert, C. H., in Jordan, D. S., & Evermann, B. W., “Fishes of North and Middle America,” 1, 1896: 25 (new name and description, ques- tions relationship to Chilean ventriosum). Scylliorhinus ventriosus, Garman, S., Mem. Mus. Comp. Zool., 24, 1899: 26 (“. . . S. ventriosus Garm., from Acapulco and northward . . . ”). Discussion: All of the records of this species with the exception of one are from the region from Monterey, California, to Ensenada, Mexico. The exception is the record of Garman, listing the species from Acapulco, Mexico. Garman, in the Acapulco record, synonymized the northern form uter, with the Chilean ven- triosus which he had described in 1880. The similarity and dissimilarity of the northern and southern forms had been recognized by Jordan & Gilbert when the name uter was given to the California specimens. Considering the lack of records from the warm waters south of Acapulco, we retain the California form as a species distinct from the South American one. Pristiurus Muller & Henle, 1838. Pristiurus xaniurus (Gilbert). File-tail Shark. 100 [XXVI: 15 Zoologica: New York Zoological Society Range: Coast of southern and Lower California, south to San Roque Bay, in 184 to 684 fathoms. Field Characters: Small sharks with two dorsal fins placed far back on the body, the first dorsal above the ventrals; anal fin present; upper edge of caudal in adult with a broad band of enlarged scales; labial fold of lower jaw slightly shorter than the fold of the upper jaw. Dark slaty brown, usually uniform above, sometimes with small whitish spots, the fins often edged with paler; belly pale. (Illustration after Garman, 1913: 552 mm.). Size: Grows to about two feet. Study Material: None. References: Catulus xaniurus, Gilbert, C. H., Proc. U. S. Nat. Mus., 14, 1891 (1892): 540 (original description, abundance; type locality, Southern and Lower California, in 184 to 684 fathoms). Gilbert, C. H., Rep’t. U. S. Fish Comm., 1893 (1895): 461 (off Central California in 200 to 456 fathoms; spines and spinelets, young, claspers). Jordan, D. S. & Evermann, B. W., Fishes of North and Middle America, 1, 1896: 24 (description). Townsend, C. H., & Nichols, J. T., Bull. Amer. Mus. Nat. Hist., 52, 1925: 5 (18 specimens from 27° 07' N., 114° 33' W., off San Roque Bay, Pacific coast of Lower California, in 284 fathoms). Parmaturus xaniurus, Walford, L. A., Div. Fish and Game California, Fish Bull., 45, 1935: 27 (short description, distribution, figure). Garman, S., The Plagiostomia, 1913: 90, Plate 9, figs. 1-5 (description, figure). Cephalurus Bigelow & Schroeder, 1941. Cephalurus cephalus (Gilbert). Text-figure 7. Range: Gulf of California; 85 to 100 miles N. W. of Cape San Lucas, Lower California, and near Clarion Island, in 85 to 460 fathoms. Field Characters: Very small, rather broad- headed sharks with wide crescentic mouth; second dorsal fin about equal to anal in size, with its posterior base termination behind that of the anal; brown, the belly same color as the rest of the body. (Illustration after Bigelow & Schroeder, 1941.) Size: Grows to about 10 inches. Study Material: 1 specimen. Mexico, off the south-western Coast of California, No. 12,831, Amer. Mus. Nat. Hist., length 206 mm. References: Catulus cephalus, Gilbert, C. IT, Proc. U. S. Nat. Mus., 14, 1891 (1892):, 541 (original description, Gulf of California at 362 fathoms, and near Clarion Island at 460 fathoms). Garman, S., The Plagiostomia, 1913: 78 (short description; comment on inadequacy of the original description; placed provisionally in Catulus). Townsend, C. IT, & Nichols, J. T., Bull. Amer. Mus. Nat. Hist., 52, 1925: 6 (speci- mens from 85 to 100 miles northwest of Cape San Lucas, in 389 to 405 fathoms). Family Triakidae. Key to tropical eastern Pacific genera. la. Spiracles present; no caudal pit Triakis lb. Spiracles absent; caudal pit present . Triaenodon Triakis Muller & Henle, 1838. Key to tropical eastern Pacific species.11 la. Grayish with cross bands and large alternating spots of brown; median cusp of teeth acumi- nate, lateral cusps short semifasciata lb. Slaty brown, with scattered small spots of black ; median cusps of teeth short, broad-based, lateral cusps short maculatus Triakis semifasciatum Girard. Range: Cape Mendocino, California, to Mag- dalena Bay, Lower California. (Mexico: Cedros Island, Port San Bartholomae, Turtle Bay, Ballenas Bay and Magdalena Bay.) Field Characters: An easily recognizable shark: gray with well-defined cross bands of black on the upper parts, interspaced with round black dots along the sides. Size: Grows to about 3 feet in males, females over five feet. Study Material: We have no material. At Clarion Island in the Revillagigedo group, a speci- men supposedly of this species was caught and lost on a feather-fly. The shark was about four feet long. References: Triakis semifasciatum, Girard, C. F., Proc. Acad. Nat. Sci. Phila., 7, 1854: 196 (original description; type locality, Presidio de San Francisco). Lockington, W. N., Proc. Acad. Nat. Sci. Phila., 1881 (1882): (Magdalena Bay). Fowler, H. W., Proc. Acad. Nat. Sci. Phila., 60, 1908: 59 (Use of felis in place of semifasciatum) . Osburn, R. C., & Nichols, J. T., Bull. Amer. ii From Garman, S, 1913: 165. 1941] Beebe & Tee-Van: Sharks 101 Mus. Nat. Hist., 35, 1916: 141 (3 specimens in seines, Cedros Island, Port San Bartholome and Ballenas Bay, Lower California). Wales, J. H., Copeia, 1932: 163 (Ensenada and Turtle Bay, Lower California; many seined at latter locality). Walford, L. A., Div. Fish and Game California, Fish Bull., 45, 1935: 32 (Short description, range, etc., figure). Barnhart, P. S., Mar. Fishes South. California, 1936; 9, fig. 14 (short descrip- tion and color; figure). Triakis maculata Kner & Steindachner. Range: Recorded by Fowler from California, Mexico, Peru and Chile. Field Characters: Small sharks; slaty brown with numerous irregular scattered spots of black on the back and flanks. Teeth small, numerous, more than two rows functioning, each tooth with the central cusp short, broad at base and directed obliquely outward; the lateral denticles of each tooth are mere rudiments separated from the principal cusp by a shallow notch. Size: Grows to about two feet. Study Material: None. References: Triakis maculatus, Kner, R., & Steindachner, F., Sitsb. Akad. IITss. Wien, 54, 1867: 391 (original description, “Sudsee,” Listed as “ Triakis scyllium Dum., vel maculata, n. sp.). Steindachner, F., Sitz. Akad. ITfss. Wien, 60, 1870: 315 (Mazatlan, Mexico). Garman, S., The Plagiostomia, 1913: 167 (Redescription of the species from Callao, Peru, specimens). Fowler, H. W., Proc. fth Pac. Sci. Congr., Java 1929, Biol. Pap. 3; 1930: 490 (Lists following range: California, Mexico, Peru and Chile). Discussion: This species was described by Kner & Steindachner from a specimen taken from the indefinite locality “Sudsee.” However, the other species of fish listed in the same paper are from Chile and Peru, so that it is not un- likely that the types came from the west coast of South America. In 1870 Steindachner listed a specimen from Mazatlan, Mexico, and in 1913 Garman redescribed the species from specimens taken at Callao, Peru. Fowler in 1930 listed the range as we have recorded it. We have been unable to find the Californian and Chilean references upon which the range is based. Triaenodon Muller & Henle, 1837. Triaenodon obesus (Ruppell). (Plate II, Fig. 1). Range: Red Sea, Indian Ocean, Oceania in- cluding Hawaii, Cocos Island and Panama. (In the eastern tropical Pacific three specimens are known, two from Cocos Island and one from Bahia Honda, Panama. The Panama specimen here recorded is the first reported occurrence of the species on the American continent.) Field Characters: A blunt-headed shark with mouth close to and paralleling the outer border of the head ; nostrils with a conspicuous triangular flap; nictitating membrane present; first dorsal fin nearer ventrals than pectorals; teeth with a long median cusp and wTith a smaller cusp, rarely two, at the base on each side; tips of the dorsal fins and the upper caudal lobe white. (Illustra- tion after Fowler, 1928: 645 mm.) Description: The single specimen secured by us agrees with current descriptions.15 For compara- tive purposes we append the following account of this 1,175 mm. (4634 inches) fish. Proportions are stated in percentages of the length to the tip of the last caudal vertebrae (1,125 mm.). Head broad, depressed, the upper surface quite flat; snout broadly rounded, the mouth close to and paralleling the outer border of the head but slightly nearer the sides of the head laterally than anteriorly; preoral length 3.6%, width of mouth 8.5; symphysis of lower jaw to line joining the angles of the jaw 4.7; mouth strongly arched with a small crease on the upper lip on each side and with a deep pit at the gape on each side in the lower jaw. Nostrils to snout 3.7; internarial space 4.6; nostrils with a large conspicuous triangular flap on their inner halves. Eye a broad horizontal oval with a large nicti- tating membrane, diameter 1.85; distance from snout to eye 6.9; distance from eye to mouth 2.7. Snout to first gill-opening 18, snout to fifth gill- opening 22; height of first gill-opening 2.65, height of fifth gill-opening 3.1; last gill-slit over the base of the pectoral fin. Snout to first dorsal fin 38, the fin rather sharp superiorly and with a sharp-pointed posterior lobe; first dorsal fin much closer to pelvics than to the pectorals; base of first dorsal 8.9, anterior edge 12.2, vertical height of fin from body 9.2, base to tip of lower lobe 4.4. Interdorsal space 17.8. Snout to 16 The eastern Pacific records of the genus (Cocos Island, Hawaii) have been considered up to the present, as referable to obesus. Whitley, however (.Australian Zoologist, 9 (3) 1939: 237), suggests that the Hawaiian form as represented by Fowler’s Laysan Island example which the latter called obesus, is an "ally" of obesus. As mentioned above our specimen agrees so closely with the current descriptions of obesus that we see no reason for using another name. 102 Zoologica: New York Zoological Society [XXVI: 15 second dorsal fin 64; base of second dorsal 6.1, anterior edge 8.5, vertical height 6, base to tip of lower lobe 3.8, origin of fin very slightly ahead of the anal. Anal fin from snout 65.5, base 5, anterior edge 9.1, base to tip of posterior lobe 3.55. Base of second dorsal to upper caudal pit 8.35; base of anal to inferior caudal pit 8.35. Upper caudal pit to tip of caudal 24.8; inferior caudal pit to tip of lower caudal lobe 12.7; tip of caudal to notch of terminal lobe 6.75. Base of pectoral from snout 24.6, the fin large and low, its tip somewhat pointed, anterior edge of fin 14.6, inner edge 5, base of pectoral 5. Pelvic fins from snout 50, the fin truncate posteriorly, base of fin 6.1, anterior edge 7.3, inner edge of fin from the inner posterior tip to the separation of the fins anteriorly 6.5. Teeth in both jaws tricuspid (rarely with five cusps), the median cusps much larger than the lateral ones, in about 45 rows in each jaw (the rear teeth difficult to count as the jaws were not dissected and the teeth extend far back into the mouth), those in the upper jaw erect anteriorly, becoming oblique and backwardly inclined on the sides posteriorly. Denticles from the upper side beneath the first dorsal fin are 5 to 7 keeled, the outer keels when 7 are present, very small, the denticles closely packed together. Color: In life dark gray above, shading into dead white below, the tips of the second dorsal fin and the upper lobe of the caudal fin dead white. Irregularly scattered dark spots on the sides and tail, all smaller than the eye. Iris pale green with faint blackish angular lines. Snodgrass & Heller state the color of their Cocos Island fish as follows: “dark uniform slate above, below livid yellowish slate; tip of 1st dorsal and of upper lobe of the caudal creamy white.” Herre gives the color of his specimen from the same locality preserved in alcohol as: “more or less brownish or rusty brown above, paler to whitish beneath; the tips of the dorsals and upper caudal lobe are milky white, the margins of the other fins dark or blackish.” Size: Grows to at least 5 feet in our region. Local Distribution: Our single specimen was speared close inshore at night. Abundance: An uncommon species, represented by three records in the eastern Pacific American waters. Food: The stomach of our fish contained a 205 mm. snapper-like fish. Study Material: 1 specimen. Panama: Bahia Honda, 1, female (26186) 1,175 mm., Mar. 15, 1938, speared. References: Carciharias obesus Ruppell, W. P. E. S., Neue Wirbelthiere zu der Fauna von Abyssinien gehorig, 1835: 64, PI. 18, fig. 2 (original description, figure; type locality, Red Sea). Triaenodon obesus, Snodgrass, R. E., & Heller, E., Proc. Wash. Acad. Sci., 6, 1905: 344 (range, description, color; Cocos Island); Fowler, H. W., Proc. 4th Pac. Sci. Congr., Java, 1929, 3 (1930) : 489 (Check list, range); Herre, A. W., Field Mus. Nat. Hist., Zool. Ser., 21, 1936: 24 (short descrip- tion, color, range; Cocos Island); Fowler, H. W., Acad. Nat. Sci. Phila., Monograph 2, 1938: 249 (check-list, Cocos Island). Discussion: Whitley has recently pointed out that the genus Triaenodon was mentioned earlier than the usually accepted first reference of Muller & Henle (Syst. Beschr. der Plagiostomen, 1838-1841 : 55) . The references referred to are as follows: Muller & Henle, Ber. Verh. kon. Preuss. Akad. Wiss. Berlin, 2, 1837: 113, and Mag. Nat. Hist. (ed. Charlesworth) 2, 1838: 38. In the last two references cited above there is no mention of a genotype. In the first reference, two species are recognized under Triaenodon, obesus and smithii. As Muller & Henle state (footnote p. 56, of last-mentioned reference), that smithii belonged to the genus Leptocarias, it is obvious that obesus must be the genotype. In our specimen of Triaenodon obesus, the nictitating membrane is complete and not a fold, as is indicated in White’s “Key to Galea” (Bull. Amer. Mus. Nat. Hist., 74, 1937: 121) in alterna- tive (55). Family Mustelidae. Mustelus Linck, 1790. 16 Key to species of the tropical eastern Pacific. la. Lower lobe of the caudal fin pointed, arcuate. lunulatus lb. Lower lobe of caudal fin rounded. 2a. Midpoint of base of first dorsal fin much closer to origin of ventral fins than to axil of the pectorals calif ornicus 2b. Midpoint of base of first dorsal fin as close to axil of pectoral fins as to the ventrals or closer dorsalis Mustelus californicus Gill. Gray Smooth Hound. Range: Cape Mendocino, California, south- ward into the Gulf of California (Mexico: “Lower California,” Magdalena Bay, Cape San Lucas, Rio Colorado, Guaymas). Field Characters: Small sharks with teeth in flattened pavement-like rows; preoral length about equal to or a little less than the width of the mouth; lower lobe of caudal fin ending obtuse- We have made no attempt to consider the species of the genus recorded from Peru and southward. Some of these may belong to other genera; they are mcnto ( edulis ), abbotti and nioromaculatus. See Bigelow & Schroeder, (Proc. Boston Soc. Nat. Hist., 41 (8) 1940; 417-438), for data on these forms. 1941] Beebe & Tee-Van: Sharks 103 ly; center of base of first dorsal fin nearer root of ventrals than pectorals. Dark lead gray, white below. (Illustration from specimen No. 443, || Mus. Comp. Zool.; 615 mm.). 1 Size: Grows to about “V/i feet. Study Material: 3 specimens. Mexico: Cape ! San Lucas, 3 (24815, 24821, 24822) female and two embryos about to be born, March 30, 1936, hand line (only embryos saved). References: Mustelus californicus, Gill, T. N., ' Proc. Acad. Nat. Sci. Phila., 16, 1864: 148 (original description, type locality, San Francisco, California). Lockington, W. N., Proc. Acad. Nat. Sci. Phila., 1881: 114 (Lower California). Starks, E. C., Copeia, 46, 1917: 61-63 (compari- son with M. henlei and M. lunulatus; Mexico: Ensenada and Magdalena Bay). Springer, S., Proc. U. S. Nat. Mus., 86, 1939: 468 (in key to genus) . Galeorhinus californicus, Breder, C. M., Jr., Bull. Bingham Oceano. Coll., 2 (1), 1926: 3 (Rio Colorado, Lower California). Galeus dorsalis, Evermann, B. W., & Jenkins, 0. P., Proc. U. S. Nat. Mus., 14, 1891: 129 (Guaymas, Mexico; embryo; see Gilbert & Starks, Fishes of Panama Bay, p. 7). Mustelus dorsalis Gill. Tollo. Range: Gulf of California, southward to Colombia and ? Peru. (Mexico: Gulf of Cali- fornia; Costa Rica: Uvita Bay; Panama: Pana- ma; Peru: Callao.) Field Characters: Small sharks with teeth in flattened, pavement-like rows; preoral length about equal to width of mouth; first dorsal fin with center of its base equidistant from base of pectoral and ventral fins; lower lobe of caudal fin not prominent, ending obtusely; uniform grayish above, whitish below, no conspicuous color markings. (Illustration from specimen No. 26,176: 488 mm.) Size: Grows to about three feet. Study Material: 1 specimen. Costa Rica: Uvita Bay, male (26176) 480 mm., March 2, 1938, taken on hand line. References: Mustelus dorsalis, Gill, T. N., Proc. Acad. Nat. Sci. Phila., 1864: 149 (original description; Panama, type No. 8068, U. S. Nat. Mus.). Gunther, A., Trans. Zool. Soc. London, 6 (7), 1868: 396 (check list), 490 (copied descrip- tion). Gunther, A., Cat. Fishes Brit. Mus., 8, 1870: 388 (note). Jordan, D. S. & Gilbert, C. H., Bull. U. S. Fish Comm., 2, 1882: 109 (name only, Panama). Jordan, D. S., & Gilbert, C. H., Proc. U. S. Nat. Mus., 5, 1882 (1883): 109 (Key differentiating this species from lunulatus and canis). Jordan, D. S., & Gilbert, C. H., Proc. U. S. Nat. Mus., 5, 1882 (1883): 373 (List of Dow’s specimens, type from Panama). Pellegrin, J., Bull. Mus. Hist. Nat., Paris, 7, 1901: 161 (Gulf of California). ? Evermann, B. W., & Radcliffe, L., U. S. Nat. Mus., Bull. 95, 1917: 7, plate 1, fig. 3, plate 2, fig. 1 (measurements, description, figures). Nichols, J. T. & Murphy, R. C., Bull. Amer. Mus. Nat. Hist., 46, 1922: 504 (Callao, Peru). Springer, S., Proc. U. S. Nat. Mus., 86, 1939: 467 (in key). Galeus dorsalis, Jordan, D. S., Proc. U. S. Nat. Mus., 8, 1885 (1886): 363 (check-list, Panama). Jordan, D. S., & Bollman, C. H., Proc. U. S. Nat. Mus., 12, 1889 (1890) : 179 (name only). Gilbert, C. IL., Proc. U. S. Nat. Mus., 13, 1890 (1891): 449 (Panama). Jordan, D. S., & Evermann, B. W., Fishes North and Middle America, 1, 1896: 30 (description, range, color). Gilbert, C. H., & Starks, E. C., Fishes of Panama Bay, 1904: 7, Plate 1, fig. 2 (comments on description, measure- ments, figure of head). Osburn, R. C. & Nichols, J. T., Bull. Amer. Mus. Nat. Hist., 35, 1916: 141 (Tiburon Island, Gulf of California; three-foot female with six unborn ten-inch young) . Galeorhinus dorsalis, Garman, S., The Plagio- stomia, 1913: 178 (synonymy, description, color). Meek, S. E. & Hildebrand, S. F., Marine Fishes of Panama, 1, 1923: 32 (short synonymy, note on embryos, description; Panama market). Mustelus lunulatus Jordan & Gilbert. Range: Southern California to Colombia. (Mexico: Santa Inez Bay, Concepcion Bay, Guaymas, Mazatlan; Costa Rica: Port Parker; Panama: Panama Bay; Colombia: Gorgona Island.) Field Characters: Small sharks with teeth in flattened pavement-like rows; preoral length greater than width of mouth; center of base of first dorsal closer to base of pectoral fins than to pelvics; lower lobe of caudal fin prominent, end- ing in a point. Uniform grayish above, pale below. (Illustration after Kumada & Hiyama, 1937: 472 mm.) Size: Grows to 5 feet 8 inches. Study Materials: 6 specimens. Mexico: Santa Inez Bay, Lower California, 2, April 13, 1936 (not saved); Santa Inez Bay, 1, 4 feet long; April 23, 1936 (not saved); Concepcion Bay, 1 (24994), April 13, 1936. Costa Rica: Port 104 [XXVI: 15 Zoologica: New York Zoological Society Parker, 1 (26114), 700 mm., Jan. 16, 1938, captured on hook, teeth saved; Colombia: Gor- gona Island, 1 (26211), 890 mm., March 27, 1938. References: Mustelus lunulatus, Jordan, D. S. & Gilbert, C. H., Proc. U. S. Nat. Mus., 5, 1882: 108 (original description; type locality: Mazatlan, Mexico; type No. 29211, U. S. Nat. Mus.). Jordan, D. S. & Gilbert, C. H., Bull. U. S. Fish Comm., 2, 1882 (1883): 105 (check-list, Mazatlan, Mexico). Jordan, D. S. & Evermann, B. W., Fishes North and Middle America, 1, 1896: 28 (description) . Jordan, D. S. & Evermann, B. W., Fishes North and Middle America, 3, 1898: 2745 (in key to west coast Mustelus and Galeus) . Jordan, D. S. & McGregor, R. C., Rep. Comm. Fish, for 1898 (1899): 274 (Ensenada, Lower California, short description). Gilbert, C. H. & Starks, E. C., Fishes of Panama Bay, 1904: 5, Plate 1, fig. 1 (comparison with northern speci- mens, proportions, measurements, etc., figure), 207 (check-list, range). Starks, E. C., Copeia, 46, 1917: 63 (comparison with Mustelus henlei and Mustelus calif or nicus). Norris, H. W., Copeia, 114, 1923: 1 (California, size, abundance, young). Kumada, T. & Hiyama, Y., Marine Fishes Pacific Coast of Mexico, 1937: 16, Plate 47 (Mexico; figure). Springer, S., Proc. U. S. Nat. Mus., 86. 1939: 464 (comparison with norrisi, etc.), 467 (in key to Mustelus). Galeus lunulatus, Jordan, D. S., Proc. U. S. Nat. Mus., 8, 1885 (1886): 363 (check-list; name only)- Evermann, B. W. & Jenkins, O. P., Proc. U. S. Nat. Mus., 14, 1891 (1892): 128 (Guaymas, Mexico). Jordan, D. S., Fishes of Sinaloa, 1895: 382 (Mazatlan, abundance). Galeorhinus lunulatus, Garman, S., The Plagio- stomia, 1913: 174 (description, color, range). Meek, S. E. & Hildebrand, S. F., Marine Fishes of Panama, 1, 1923: 33 (short synonymy, description, comparison with dorsalis; range; Panama) . Cynias lunulatus, Starks, E. C. & Morris, E. L., Univ. Cal. Publ. Zool., 3 (11), 1907: 163, 164 (comparison with M. calif ornicus) . Family Galeorhinidae. Key to genera of the tropical eastern Pacific, la. Spiracles absent. 2a. Lower labial folds wanting or rudimentary. 3a. First dorsal fin inserted posteriorly, the mid-point of its base nearer ventral than pectoral fins Prionace 3b. First dorsal fin inserted anteriorly, the mid-point of its base nearer pectoral than ventral fins. 4a. Second dorsal fin almost as large in area as the first; teeth of both jaws smooth, narrowly triangular, with wide basal shoulders, the shoulders extending on each side of the central triangular portion Aprionodon 4b. Second dorsal fin very much smaller than the first dorsal; teeth of at least the upper jaw serrated, more or less triangular Eulamia 2b. Labial folds well developed, present on both jaws; teeth not serrate Scoliodon lb. Spiracles present, small, situated behind the eye. 5a. Teeth large, alike in both jaws, deeply notched on the outer edge and convex on the inner, heavily serrated; body in small and medium sized fish usually with irregu- lar dark spots on the sides which often coalesce to form bars; these spots and bars tend to disappear with age. . . .Galeocerdo 5b. Teeth not deeply notched, curved on both sides; color and pattern not as above Galeorhinus Prionace Cantor, 1849. Prionace glauca (Linnaeus). Text-figure 13. Range: Tropical seas throughout the world; on the Pacific coast of America from Puget Sound southward to the Gulf of California; on the southern border of the tropical eastern Pacific it is recorded from Chile. (Mexico: Carmen Island, Gulf of California.) Field Characters: Large, elongate sharks with short head and long tapering snout; first dorsal fin closer to pelvic fins than to pectorals; upper teeth triangular, convex externally, concave in- ternally, all strongly serrated. Color above rich deep blue. (Illustration after Garman, 1913.) Size: Grows to 15 to 20 feet. Study Material: None. References: Galeus glaucus, Linnaeus, Systema Naturae, 10th Ed., 1, 1758: 235 (original descrip- tion) . Prionace glauca, Osburn, R. C., & Nichols, J. T., Bull. Amer. Mus. Nat. Hist., 35, 1916: 141 (young specimen, hand-line, Carmen Island, Gulf of California). Discussion: With the exception of the one Gulf of California specimen, records of this species are conspicuous by their absence in our region. The species is recorded north and south of the tropical eastern Pacific, both along the United States coast and in Chile. Fowler has synonymized the Chilean Car- charias pugae Perez, C. gracilis Philippi and C. aethiops Philippi with this species. In addition to these Chilean records there are a number of others under the name of glaucus. Whitley17 uses the name Carcharhinus in place of Prionace for the blue sharks. The former n Whitley, Fishes of Australia, Part 1, The Sharks. 1940: 107. Beebe & Tee-Van: Sharks 105 1941] name has been hurled about to such a degree that we prefer retaining Prionace for the present for this genus. Aprionodon Gill, 1861. Aprionodon fronto (Jordan & Gilbert). Range: Mexico, Costa Rica. (Mexico: Guay- mas, Mazatlan; Costa Rica: Port Parker. Our Port Parker specimen extends the range of this species some 1,500 miles southeastward along the coast.) Field Characters: A shark with the second dorsal fin almost as large as the first, its base 4/5th as long as that of the first, the fin larger than the anal and originating slightly further forward; pectoral fin broad, its breadth 70% of the length of the fin; teeth of both jaws narrowly triangular with wide basal shoulders, edges of teeth smooth. (Illustration from Specimen No. 26,116:704 mm.) Description: The proportions, expressed in percentages of the total length, of the specimen obtained by the Zaca, are as follows: Specimen No. 22,116, 704 mm. total length. Length to caudal notch 76%; snout to origin of first dorsal fin 36.2; snout to origin of second dorsal fin 63; snout to origin of pelvic fins 50.5; snout to origin of anal fin 64.2; vertical of posterior pectoral base to origin of first dorsal fin 6.95; vertical of inner tip of pectoral fin to origin of first dorsal fin 2.8; First dorsal fin base 9.5, anterior edge 11.2, posterior edge 4.25, vertical height from body 6.4. Interdorsal space 19.2. Second dorsal fin base 6.95, anterior edge 8.7, posterior edge 3.7, vertical height of fin from body 6. Caudal fin length 22.5, tip to origin of terminal lobe 5.7, length lower lobe 11.2. Pectoral fin base 6.1, anterior edge 14.8, posterior edge 6.5, greatest breadth 10.7. Pelvic fin base 6.25, anterior edge 7.6, posterior edge 3.62. Anal fin base 5.35, anterior edge 7.7, posterior edge 3.05, vertical height from body 5. Snout to eye 7.95, eye diameter 1.8. Snout to mouth 5.26, mouth width 9.2, symphysis to line joining angles of the mouth 5.1. Snout to outer angle of nostrils 4.83, internarial space 5.55, nostril to mouth 2.7, length of nostril 1.63. Snout to 1st gill-slit 19.5, snout to 5th gill-slit 23.3. Height of 1st gill-slit 4.1, height of 5th gill-slit 4.1. Teeth of both jaws narrowly triangular with broad, shoulder-like basal portion and with no trace of serrations. Laterally the teeth tilt backwards, increasingly so posteriorly. The teeth farthest back, because of this tilting, have the posterior basal portion accentuated. The edges of the teeth are translucent and rather delicate and a ragged broken edge can be pro- duced by the slightest effort. Color: In life the upper surfaces are in general yellow-green; two broad dark bands begin at the nape and unite in back of the second dorsal fin. Below pale green, becoming almost white on the lower snout, chin and between the pectoral fins. Size: The three known specimens of this species are small, 704, 723 and 915 mm. (27 to 36 inches). Food: Our specimen contained the remains of an unidentifiable fish about 115 mm. long. Local Distribution: Seined close inshore in daylight over a sandy beach. Study Material: 1 specimen. Costa Rica: Port Parker, 1 female (26,116) 704 mm., Jan. 18, 1938, seined. References: Carcharias fronto, Jordan, D. S. & Gilbert, C. H., Proc. U. S. Nat. Mus., 5, 1882: 102-103, not large specimen mentioned on page 104 (description; type-locality: Mazatlan, Mex- ico; type No. 28167, U. S. Nat. Mus.). Jordan, D. S. & Gilbert, C. H., Bull. U. S. Fish Comm., 2, 1882: 105 (Mazatlan, Mexico; common name). Eulamia fronto, Evermann, B. W., & Jenkins, O. P., Proc. U. S. Nat. Mus., 14, 1891 (1892): 129 (Guaymas, Mexico; 28 inch specimen). Carcharinus fronto, ? Jordan, D. S., Fishes of Sinaloa, 1895: 382 (erroneous note referring to a large, man-eating shark). Jordan, D. S. & Evermann, B. W., Fishes North and Middle America, 1, 1896: 39 (description; length errone- ously given as 10 feet). Jordan, D. S., Proc. U. S. Nat. Mus., 8, 1885 (1886): 363 (check-list). Carcharinus milberti, Garman, S., The Plagio- stomia, 1913: 133 (synonymy only, in part). Carcharhinus milberti, Meek, S. E. & Hilde- brand, S. F., Marine Fishes of Panama, 1, 1923: 38 (synonymy only, in part). ? Etdamia plumbea, Fowler, H. W., Proc. 4th Pac. Sci. Congr., Java, 1929 3, 1930: 492 (? Mazatlan record). Discussion: The 704 mm. specimen that we refer to this species agrees well with the descrip- tion of the 915 mm. type, except for the following : there is but one gill-opening definitely over the pectoral base, and the anterior margin of the pelvic fins is shorter than the distance between the angles of the mouth. However, the general agreement is so close that we have no hesitation in assigning our Costa Rican specimen to fronto. In the original description of fronto, based on an adult and a young shark, it is obvious that two species are represented, the first half of the description referring to a fish with an exception- ally large second dorsal fin and narrowly triangu- lar, broad-based teeth. The second half of the 106 [XXVI: 15 Zoologica: New York Zoological Society description, in which the body-measurements are admittedly inaccurate, as they were made with- out instruments, is of a ten-foot shark with much smaller second dorsal fin and serrated teeth. The latter fish may represent a specimen of Eulamia azureus. We here delimit the first half of this description, referring to the small specimen designated as the type, as the original description of fronto. Our transfer of the species to the genus Aprionodon is based upon the following. Jordan & Gilbert in the original description of fronto state: “Edges of teeth appearing minutely serrulate under a lens.” This condition is not true of our specimen as even under fairly high power the tooth edges are entire except where their rather delicate borders have become irregu- larly nicked from some external agency. In order to check on this discrepancy, Dr. Leonard P. Schultz of the U. S. National Museum was asked to examine the teeth of the type speci- men of C archarias fronto. He reports: . . the teeth do not have serrations, only rough here and there from some external cause.” We thus have evidence from the type and an additional shark that the teeth are smooth and we consequently assign the species to Aprionodon. It may be mentioned that Aprionodon fronto differs considerably in appearance from the species of Eulamia found along the tropical Pacific coast, the similarity in size of the two dorsal fins giving the fish a distinctive appear- ance. In this it resembles the Atlantic Hypoprion brevipinnis and some of the western Pacific species of Aprionodon. Probably as the result of true fronto being considered as the young of a large, serrated- toothed shark (a natural conclusion, considering the inference supplied by the two specimens in the original description), Garman synonymized this species and the eastern Pacific Eulamia azureus with the Atlantic Ocean Eulamia milberti. In this he has been followed by other authors. Eulamia Gill, 1861. Key to tropical eastern Pacific species.18 la. Sides of the body with a band-like continuation of the dark color of the upper surfaces ex- tending backward along the sides to above the pelvic fins, enclosing above it a section of the white of the underparts; tips of fins black; teeth of both jaws narrowly triangular with a broad shoulder-like base, all of the teeth distinctly serrated on their margins; lower teeth considerably narrower than the upper. aethalorus lb. Color pattern not as above. Teeth, especially those of the upper jaw, not as above, either broadly triangular, or with one side notched, or with a broad shoulder-like base on one side only. 2a. Snout exceptionally long and thin; inter- narial space equal to or only slightly exceed- ing the length of one of the nostrils; nostril openings almost transverse velox 2b. Snout not especially long and narrow; inter- narial space at least two and a half times or more as broad as the length of a nostril. 3a. Origin of the 2nd dorsal fin opposite or be- hind the vertical of the middle of the anal fin cerdale 3b. Origin of the 2nd dorsal fin opposite or in advance of the origin of the anal fin. 4a. Snout short and bluntly rounded, the preoral portion 1.66 to 1.9 in the dis- tance between the angles of the mouth ; origin of the 2nd dorsal fin conspicu- ously in advance of that of the anal fin azureus 4b. Snout not as short and blunt, the preoral portion slightly less to slightly longer than the width of the mouth, not as short as mentioned above; origin of the 2nd dorsal fin slightly in advance of that of the anal, or directly above. 5a. Teeth finely serrate, not notched; ori- gin of the 2nd dorsal fin slightly in advance of the origin of the anal fin. lamiella 5b. Teeth coarsely serrate, usually notched; origin of 2nd dorsal fin approxi- mately above that of the anal. 6a. Tips of some of the fins white. platyrhynchus 6b. Tips of the fins dusky, never white. galapagensis Eulamia aethalorus (Jordan & Gilbert). Cazon. (Plate II, Fig. 2). Range: Pacific mainland from Lower California and the Gulf of California to Peru. (Mexico: Concepcion Bay, La Paz, Arena Bank, Mazatlan; Guatemala: Chiapam; Panama: Panama Bay; Peru: Lobos de Afuera Island). Field Characters: A medium-sized shark with conspicuous black tips to the fins and a dark band along the sides as far back as the pelvic fins, enclosing above it a section of the white underparts. Teeth of the jaws narrowly tri- angular with broad, shoulder-like basal portions; upper teeth finely serrate, lower teeth almost 18 The difficulties of identifying some of the species of this genus have long been manifest, and during the Zaca and Arcturus expeditions this fact was appreciated more than ever when sharks too large to preserve were caught. While working over our preserved material, it became even more evident that the synonymies and concepts of the species of Eulamia of the Pacific were far from being clean-cut and exact. In an attempt to straighten out the confusion that existed in our minds while reviewing the literature of the sharks of the eastern tropical Pacific, we decided to dis- card all references that merged the eastern Pacific forms with species of supposedly world-wide distribution and to start over again with types and original descriptions of the species that have been described from the region under discussion. We have not attempted to recorrelate the eastern Pacific forms with species from other regions. These littoral species seem to represent valid local species. The change with age of many characters in this group is considerable, and because of this we have given the exact length of each specimen discussed. Two imperfectly described species that apparently be- long to the genus Eulamia are described from the west coast of South America. We have not been able to satisfy ourselves as to their relationships. They are Carcharias robustus Philippi (Ann. Univer. Chile AYS, 1896: 389) and Eulamia philippi Fowler, described as Carcharias brachyrrhynchus by Philippi (Ann. Univer. Chile, 71, 1887: 540) and renamed by Fowler. 1941] Beebe & Tee-Van: Sharks 107 smooth. (Illustration after Meek & Hildebrand, 1923: 850 mm.) Color: Dusky bronze above; dead white on lower snout, head and lower surfaces of the pectoral fins; belly, first dorsal and anal fins grayish; the dusky bronze covers the upper surface of the pectoral fins and extends back in a long narrowing band along the lower sides to above the pelvic fins, enclosing a wider band of whitish from the sides of the peduncle forward, dying out at the vertical of the center of the first dorsal fin. Pelvic fins above, and caudal fin bronzy dusky like the back. Paired fins white below with jet black tips, the black extending down the posterior edge of the pectorals. Pos- terior edge of the first dorsal black; distal half of second dorsal black; posterior edge of caudal black. Anal fin pale with a large black spot at the tip. Iris pale brassy. (2,070 mm. fish.) In the embryos from the specimen just de- scribed, the color pattern of the sides is even more marked than in the adult. They also possess conspicuously black-tipped fins. Size and Weight: Grows to at least seven feet. A 2,070 mm. (81 inches) shark weighed 123 pounds. A 630 mm. embryo weighed 3 pounds. Parasites: Eighteen copepod parasites, Alebi- aon sp. taken from this species plus a specimen of Racinela aries, from the gills. Food: This shark took dolphin-fish bait while it was associated with several other sharks in the vicinity of a school of hundreds of large crevalle ( Caranx caninus). In the stomach of the shark were 24 California sardines ( Sardinia caerulea). Breeding: Four embryos averaging 630-650 mm. in length were taken from the 2,070 mm. adult; they were close to being born. These full-grown embryo sharks were roughly one-third the total length of the adult. Twenty- four comparative measurements of mother and young showed a slight increase in the embryos in relative lengths of the predorsal part of the fish, such as snout to mouth, snout to pectoral, snout to first dorsal fin. The eye was somewhat larger. The pectorals were slightly smaller as was the second dorsal, and the depth was proportionately less. On the whole, however, there was an astonishing agreement in the proportions of the mother and unborn offspring. This absence of marked dissimilarity in pro- portions seems to find its explanation in the total lack of larval or real adolescent life. These embryos, when freed from the mother and the umbilical cord, straightway swam off, showing perfect correlation, avoiding the sides and bottom of a large tub, snapping at anything offered to them and apparently functioning in almost every way that a shark requires in order immediately to begin a successful career in the open sea. At the first impact of the outside world these embryos are perfect sharks in miniature, quite unlike the condition of those fish which hatch from eggs near shore or the bottom of shallow waters. Study Material: 4 specimens. Mexico: Arena Bank, Lower California, 3, adult female (teeth saved) and two young (25,471, 25,472, 25,472B.), 2,070, 630, 650 mm., April 30, 1936, adult captured on rod and line with sailfish bait; 1, Concepcion Bay, Lower California (U. S. Nat. AIus. 46851), collected by the Albatross. References: Carcharias aethalorus, Jordan, D. S. & Gilbert, C. H., Proc. U. S. Nat. Mas., 5, 1882: 104 (original description, color; type- locality: Alazatlan, Mexico). Jordan, D. S. & Gilbert, C. H., Bull. U. S. Fish Comm.., 2, 1882: 105 (Mazatlan, Alexico), 109 (Panama). Gil- bert, C. H. & Starks, E. C., Fishes of Panama Bay, 1904: 9 (teeth, claspers, proportions; Panama), 207 (distribution, Gulf of California). Carcharhinus aethalorus, Jordan, D. S., Proc. U. S. Nat. Mus., 8, 1885 (1886): 363 (check-list). Jordan, D. S., Fishes of Sinaloa, 1895: 383 (relationship with limbatus). Jordan, D. S. & Evermann, B. W., Fishes of North and Middle America, 1, 1896: 40 (description, range). Carcharinus aethalorus, Garman, S., Bull. Mus. Comp. Zool., 46 (12) 1906: 229 (Panama). Carcharias limbatus, Giinther, A., Cat. Fishes Brit. Mus., 8, 1870: 373-374 (Reference to speci- men from Chiapam only: Guatemala; refers to maculipinnis reference) . Carcharinus limbatus, Garman, S., The Plagio- stomia, 1913: 127 (synonymy referring to aethalorus ) . Carcharhinus limbatus, Pellegrin, J., Bull. Mus. Hist. Nat. Paris, 7, 1901: 161 (Gulf of California). Nichols, J. T. & Murphy, R. C., Bull. Amer. Mus. Nat. Hist., 46, 1922: 504 (jaw from Lobos de Tierra, Peru). Meek, S. E., & Hildebrand, S. F., Marine Fishes of Panama, 1, 1923: 41 (description; synonymy referring to aethalorus; Panama) . Carcharinus natator, Meek, S. E. & Hildebrand, S. F., Marine Fishes of Panama, 1, 1923: 41, Plate 1, fig. 1 (original description, color, figure; type locality, Panama City, Panama; type No. 79310, U. S. Nat. Mus.). Carcharias maculipinna (not of Poey) Gunther, A., Trans. Zool. Soc. London, 6, 1868: 490, teeth, dimensions; Guatemala, Chiapam. Discussion: We have two embryos of this species at hand that agree excellently with the original description of aethalorus, and the teeth, 108 [XXVI: 15 Zoologica: New York Zoological Society field description, measurements and photo- graphs of the 2,070 mm. parent; these specimens were taken at Arena Bank, Lower California, 160 miles N. by W. of Mazatlan, the type locality of the species. The embryos differ from the original description in having the base of the first dorsal fin slightly greater than the height (the opposite is true of the parent) and in the caudal fin being 29% of the length instead of 25% (the parent has the same measurement 27%). Comparing the adult female with the original description of aethalorus, which is of a male 30 inches or 762 mm. in length, the following slight differences are observed: the preoral distance is 81% of the mouth width instead of being equal; the teeth are ^ instead of — - (Panama specimens 29 are recorded as ^ ; the base of the first dorsal is 1.2 in the interorbital width, not equal; the second dorsal fin base is 5.2 instead of 4 in the interdorsal space ; the base of the 2nd dorsal fin is 36% (instead of 50%) of the base of the first dorsal (in our embryos this distance is approxi- mately 50%). Carcharinus natalor, described by Meek & Hildebrand from Panama Bay, we believe to be identical with aethalorus. The characters differ- entiating the two forms overlap when our em- bryos and large specimens are compared with the descriptions. Etdamia aethalorus and natator both possess the same type of teeth, general proportions and relationships of fins. It must be admitted that our specimens do not show the peculiar form of the snout, when viewed from the side, that is shown for natator in Meek & Hildebrand’s original illustration of their type. This peculiar conformation may have been caused by preservation. As far as coloration is concerned the two forms are the same. Both have conspicuous black- tipped fins and possess a distinctive color pattern on the sides. In the original description of ■natator two ill-defined dark bands are mentioned, and in our two Gulf of California specimens these bands are strongly evident. However, the upper- most of these two bands is merely an intensifica- tion of the lower border of the dark pigmentation of the dorsal surfaces and in the adult female this band has completely merged with the dark of the upper surfaces, leaving only the shorter, lower band conspicuously outlined. This band is mentioned above in the description under Color and can indistinctly be seen in the illustration; it was plainly visible in life and is so mentioned in our field notes. This shark is closely related to Eulamia limbatus of the Atlantic. Eulamia velox (Gilbert). (Plate II, Fig. 3). Range: Lower California, Costa Rica and Panama Bay. (Mexico: Santo Domingo Point and Cape San Lucas, Lower California; Costa Rica: Port Culebra; Panama: Panama Bay. Previously known from three specimens taken in Panama Bay ; our two specimens extend the range some 2,100 miles northeastward along the coast.) Field Characters: A small elongate shark with a long narrow snout; nostrils large, the inter- narial space narrow, about equal, more or less, to the relatively large nostril openings. (Illus- tration after Gilbert & Starks, 1904: 1200 mm.) Color: Bronzy brown above, changing to silvery iridescence and dead white below. Second dorsal fin with a dusky tip. Size: The largest recorded specimen is 1,200 mm. (47)/£ inches). Food: The stomach of our Santa Domingo specimen contained the chelae of a crab, Ovalipes punctatus. Study Material: 3 specimens. Mexico: Santo Domingo Point, Gulf of California, 1 female (25264), 945 mm., April 16, 1936, hand-line; San Lucas Bay, Lower California, 1, April 23, 1936, hand-line. Costa Rica: Port Culebra, 1 female (26134) 735 mm., Jan. 25, 1938. References: Carcharhinus velox, Gilbert, C. H., in Jordan, D. S. & Evermann, B. W., Fishes North and Middle America, 3, 1898: 2747 (original description, color; type locality, Pana- ma; type, No. 11893, University Museum, Stan- ford University). Meek, S. E. & Hildebrand, S. F., Marine Fishes of Panama, 1, 1923: 45 (description, Panama market). Carcharias velox, Gilbert, C. H. & Starks, E. C., Fishes of Panama Bay, 1904: 5 (deposition of type) 9, Plate 1, fig. 3 (copy of original descrip- tion, figure), 207 (distribution). Carcharinus velox, Garman, S., The Plagio- stomia, 1913: 130 (description). Discussion: These are the fourth, fifth and sixth recorded specimens of this rare species. The previously known ones are the type, taken by Beebe & Tee-Van: Sharks 109 1941] Gilbert in Panama Bay, and two specimens taken by Meek & Hildebrand, also at Panama. The present specimens agree perfectly with Gilbert’s description of a 1,200 mm. fish and with the description given by Meek & Hildebrand. Eulamia cerdale (Gilbert). Text-figure 17. Range: Mexico to Ecuador and the Galapagos Islands. (Mexico: “coast of Mexico” — Kumada & Hiyama; Panama: Panama Bay, and at sea 130 miles S. W. of Burica Point, Panama; Colombia: Buenaventura; Ecuador: Guayaquil: Galapagos Islands, South Seymour and Albe- marle. Also known, according to Meek & Hildebrand, from the Atlantic side of the Isthmus of Panama.) Field Characters: A small shark lacking con- spicuous folds and grooves about the angles of the mouth and with the origin of the 2nd dorsal fin at or behind the vertical of the middle of the anal fin; teeth serrate, those in the upper jaw broader and more oblique than those in the lower and with broader bases; a rather prominent notch behind the large triangular cusp of the lateral teeth in the upper jaw. Bluish-gray above, pale below. (Illustration after Kumada & Hiyama, 1937: 537 mm.) Size: Herre’s largest specimen was 1,235 mm. (49 inches). This, judging by Meek & Hilde- brand, is a small species. Study Material: 1 specimen, Panama (U. S. Nat. Mus. 50438) collected by Gilbert. References: Carcharinus sp. indes., Jordan, D. S., Proc. U. S. Nat. Mus., 8, 1885 (1886): 363 (name only; specimens destroyed by fire before publication of description). Carcharhinus cerdale, Gilbert, C. H., in Jordan, D. S. & Evermann, B. W., Fishes of North and Middle America, 3, 1896: 2746 (original descrip- tion, color, abundance, comparison with aethalo- rus; type locality : Panama ; type No. 11,884, Uni- versity Museum, Stanford University). Meek, S. E. & Hildebrand, S. F., Marine Fishes of Panama, 1, 1923: 47 (description, synonymy, color, discussion of Atlantic coast relative, range) . Carcharias cerdale, Gilbert, C. H. & Starks, E. C., Fishes of Panama Bay, 1904: 5 (number and disposition of type), 10, Plate 2, fig. 4 (description, color, abundance, comparison with aethalorus, figure; Panama). Starks, E. C., Proc. U. S. Nat. Mus., 30, 1906: 762 (range), 763 (Ecuador) . Carcharinus cerdale, Garman, S., Bull. Mus. Comp. Zool., 46, 1906: 229 (Panama). Herre, A. W., Field Mus. Nat. Hist., Zool. Ser., 21, 1936: 22 (synonymy, short description, color, size; Galapagos Islands and at sea) . Charcharhinus cerdale, Wilson, C., Ann. Cam. Mus., 10. 1916: 58 (Colombia, Ecuador). Eulamia cerdale, Fowler, H. W., Acad. Nat. Sci. Phila., Monograph No. 2, 1938: 249 (men- tioned in check-list of GaMpagos Island fishes). Carcharinus menisorrah (in part) Garman, S., The Plagiostomia, 1913: 135 (synonymy referring to cerdale only) . Carcharhinus lamiella (not of Jordan & Gilbert), Kumada, T., & Hiyama, Y., Marine Fishes Pacific Coast of Mexico, 1937: 16, Plate 48 (short description, referring to almost any shark; figure referable to cerdale). Discussion: The figure of Kumada & Hiyama agrees almost entirely with the original figure of cerdale, and even though this record extends the distribution of the species to an indefinite locality on the coast of Mexico, we consider this extension justified. Eulamia azureus (Gilbert & Starks). Range: Mexico ?, Costa Rica, Panama and Ecuador. (? Mexico: Mazatlan; Costa Rica: Piedra Blanca Bay; Panama: Panama Bay; Ecuador: Guayaquil.) Field Characters: A large shark with broadly rounded head and with the origin of the 2nd dorsal fin well in advance of that of the anal; teeth of the upper jaw broadly triangular with their inner margins oblique and usually slightly convex, the outer margins concave or sometimes with a very slight notch; lower jaw with much narrower, triangular teeth on a broad base ; upper jaw teeth serrate along entire border, lower teeth more finely serrate, the serrations less prominent and sometimes absent on the shoulders of the teeth. (Illustration after Gilbert & Starks, 1904: 920 mm.) Color: Dark gray above, white below; fins with dusky tips. Size and Weight: Our 2,820 mm. (9 feet, 3 inches) shark weighed 210 pounds. Food: This species at times feeds upon sting rays, as the spines of four of these animals were found embedded in the skin of the jaws of our shark. Parasites: Two leeches, Pontobdella muricata (Linn.), were taken from our Piedra Blanca 110 shark. A copepod parasite, Rocinela aries, taken from the gills. Study Material: 2 specimens. Costa Rica: Piedra Blanca Bay, 1 male (26148), 2,820 mm., Feb. 4, 1938, harpooned. Ecuador: Guayaquil, 1 male (U. S. Nat. Mus. 53528) 1025 mm. References: Carcharias azureus, Gilbert, C. H. & Starks, E. C., Fishes of Panama Bay, 1904: 5 (type and disposition), 11, Plate 2, fig. 5 (original description, color, figure; type locality, Panama; type No. 11884, University Museum, Stanford University; comparison with C. nica- raguensis; size of male with undeveloped clasp- ers), 207 (distribution). Starks, E. C., Proc. U. S. Nat. Mus., 30, 1906: 762 (distribution), 763 (Guayaquil, Ecuador; specimen compared with type). Garman, S., Bull. Mus. Corny. Zool., 46, 1906: 229 (Panama). Carcharias milberti, Garman, S., The Plagio- stomia, 1913: 133 (references referring to azureus only; not description). Meek, S. E. & Hilde- brand, S. F., Marine Fishes of Panama, 1, 1923: 38 (references referring to Pacific specimens and description of Ecuadorian skin on pp. 39-40; description) . Eulamia plumbea, ? Fowler, H. W., Proc. fill Pac. Sci. Congr., Java 1929, 3, 1930: 492 (? Panama record). Discussion: We have had for comparison with our Costa Rican specimen, the 1,025 mm. shark recorded by Starks from Ecuador; we find that they are the same. The Ecuadorian shark was compared with the type by Starks and partially redescribed by Meek & Hildebrand. The large specimen recorded under the original description of Carcharias fronto (Proc. U. S. Nat. Mus., 5: 103-104) may be this species. As is stated, all of the measurements were taken without instruments and are questionable. However, there are no strong points of difference and the description of the teeth is in agreement with our specimen. This species is closely related to the Atlantic Eulamia milberti. Eulamia lamiella (Jordan & Gilbert). Bay Shark; Tiburon. Range: Southern California to the Gulf of California. (Mexico: Concepcion Point and Mazatlan.) [XXVI: 15 Field Characters: Broad-snouted shark growing to 15 feet, with origin of the 2nd dorsal fin slightly in advance of that of anal, preoral por- tion of the snout slightly less or slightly greater than the width of the mouth; teeth triangular, not notched, with fine serrations; grayish, tips of the pectoral fins slightly dusky. (Illustration from specimen No. 25263: 840 mm.) Color: Dark gray above, grayish-white below; no markings; tips of the pectorals slightly dusky, the upper side dark; all other fins plain. The color of six embryos averaging 840 mm. in length, was as follows: Above slaty -black with sheen; dead white below; sides iridescent, changing from purplish to bronze to silvery ; first dorsal dark bronze-gray, upper posterior margin broadly black; second dorsal same with the distal half black, dark at base with large, black distal spot, white below with corresponding distal black spot; pelvics pale gray with black tip; anal same with larger black spot; caudal with entire outline narrowly jet black, tips of lobes black, the rest gray. Iris silvery with greenish tinge. Size and Weight: Grows to 15 feet. A 2,616 mm. (8 foot 6 inch) shark weighed 270 pounds. Abundance: Rare north of San Diego, fairly common in San Diego Bay and southward. Parasites: Five copepod parasites, Alebion sp., taken from this species. Breeding: Our 2,616 mm. female had six young on April 16, 1936, about to be born. The young averaged around 840 mm. in length and were from 4 pounds 4 ounces to 8 pounds 12 ounces in weight. The litter was composed of one male and five females. Study Material: 3 specimens. Mexico: Con- cepcion Point, Lower California, 1 adult female, not saved (25262) and 1 embryo, saved (25263), 2,616 and 840 mm., April 16, 1936, caught on hook and line. The type in the U. S. National Museum was also examined. References: Eulamia lamia, Jordan, D. S., & Gilbert, C. H., Proc. U. S. Nat. Mus., 4, 1881: 32 (First recording of specimen which later became the type of lamiella). Carcharias lamiella, Jordan, D. S., & Gilbert, C. H., Proc. U. S. Nat. Mus., 5, 1882: 110 (original description, color; type locality, San Diego, California; type, No. 27366, U. S. Nat. Mus.). Carcharhinus lamiella, Jordan, D. S., Fishes of Sinaloa, 1895: 382 (Mazatlan, Mexico; deformed tail). Jordan, D. S., & Evermann, B. W., Fishes North and Middle America, 1, 1896: 37 (de- scription; range). Carcharinus commersonii (in part), Garman, S., The Plagiostomia, 1913: 140 (synonymy referring to lamiella only). Meek, S. E., & Hildebrand, S. F., Marine Fishes of Panama, 1, 1923: 43 (synonymy referring to lamiella only, not description). Eulamia commersonii (in part), Fowler, H. W., Zoologica: New York Zoological Society Beebe & Tee-Van: Sharks 111 1941] Proc. 4th Pac. Sci. Cong., Java, 1929 (1930): 493 (localities referring to lamiella only; refer- ences include platyrhynchus) . Discussion: The relationship of this species, and of platyrhynchus and galapagensis are not as clearly defined as might be desired. Thus we find that the original description of platyrhynchus states the following: “From lamiella it differs in the notched teeth and the anterior position of the first dorsal.” The type of lamiella, however, is not in accord with its original description and the first dorsal fin is much further forward than is stated. It is actually the same in position in both species. Relative fin positions are thus invalid as distinguishing specimens of lamiella and platyrhynchus. The teeth serrations in the two species con- stitute a real difference. Thus in lamiella, of which we have examined adult and young in addition to the young type, the serrations on the teeth are very fine, while in the more or less equal-sized specimen of platyrhynchus and the eight specimens of the closely related galapagensis the serrations are exceptionally coarse. As far as galapagensis and platyrhynchus are concerned, the original description of the former states: “The same in every respect as C archarias platyrhynchus (Gilbert) except that the fins at all ages are of uniform coloration with the body, being never margined with white.” In the materials that we have examined the presence or absence of white fin tips is clearly demonstrable, and the specimens are easily assigned to one species or the other on that basis, but the materials also indicate that further studies are needed involving series of young and old before these two forms are clearly defined. There seems to be considerable variation in our series of galapagensis, but owing to differences in size and lack of pertinent sized material we are unable to correlate these variations. It is of interest that, so far, all locality records of lamiella are continental, while those of platy- rhynchus and galapagensis, with the exception of a single questionable continental record for each species, are off-shore and insular. Carcharhinus lamiella possesses a low dermal keel between the first and second dorsal fins. This species has usually been synonymized under the Atlantic and Mediterranean Eulamia commersonii. Eulamia platyrhynchus Gilbert. White-tipped Shark. Range: Mexico: Magdalena Bay and Mazatlan (?), Revillagigedo Islands; Cocos Island, Clip- perton Island, Galapagos Islands and at sea, 230 miles N. W. of Clipperton Island. Field Characters: Small to medium-sized, broad-snouted sharks with the origin of the 2nd dorsal fin approximately above that of the anal ; teeth of the upper jaw triangular with the outer margin notched or concave, serrated; lower teeth erect, narrow, serrulate. Gray, with the tips of the dorsal, pectoral and caudal fins tipped with white. (Illustration from Specimen No. 17,521, Field Museum; 700 mm.) Color: Dorsal fins, sometimes only the first dorsal, tipped with white; occasionally with the dorsal and pectoral fins tipped and posteriorly bordered with white, the marginal parts of the fins pale. Size: Grows to nine feet. Abundance: A locally abundant fish. Beebe reports 16, five- to six-foot sharks seen while diving in one spot at Cocos Island. Food: Fish; a new species, Pontinus strigatus, was taken from the stomach of this species. There is also a record of Diodon hystrix. Study Materials: 2 specimens. Clarion Island, 1 male (25594) 1,524 mm., May 10, 1936, hand line, teeth preserved and photograph. 1, Tagus Cove, Galapagos Islands, 700 mm. (Field Mus. Nat. Hist. 17521), Jan. 9, 1929. References: Eidamia ( Platypodon ) platyrhyn- chus, Gilbert, C. H., Proc. U. S. Nat. Mus., 14, 1891: (1892): 543 (original description: part refers to galapagensis: Clarion Island, Magdalena Bay). Carcharias platyrhynchus, Snodgrass, R. E., & Heller, E., Proc. Wash. Acad. Sci., 6, 1905: 344 (Clarion Island, near Clipperton Island, color note), 414 (280 mm. Diodon hystrix from stom- ach). Beebe, W., The Arcturus Adventure, G. P. Putnam’s Sons 1925: 246, 435 (sociability), 302, 435 (habits) 412 (Cocos Island). Carcharinus platiyrliynchus, Herre, A. W., Field Mus. Nat. Hist., Zool. Ser., 21, 1936: 23 (brief note on proportions, color and abundance ; Cocos and Galapagos Islands) . Carcharhinus platyrhynchus, Jordan, D. S., & McGregor, R. C., Rep’t. U. S. Fish Comm., 1898 (1899): 274 (Clarion Island). Jordan, D. S. & Evermann, B. W., Fishes North and Middle America, 1, 1896: 36 (description in part, re- mainder belongs to galapagensis). Heller, E., & Snodgrass, R. E., Proc. Wash. Acad. Sci., 5, 1903: 209 (Galapagos Is., type of Pontinus strigatus taken from stomach) . Carcharinus commersonii (in part) Garman, S., The Plagiostomia, 1913: 140 (synonymy referring to platyrhynchus only). Meek, S. E., & Hildebrand, S. F., Marine Fishes Panama, 1, 1923: 43 (synonymy referring to platyrhynchus). Carcharias sp. incog., Jordan, D. S., & Gilbert, C. H., Proc. U. S. Nat. Mus., 5 1882 (1883): 107 (teeth of Mazatlan fish, may refer to this species). Discussion: See under C. lamiella. 112 Eulamia galapagensis (Snodgrass & Heller). Range: Mexico: Lower California (?), Clarion Island; Clipperton Island; Cocos and Galapagos Islands; and at sea 75 miles northeast of Malpelo Island. Field Characters: Small to medium sized, broad-snouted sharks with the 2nd dorsal fin origin approximately above that of the anal. Teeth of the upper jaw triangular, notched or concave on the outer margin and strongly serrated. Lower teeth, narrow, erect, with much smaller serrations. Color uniform gray, the tips of the fins dusky, never tipped with white. (Illustration from specimen No. 5257 : 798 mm.) Color: Iris silvery-gray. Abundance: Herre states: “This shark swarms in the waters of the Galapagos Islands and about Cocos Island. I have never seen sharks of this genus in such abundance as in these two localities. A great many of this species were caught by hook and line or harpooned. . .” Size and Weight: Grows to 8 feet. A 790 mm. (31 inches) fish weighed 7 pounds and a 1,260 mm. (50 inch) shark weighed 24 pounds. Study Material: 8 specimens. Clarion Island, 3 males and 1 female (25498, 25674, 25675, 25676; 590, 735, 768 and 842 mm.) May 10, 1936, hook and line and harpooned. Tower Island, Galapagos Islands, 2 (5254, 5257) 790, 798 mm., April 8, 1925, harpooned. GaMpagos Islands, 2 (Field Museum No. 17520) Jan. 9, 1929. Cocos Island, 1 (Field Museum No. 17519), Jan. 2, 1929. References: Carcharias galapagensis, Snodgrass, R. E., & Heller, E., Proc. Wash. Acad. Sci., 6, 1905: 343 (original description; short synonymy; range, size, food; type locality: Galapagos Islands; type No. 12324, University Museum, Stanford University) Beebe, W., The Arcturus Adventure, G. P. Putnam’s Sons, New York, 1926: 184, 302 (color of eye), 412 (4-foot specimen harpooned, food; Galapagos Islands). Eulamia galapagensis, Fowler, H. W., Proc. U. S. Nat. Mus., 80 (6) 1932: 1 (teeth; GaMpagos Islands). Fowler, H. W., Acad. Nat. Sci. Phila., Monograph No. 2, 1938: 13 (jaw and ventral fins of female from 75 miles northeast of Malpelo [XXVI: 15 Island), 19 (references; jaws, ventral fins and section of skin; GaMpagos Islands), 248 (check- list; range among GaMpagos Islands). Carcharinus galapagensis, Herre, A. W., Field Mus. Nat. Hist., Zool. Ser., 21, 1936: 22 (color, abundance, size; GaMpagos and Cocos Islands). Eulamia lamiella (?), Jordan, D. S., & Boll- man, C. H., Proc. U. S. Nat. Mus., 12, 1889: 179 (name only; GaMpagos Islands). Eulamia ( Platypodon ) platyrhynchus (in part), Gilbert, C. H., Proc. U. S. Nat. Mus., 14, 1891 (1892): 543 (non-white-margined specimens; Revillagigedo Islands) . Carcharhinus platyrhynchus, Jordan, D. S., & Evermann, B. W. (in part), Fishes North and Middle America, 1, 1896: 36 (part of description referring to galapagensis) . Discussion: See discussion under lamiella. Scoliodoti Muller & Henle, 1837. Scoliodon longurio (Jordan & Gilbert). Text-figure 22. Range: Gulf of California, and coast of Mexico, Panama. (Mexico: Santa Inez Bay, Guaymas, Mazatlan, San Lucas Bay, Banderas Bay, Tangola-Tangola; Panama: Panama Bay.) Field Characters: A small, sharp-snouted shark with origin of the second dorsal fin above or posterior to the middle of the base of the anal fin; conspicuous labial fold and groove at the angle of the mouth, paralleling the jaw and extending forward about one third the distance from the gape to the front of the mouth. (Illus- tration after Meek & Hildebrand, 1923.) Color: Gray with a bluish tinge above, white below; tips of fins margined with dusky. Size and Weight: Grows to 1,068 mm. (42 V2 inches). A specimen of this size weighed nine pounds. Local Distribution: All of the records and our specimens seem to indicate that this is a bay shark. Abundance: Locally abundant, based on Jordan’s and our experiences. Study Materials: 2 specimens. Mexico: Santa Inez Bay (24,993) 980 mm., April 13, 1936, hand-line; Puerto Vallarte, Banderas Bay, 1 (27,054), 815 mm., Nov. 15, 1937, hand line (teeth saved). We have also examined six speci- mens from San Lucas Bay, and an additional one from Santa Inez Bay. Zoologica: New York Zoological Society 1941] Beebe & Tee-Van: Sharks 113 References: Carcharias longurio, Jordan, D. S., & Gilbert, C. H., Proc. U. S. Nat. Mus., 5, 1882: 106 (original description, color; type locality: Mazatlan, Mexico; types, Nos. 28,306, 28,330, 28,331, 29,451, 29,551, U. S. National Museum). Carcharhinus longurio, Jordan, D. S., Proc. U. S. Nat. Mus., 8, 1885 (1886): 363 (check-list). Scoliodon longurio, Jordan, D. S., & Gilbert, C. H., Bull. U. S. Fish. Comm., 2, 1882: 105 (Mazatlan, Mexico). Evermann, B. W., & Jenkins, 0. P., Proc. TJ. S. Nat. Mus., 14, 1891 (1892): 130 (references; Guaymas, Mexico). Jordan, D. S., Fishes of Sinaloa, 1895: 382 (common at Mazatlan). Jordan, I). S., & Ever- mann, B. W., Fishes North and Middle America. 1, 1896: 42 (description). Jordan, D. S., & Evermann, B. W., Fishes North and Middle America, 3, 1898: 2748 (note on teeth and size of first dorsal). Gilbert, C. H., & Starks, E. C., Fishes of Panama Bay, 1904: 12 (6 specimens from Panama; note on porportions and teeth), 207 (range; Panama and Gulf of California). Garman, S., The Plagiostomia, 1913: 114 (short synonymy, description, color; range). Meek, S. E., & Hildebrand, S. F., Marine Fishes of Pana- ma, 1, 1923: 52, Plate 2, fig. 1 (short synonymy, description, color, figure, comparison with types, claspers; Panama fish market, 525 and 700 mm. males). Breder, C. M., Jr., Bull. Bingham Oceanogr. Coll., 2 (1), 1928: 3 (specimen from unknown locality). Seale, A., Allan Hancock Pacific Expeditions, 9 (1), 1940: 1 (490 mm. specimen, Tangola-Tangola, Mexico). Galeocerdo Midler & Henle, 1838. Galeocerdo arcticus (Faber). Tiger Shark. Range: Tropical and temperate seas, north rarely to 70°. In the eastern coastal Pacific north to San Diego. (Mexico: Gulf of California, Santa Inez Bay, Concepcion Bay, Guaymas, Mazatlan, Tangola-Tangola; Guatemala: San Jose de Guatemala ; Costa Rica : Golfito ; Panama : Panama Bay and Pearl Islands; Clarion Island; Clipperton Island; Cocos Island; Galapagos Islands: Narborough, Albemarle and Guy Fawkes Island.) Field Characters: A large, heavy shark with blunt head ; caudal fin large with very long upper lobe, and well-developed lateral keels at base; teeth alike in both jaws, semicircular, with a deep notch and coarsely serrated edges, the tips turned obliquely outward. (Illustration after Norman, 1937.) Color: Dark gray above, white below; numer- ous black, rectangular spots on body and fins, usually forming vertical bars, becoming roundeci on upper caudal lobe. This pattern is lost on older individuals. Iris greenish-brown. Size: Reaches a length of at least 20 feet (Record of 30 feet unconfirmed). Weight: A shark of 1,625 mm. (5 feet, 4 inches) weighed 137 pounds; 1, 3,073 mm. (10 feet, 1 inch) 366 pounds (liver 97 lbs. 26% of whole) ; 1, 3,200 mm. (10 feet, 6 inches) 505 pounds; 1, 3,886 mm. (12 feet, 9 inches) 780 pounds (liver 188 lbs., 24% of whole). Local Distribution: Well offshore and in bays of only four fathoms depth. Abundance: Tiger sharks are fairly common throughout the area under consideration. Food: Almost any invertebrate or vertebrate of sufficient size may find a place in the diet of this shark. Our list is as follows: garbage (3 stomachs), octopus (400 mm.), Heterodontus quoyi (375 mm.), sting rays (7 in 3 stomachs, four of them Urobatis halleri), Gymnosarda alletterata (400 mm.), Mycteroperca jordani (600 mm.), Diodon holacanthus (200 mm.), Ogcocepha- lus sp. (150 mm.), Iguana iguana (1,371 mm.), Chelone mydas, full of eggs (760 mm.), feathers (3 stomachs), 2 Clarion shearwaters, Puffmus auricularis; and Galapagos sea-lion pup, Otaria jubata. Parasites: Two copepods taken from near the gills, Pandarus satyra Dana and Rocinela aries. Study Material: Definite notes were made on 11 tiger sharks, and several other individuals were seen. All were hooked from the deck of vessels. Mexico: Santa Inez Bay, 1 (24,894), 1,625 mm., April 9, 1936; Santa Inez Bay, 1, 1,422 mm., April 9, 1936; Concepcion Bay, 1, 1,625 mm., April 16, 1936; Clarion Island, 1, ca. 1,400 mm., May 11, 1936; Clarion Island, 1 (25,655), 3,886 mm., May 13, 1936; Tangola- Tangola Bay, 1 (26,051), 3,073 mm., Dec. 10, 1937; Costa Rica: Golfito, 1 (26,184), 3,200 mm., March 8, 1938; Panama: Pearl Islands, 1, 1,882 mm., June 27, 1933; Galapagos Islands: Tagus Cove, Albermarle Island, 1 (6159), 2,133 mm., June 7, 1925; Guy Fawkes Islands, 1, “18 or 20 feet,” March 31, 1923, “seen to kill and devour a sea-lion pup”; Cocos Island: 1, “ 15 to 18 feet,” May 17, 1925. References: Squalus arcticus, Faber, F., Fische Islands, 1829: 17 (Iceland and neighboring seas). Galeocerdo tigrinus, Jordan, D. S., & Gilbert, C. H., Bull. U. S. Fish Comm., 2, 1882 (1883): 112 (San Jose de Guatemala). Gilbert, C. H., Bull. U. S. Fish Comm., 2, 1882 (1883): 105 (Mazatlan, Mexico). Pellegrin, J., Bull. Mus. Hist. Nat., Paris, 7, 1901: 161, 166 (Gulf of California, danger to man). Snodgrass, R. E., & Heller, E., Proc. Wash. Acad., 6, 1905: 342 (Albemarle and Narborough Islands, Galapagos 114 Zoologica: New York Zoological Society [XXVI: 15 Islands). Fowler, H. W., Proc. Acad. Nat. Sci. Phila., 60, 1908: 61 (jaws from Guaymas, Mexico). Beebe, W., “Galapagos: World’s End,” New York, 1924: 201, 434 (Guy Fawkes Islands, Galapagos ; eating sea-lion pup). Galeocerdo maculatus, Jordan, D. S., & Boll- man, C. H., Proc. U. S. Nat. Mus., 12, 1890: 179 (Panama) . Galeocerdo arcticus, Meek, S. E., & Hildebrand, S. F., Marine Fishes of Panama, 1, 1923: 56 (once taken on the Pacific Coast by the Alba- tross). Beebe, W., “The Arcturus Adventure,” New York, 1926: 247, 435 (Cocos Island). Breder, C. M., Jr., Bull. Bingham Oceano. Coll., 2 (1), 1928: 3 (specimen from unknown locality). Schmitt, W. L., Annotated List of Fishes, Presidential Cruise, 1938, privately printed, 1938: v (Weights; Cocos, Clipperton, and Galapagos Islands). Galeorhinus Blainville, 1816. We have no material referable to this genus. It is evident from the confused literature that careful study should be made of the relation- ships of the northern Galeorhinus zyopterus and of the specimens reported from Peru and Chile as zyopterus, 19 galeus ,20 molinae,n and chilensis ,n Fowler23 places all of these records under the name galeus. Apparently there are no records of the genus from the tropical eastern Pacific, beyond those from Cedros Island and Peru, the northern and southern boundaries respectively, of our region. Family Sphyrnidae. Sphyrna Rafinesque, 1810. Hammerhead and Shovelhead Sharks. Key to tropical eastern Pacific species.24 la. Second dorsal fin with a long posterior lobe, which when lifted upward, will reach about twice as high as the fin; anterior margin of the head three-lobed zygaena lb. Second dorsal fin with a short posterior lobe, which when lifted upward, will reach about as high as the fin. 2a. Front margin of the head between the nasal apertures lobed, the front margin not forming a continuous curve. 19 Evermann & Radcliffe, U. S. Nat. Mus., Bull. 95: 1917: 10. 20 Fowler, Proc. blh Pac. Sci. Congr. Java 1929, 3, 1930: 490. si Philippi, Ann. Univers. Chile, 71, 1887: 543, Plate 4, flg. 2. 22 Perez, Estudios sobre algunos escualos de la costa de Chile, 1886: 3. 23 Fowier, 1. c., 490. 23 Adapted with slight modifications from Springer (1940). Sphyrna peruana Phillipi from Chile and Peru has not been considered here. The amount of material in our collection of zygaena and tudes is so small that we have made no attempt to determine whether the eastern Pacific species should be considered as local races, as seems to be indicated in many of the littoral tropical eastern Pacific sharks. 3a. Head hammer-shaped; oculo-narial expanse irregularly quadrangular, almost exactly transverse in old adults; teeth heavy, serrate tudes 3b. Head not definitely hammer-shaped; oculo- narial space irregularly oval; teeth slender, not serrate corona 2b. Front margin of the head between the nasal apertures not lobed, the front margin form- ing a continuous curve. 4a. Teeth with low cusps, the cusps pro- gressively smaller towards the angles of the jaws, entirely absent on one or two rows in the upper jaw and on four or five rows in the lower jaw; head broadly spade-shaped; length of snout to mouth 1.5 to 1.75 in the internasal distance vespertina 4b. All teeth with cusps; oculo-narial space broadly oval; length of snout to mouth 2.2 in internasal distance media Sphyrna zygaena (Linnaeus). Cruz, Pez Martillo. Range: Tropical and temperate seas; known in the eastern tropical Pacific from southern Cali- fornia, Mexico, Panama, Peru and the Galapagos Islands. (Mexico: San Lucas Bay, Mazatlan, Guaymas; Panama: Panama; Peru: Lobas de Tierra, Callao; Galapagos Islands). Field Characters: A large shark with head expanded laterally, hammer-shaped; anterior edge of head between nostrils three-lobed; a line connecting the centers of the eyes passes through the mouth; diameter of eye greater than anterior extension of head immediately in front of eye; posterior lobe of second dorsal fin, when lifted upward, reaches twice as high as the fin. (Illus- tration from specimen 25,549; 1,030 mm.) Size: Grows to 17 to 20 feet and a weight of 1,500 pounds. Study Material: 1 specimen. Mexico: San Lucas Bay, Lower California, 1 (25,549), 1,030 mm., May 5, 1936, harpooned. References: Squalus zygaena, Linnaeus, Syst. Nat., ed. X, 1758: 234 (original description; Europe, America). Sphyrna zygaena, Jordan, D. S. & Gilbert, C. H., Bull. U. S. Fish Comm., 2, 1882: 105 (Mazat- lan). Jordan, D. S., & Gilbert, C. H., Bull. U. S. Fish Comm., 2, 1882: 109 (Panama). Evermann, B. W., & Jenkins, O. P., Proc. U. S. Nat. Mus., 14, 1891 (1892): 131 (2}4 foot specimen from Guaymas, Mexico). Jordan, D. S., Fishes of 1941] Beebe & Tee-Van: Sharks 115 Sinaloa, 1895: 383 (Mazatlan). Jordan, D. S., & Evermann, B. W., Fishes of North and Middle America, 3, 1898: 2748 (Mazatlan). Pellegrin, J., Bull. Mus. Hist. Nat., Paris, 7, 1901: 161 (Gulf of California). Gilbert, C. H., & Starks, E. C., Fishes of Panama Bay, 1904: 13 (abun- dance in Panama Bay), 207 (range). Starks, E. C., Proc. U. S. Nat. Mus., 30, 1906: 762 (eastern Pacific range), 763 (Callao, Peru). Garman, S., Bull. Mus. Comp. Zool., 46 (12), 1906: 229 (Panama). Hussakof, L., Copeia, 34, 1916: 63-64 (comparison of Atlantic and Pacific hammerheads, development of teeth). Ever- mann, B. W., & Radcliffe, L., Bull. 17. S. Nat. Mus., 95, 1917: 5 (Lobos de Tierra, Peru; meas- urements of a 1,000 mm. shark, short synonymy, questioning relationship of S. peruana Philippi). Nichols, J. T., & Murphy, R. C., Bull. Amer. Mus. Nat. Hist., 46, 1922: 504 (Lobos de Tierra, Peru). Walforcl, L. A., Fish and Game of Cali- fornia, Bull. 45, 1935: 40, fig. 38 (figure, notes). Walford, L. A., California Fish and Game, 17, 1931: 404 (off southern California coast, landed at San Pedro and Santa Monica). Clark, H. W., Proc. Cal. Acad. Sci., (4) 21 (29) 1936: 395 (Galapagos Islands, pile of dried skins). Wal- ford, L. A., Marine Game Fish of the Pacific Coast, 1937: plate 25 (figure). Seale, A., Allan Hancock Pacific Expeditions, 9 (1), 1940: 2 (Galapagos Islands; abundance). Cestracion zygaena, Garman, S., The Plagio- stomia, 1913: 157, plate 1, fig. 1-3 (Synonymy, des. of Atlantic specimen, figure). Discussion: Our specimen from San Lucas Bay agrees well with illustrations of specimens re- corded as zygaena from the Atlantic and other localities. Sphyrna tildes Valenciennes. Range: Tropical seas. (All of the definite records within the coastal tropical eastern Pacific area, are concentrated in the Gulf of California: Concepcion Bay, Arena Bank, San Francisquito Bay, Guaymas.) Field Characters: A large shark with head expanded laterally, hammer-shaped; anterior border of head between the nostrils four-lobed; a line joining the centers of the eyes passes in front of the mouth; diameter of eye equal to anterior extension of head immediately in front of eye; posterior lobe of second dorsal fin, when lifted upward, reaching about as high as the fin. (Illustration from specimen No. 25247; 1,334 mm.). Size and Weight: Grows to about 5 feet. Study Material: 2 specimens. Mexico: Con- cepcion Bay, Lower California, 1 (25,247), 1,334 mm., April 16, 1936 (head and pectoral fins preserved). Arena Bank, Lower California, 1 (25,485), May 1, 1936, dried head picked up on beach, 767 mm. across the “hammer.” References: Sphyrna tudes, Valenciennes, A., Mem. Mus. Hist. Nat. Paris, 9, 1822: 225, PI. 12, fig. 1 (original description, figure). Evermann, B. W., & Jenkins, O. P., Proc. U. S. Nat. Mus., 14, 1891 (1892): 131 (Guaymas; synonymy wrong). Pellegrin, J., Bull. Mus. Hist. Nat. Paris, 7, 1901 : 161 (name only, Gulf of Cali- fornia), 166 (abundance). Breder, C. M., Jr., Bull. Bingham Oceanogr. Coll., 2 (1) 1928: 4 (San Francisquito Bay, Gulf of California). Discussion: The two fish at hand constitute too small a sample for adequate comparison with specimens of tudes from other localities. The large head agrees in shape and form with the head of the smaller example and there is no doubt that the two heads represent the same species. They both agree well with descriptions of tudes from other localities. The teeth of the large fish are lost; those of the 1,334 mm. shark are non-serrated. This species, although supposedly wide spread in tropical seas, seems to be known in the eastern tropical Pacific only from the Gulf of California. Springer (1940) has shown that some of the older records within our faunal area that were assigned to tudes belong to new species described by him. The Galapagos Islands record of tudes2S of Snodgrass & Heller, is inconclusive and may refer to some of Springer’s species. Sphyrna corona Springer. Range: Panama (Panama City) and “west coast of Mexico.” Field Characters: Medium sized sharks with head expanded laterally, kidney-shaped; front of head lobed, the posterior border of the head not parallel to the anterior border. (Illustration after Springer, 1940). Size: Grows to about three feet. Study Material: None. References: Sphyrna corona, Springer, S., Stan- ford Ichthy. Bull., 1 (5) 1940: 163, fig. 4 (original 26 Snodgrass, R. E., & Heller, E., Proc. Wash. Acad. Sci., 6, 1905: 345. 116 Zoologica: New York Zoological Society [XXVI: 15 description, figure; type-locality Panama; type, No. 11,882, Stanford University). Sphyrna hides, Jordan, D. S., & Evermann, B. W., Fishes of North and Middle America, 1, 1896: 44 (in part). Sphyrna tibnro, Kumada, T., & Hiyama, Y., Marine Fishes of the Pacific Coast of Mexico, 1937: 17, plate 1 (figure and short comment). Discussion: Judging by head shape, the approx- imately 915 mm. shark figured by Kumada & Hiyama ( l . c .) under the name of Sphyrna tiburo, is the same as corona. Sphyrna vespertina Springer. Range: Panama and Ecuador. Field Characters: A small shark with expanded spade-shaped head, the front margin of the head between the nostrils not lobed, the head slightly pointed anteriorly; teeth with low cusps, the cusps becoming progressively smaller toward the angles of the jaws, entirely absent on one or two rows in the upper jaw and on 4 to 5 rows in the lower jaw. (Illustration after Springer, 1940.) Size: Grows to about three feet. Study Material: None. References: Sphyrna vespertina, Springer, S., Stanford Ichthy. Bull., 1 (5), 1940: 161, fig. 2 (original description, figure; type-locality, Pana- ma; type, No. 11,584, Stanford University). Sphyrna tiburo, Gilbert, C. H., & Starks, E. C., Fishes of Panama Bay, 1904: 13 (Discussion of so-called tiburo and tudes). Wilson, C., Ann. Carn. Mus. Pitts., 10, 1916: 58 (Guayaquil, Ecuador) . Discussion: Probably most of the eastern Pacific records of tiburo will be found to refer to this species. Springer in his 1940 paper infers that tiburo is Atlantic only, as he states of vespertina “ . . . It is closely allied to Sphyrna tiburo of the Caribbean, Gulf of Mexico and the Atlantic coast of the Americas. . .” The reference to tiburo of Kumada in “Marine Fishes of the Pacific Coast of Mexico,” 1937: Plate 1, has been placed under Sphyrna corona. Specimens mentioned in the following references require study: Meek, S. E., & Hilldebrand, S. F., Marine Fishes of Panama, 1, 1923: 60-61 (Panama); Starks, E. C., Proc. U. S. Nat. Mus., 30, 1916: 762 (range), 763 (specimen from Guaya- quil, Ecuador), and Jordan, D. S., Fishes of Sinaloa, 1895:383 (one specimen from Mazatlan). Sphyrna media Springer. Range: Mexico and Panama. (Mexico: Mazat- lan; Panama: Panama market.) Field Characters: Medium-sized sharks with head expanded laterally, the anterior margin between the nostrils in a continuous curve, not lobed; all teeth with cusps. Study Material: None. References: Sphyrna media, Springer, S., Stan- ford Ichthyological Bull., 1 (5), 1940: 162, fig. 3 (original description, figure; type-locality, Ma- zatlan, Mexico; type, No. 11,583, Stanford University). Sphyrna tudes, Jordan, D. S., Fishes of Sinaloa, 1895: 383 (Mazatlan, Mexico). Jordan, D. S., & Evermann, B. W., Fishes North and Middle America, 1, 1896: 44 (in part). Gilbert, C. H., & Starks, E. C., Fishes of Panama Bay, 1904: 13, (Panama market) . 1941] Beebe & Tee-Van: Sharks 117 Family Heterodontidae. Heterodontus Blainville, 1816. Key to tropical eastern Pacific species. la. Origin of first dorsal fin over the base of the pectoral fins; distance from snout to origin of first dorsal 28 to 30% of the length; black spots on body separate and widely spaced, rarely merged together francisci lb. Origin of first dorsal fin behind the base of the pectoral fins; distance from snout to origin of first dorsal fin 32 to 37% of the length; black spots on body larger, three or four joining together to form larger spots or bands, .quoyi Heterodontus francisci (Girard). Text-figure 29. Range: Southern California, Lower California, Gulf of California. Recorded south of San Diego, California, from the following localities, Mexico: Todos Santos Bay, Magdalena Bay, Angel de la Guardia Island, Concepcion Bay and San Francisquito Bay. Field Characters: Small robust sharks with eyes beneath heavy suborbital ridges and with a spine in front of each dorsal fin; spiracle present below posterior part of eye; origin of first dorsal fin over the base of the pectoral fin; upper surfaces of the body with black subcircular spots which are separate and detached from each other, rarely close together or merging. (Illus- tration after Kumada and Hiyama, 1937: 532 mm.) Size: Grows to three feet. Study Materials: No specimens were obtained by the Zoological Society’s expeditions. We have examined fishes from Concepcion Bay, San Francisquito Bay and Angel de la Guardia Island, Gulf of California. References: Cestracion francisci, Girard, C. F., Proc. Acad. Nat. Sci. Phila., 7, 1854: 196 (original description; type locality: Monterey Bay, Cali- fornia). Heterodontus francisci, Smith, R., Proc. U S Nat. Mus., 6, 1883 (1884): 233 (Egg-cases on beach, Todos Santos Bay, Mexico). Pellegrin, J., Bull. Mus. Hist. Nat. Paris, 7, 1901: 161. (Gulf of California.) Gyropleurodus francisci, Osburn, R. C., & Nichols, J. T., Bull. Amer. Mus. Nat. Hist., 35, 1916: 141 (8" specimen dredged in 13 fathoms, Magdalena Bay, Mexico). Breder, C. M., Jr., Bidl. Bingh. Oceanogr. Coll., 2 (1), 1928: 3 (Concepcion Bay and San Francisquito Bay, Gulf of California). Kumada, T., & Hiyama, Y., Marine Fishes West Coast of Mexico, 1937: 15, Plates 44, 45 (figure). Heterodontus quoyi (Freminville). Range: Peru (Lobos de Tierra and Lobos de Afuera Islands); Galapagos Islands. An in- definite record from the west coast of Mexico, and a possible egg-case of this species from San Elena Bay, Ecuador. Field Characters: Small, robust sharks with eyes beneath heavy supraorbital ridges and with a short spine in front of each dorsal fin ; spiracle present, beneath posterior part of eye; origin of first dorsal fin behind the vertical of the base of the pectoral. Brownish to grayish, the upper and lower surfaces of body with black spots which merge to form larger spots and bands. (Illustration from specimen No. 6161: 375 mm.) Size: Grows to at least eighteen inches. Description: Our Galapagos Island specimen is described as follows, the proportions being ex- pressed in percentages of the total length, which is 375 mm.: Body robust, tapering posteriorly; snout to 1st gill-slit 18%, snout to fifth gill-slit 23.7; snout to eye 9.2, horizontal diameter of eye 3.3; a strong supra-orbital ridge ending less than an eye’s diameter back of the eye, the ridges anterior to the eye converge at first and then become parallel and less evident on the snout. Snout to first dorsal fin 34, the fin originating about 2.6 back of the vertical of the posterior part of the pectoral base; base of first dorsal fin 8.25, anterior edge 13.6, base to tip of posterior lobe 5.1, vertical height of fin from body 10.4, the fin rounded above, its free edge slightly concave and with no posterior extension. Origin of first dorsal to tip of spine 6.7, vertical height of spine from body 5.85. Interdorsal space 17. Snout to second dorsal fin 54, the fin similar in shape but slightly smaller than the first dorsal fin. Caudal fin length 22, the fin with a large terminal lobe and a large inferior lobe, the length of the latter being 13.3. Snout to pectoral fin 19, the fin low, large and horizontal; pectoral base 8.8, anterior border 21.4, inner border 7.2, free border 16, greatest breadth 14.4. Snout to pelvic fin 42.7, the fin originating slightly posterior to the verti- cal of the posterior base of the first dorsal fin, pelvic fin base 7.85, anterior border 10.7, inner border 8.5; the claspers do not reach the inner tip of the fin. Snout to anal fin 68, the fin originating under the posterior tip of the second dorsal fin, the tip of the anal reaching to the base of the caudal, base of anal 5.5, anterior border 9.9, inner border 3.3. Height of first gill-slit 4, height of 5th gill-slit 2.27, the posterior three slits above the pectoral fin. Spiracle small, beneath and posterior to the eye. Teeth similar to those mentioned in Garman’s account (1913) and figure, except that the an- 118 Zoologica: New York Zoological Society [XXVI: 15 terior teeth have five rather dull cusps. The teeth in the Peruvian example that we have studied are much sharper and tricuspid, so that in this species as well as in francisci, there is an evident reduction with age in the number of cusps on the anterior teeth. Denticles from the side immediately beneath the first dorsal fin are X-shaped with one of the arms of the X stronger and heavier toward its outer extremities. Denticles from the posterior part of the body have the spaces between the arms of the X filled in and the denticles are roughly rhomboidal in shape. Color: Immediately after being taken from the stomach of a shark, above neutral gray, below and on sides dotted with large round black spots. Size and Weight: Our 375 mm. fish weighed 456 grams (1 pound). Enemies: The Ardurus specimen was taken from the stomach of a seven-foot tiger-shark C Galeocerdo ardicus ). Egg Case: See under Discussion. Study Materials: 2 specimens. Galapagos Islands: Albemarle Is., 1, male (6161), 375 mm., June 9, 1925 ( Ardurus Expedition), stomach of tiger-shark. Peru: Lobos de Afuera Island, 1 female (Amer. Mus. Nat. Hist. 7489) washed ashore dying. References: Cestracion quoyi, Freminville, C. P., Mag. de Zool., (2) 2, 1840: pi. 3 (description, figure; type-locality: Gal&pagos Islands). Gun- ther, A., Cat. Fishes Brit. Mus., 8, 1870: 416 (short description, synonymy). Cestracion pantherinus, Valenciennes, A., Voy- age “Venus,” Zoology, 1845: 350, pi. 10, fig. 2 (description, excellent figure, Galapagos Islands). Tropidopus pantherinus, Gill, T. N., Proc. Acad. Nat. Sci. Phila., 1862: 489, 490 (new genus established for this species). Heterodontus ( Cestracion ) quoyi, Dumeril, A. H. A., Hist. Nat. Poiss., 1, 1865: 427; atlas, pi. 3, fig. 16, 17 (description, color, figure of teeth). Heterodontus quoyi, Miklukho-Maklai, N. N., & Macleay, W., Proc. Linn. Soc. N. S. Wales, 3, 1878: 316, pi. 26, fig. 35 (comments, no specimens, copy of original plate of Freminville). Fowler, H. W., Proc. 4th Pac. Sci. Congress, Java 1929, 3, 1930: 484 (check-list only, Galapagos Islands). Gyropleurodus quoyi, Jordan, D. S., & Ever- mann, B. W., Fishes North and Middle America, I, 1896: 21 (poor description, synonymy), Regan, C. T., Ann. Mag. Nat. Hist. (8) 1 : 494. Cestracion quoyi, Garman, S., The Plagio- stomia, 1913: 187, pi. 47, fig. 1-3 (description, synonymy, figure of teeth). Gyropleurodus peruanus, Evermann, B. W., & Radcliffe, L., Bull. U. S. Nat. Mus., 95, 1917: 2, pi. 1, fig. 1 (description, figure; type-locality : Lobos de Tierra Island, Peru). Nichols, J. T., & Murphy, R. C., Bull. Amer. Mus. Nat. Hist., 46, 1922: 504 (Lobos de Afuera, Peru; specimen 1941] Beebe & Tee-Van: Sharks 119 washed ashore in dying condition). ? Kumada, T., & Hiyama, Y., Marine Fishes of the Pacific Coast of Mexico, 1937: 16, pi. 46 (west coast of Mexico; figure), see comment under Discussion. Heterodontus peruanus, Fowler, H. W., Proc. ^th Pac. Sci. Congr., Java 1929, 1930: 484 (check-list; Peru). Discussion: Two species of Heterodontus have been reported from the equatorial eastern Pacific, quoyi from the Galapagos Islands and peruanus from the coast of Peru. The original description of quoyi (Freminville, 1840) is very poor and the figure that illustrates the description is crude. The plate shows a fish with an exceedingly small pectoral fin, an anal fin with its origin beneath the tip of the posterior extension of the second dorsal fin and with very large, irregularly-formed dark spots. Valenciennes, in 1855, redescribed the Gala- pagos fish under the name Cestracion panther- inum, giving an excellent figure with details of anatomy and color that agree well with our recent Galapagos material. Garman (1913) described an 18-inch female of quoyi from the Galapagos Islands and gave splendid figures of the teeth. Unfortunately, we have been unable to find the specimen or specimens upon which his description was based. Evermann & Radcliffe first described and figured the South American coastal form, peruanus, the type being a 565 mm. fish from Lobos de Tierra, Peru. For purposes of comparing quoyi and peruanus, we have, in addition to the descriptions men- tioned above, two topotypical specimens, a 375 mm. fish from Tagus Cove, Albemarle Island, Galapagos Islands, and a 523 mm. fish (Amer. Mus. Nat. Hist. 7489) previously recorded by Nichols & Murphy from Lobos de Afuera, Peru, some 20 miles from the type locality of peruanus. The latter specimen agrees completely with the original description and figure of peruanus. As these fish are of different sizes, there is a slight question as to their identity. Consequently it is necessary to take up the discussion in some detail, although there is no doubt in our minds that they belong to the same species. When Evermann & Radcliffe described peru- anus, they stated: “This species appears to be most closely related to the poorly described quoyi, but differs in coloration, in insertion of the anal and relative size of the pectoral.” However, when the two specimens at hand, representing topotypical examples of both quoyi and peruanus, are compared, we find that the relative sizes of the pectoral fins are exactly the same (the proportional measurements are identi- cal), and that the insertion of the anal fin is in the same relative position in the two fishes. This leaves only the color as supposedly different. As our Galapagos Island shark is smaller than the two known Peruvian fishes and as it was taken from the stomach of a tiger-shark, its color cannot be depended upon for direct com- parisons. The general pattern, where observable, consists of large diffuse dark spots similar to those of the Peruvian shark, and in this, it agrees with Valenciennes’ figure of a Galapagos fish illustrated in the “Voyage of the Venus.” In the original description of the Peruvian form, conspicuous black bands on the dorsal surface are recorded. In Garman’s description of a similar sized GaMpagos specimen the same pattern is mentioned. However, in the American Museum Peruvian fish, in our Galapagos Island specimen and in Valenciennes’ plate there is no indication of these bands. Evidently their presence or absence is variable. Under any circumstances, it is obvious that the difference in pattern cannot be assigned to a geographical range. Thus we find that the coloration and pattern in the two nominal forms are the same, and as there are thus no distinguishing characters left between the coastal and the South American mainland form, we consider them as identical. All of the known Galapagos and Peruvian specimens of Heterodontus have relatively low and robust dorsal spines. The northern species, francisci, has much longer and slimmer spines. This distinction is most marked in smaller specimens, but it also holds true when compari- sons are made with equal-sized larger specimens. The specimen illustrated by Kumada & Hiyama (Marine Fishes of the Pacific Coast of Mexico, 1937: 16, Plate 46) introduces an un- certain note into the discussion of the American species of Heterodontus. This fish was found far north of the range of quoyi as previously re- corded and within or near the range of francisci. Kumada & Hiyama ’s plate, which was based upon a photograph, shows a fish with the dorsal fin insertion like that of francisci, while the shape and size of the dorsal fins is that of quoyi. The color pattern is intermediate between the two forms, the spots being larger than in francisci and more isolated and separated than is the case in quoyi. The general appearance, however, is predominantly quoyi- like. For the present, we assume that this fish was correctly assigned to peruanus, and that the differences are merely variations from our present conception of the species. The exact locality from which the Kumada & Hiyama shark was taken is unknown — somewhere along the west coast of Mexico. An egg-case of a heterodontid shark dredged by us at Station 136:D-16, Arena Bank, Gulf of California, April 20, 1936, in 45 fathoms, amply validates the existence of a second species of bull- head shark in the northern part of the tropical eastern Pacific in addition to francisci. This egg-case probably belongs to the present species as exemplified by Kumada & Hiyama’s specimen. The egg-case of Heterodontus francisci is well known, and has been figured by Daniels26 and Barnhart ;27 it is screw-shaped with characteristic 26 Daniels, J. F., The Elasmobrancli Fishes, 1922: 318. 22 Barnhart, P. S., Bull. Scripps Inst. Oceanogr. Tech, ser.. 3 (4) 1932: 88. 120 Zoologica: New York Zoological Society [XXVI: 15 flanges and lacks tentacles. Our specimen is quite different and resembles in general form and the possession of long tentacles, the egg-case of Heterodontus galeatus as illustrated by Waite.* * 3 When first secured our egg-case was broken at the top (considering the smaller tentacled end as the bottom, as shown in Waite’s figure'28 and since then it has become even more fractured and fragmentary. However, even from this material it is evident that this egg-case can be dis- tinguished from that of galeatus by the fact that the spiral flanges have their outer border vertical ; i. e., more or less parallel to the long axis of the egg-case and that the upper aspect of the flange is a flat ledge at a right angle with the body of the egg-case. In Waite’s illustration of the galeatus egg-case, these flanges are oblique to the vertical axis, more or less continuing the general direction of the body of the egg-case. The width across the widest portion of the egg-case is approximately 44 mm., the width of one of the flanges at about the middle of the egg-case is 7.5 mm. Tortonese29 reports a somewhat similar egg from Ecuador under the name of Heterodontus sp., although we do not understand his reference to four appendages. This must certainly belong to quoyi; it was 64 mm. long. 28 Waite, E., Journ. Linn. Soc. Zool. Land., 25, 1896: Plate 12, and ‘‘The Fishes of Australia,” Part 1, Sharks, etc., 1940: 38. 28 Tortonese, E., Boll. Musei Zool. Anat. Comp. Torino, (3) 47, no. 89, 1939: 2. Family Squalidae. Cent r os cy Ilium Midler & Henle, 1838. Centroscyllium nigrum Garman. Text-figure 32. Range: Offshore Panamic region in 546 to 625 fathoms; taken 15 miles west of Mariato Point, Panama, 65 miles south-west of Mariato Point, Panama, and 60 miles south of Cocos Island, Offshore, Kauai Island, Hawaiian Islands in 385 to 500 fathoms. Field Characters: Very small, large-eyed sharks with well-developed spiracle and with a spine before each dorsal fin; anal fin lacking. Black with white outer borders to the dorsal, pectoral and ventral fins. (Illustration from specimen No. 6013; 278 mm.) Color: Black with white dorsal spines and posterior borders to the dorsals, pectorals and pelvic fins; eye emerald green. Size: Grows to about a foot in length. Food: Our specimens had fed upon scarlet prawns. Study Material: 8 specimens. Station 74 :OT-4, Arcturus Oceanographic Expedition, 4° 50' N., 87° 00' W., 60 miles south of Cocos Island, 625 fathoms (1,145 meters), 125 to 278 mm. long, May 30, 1925, otter trawl. Cat. Nos. 6013, 6014a, b, c, d, e, f, g. References: Centroscyllium nigrum, Garman, S., Mem. Mus. Corny. Zool., 24, 1899: 28, Plate 1, fig. 2, Plates 4 and 5, Plate 69, fig. 1 (original description, skull and skeletal characters, heart, skin, sensory system, color; figure, figures of head, skull, teeth, denticles, branchial cartilages, intestine, shoulder and pelvic girdles and lateral line systems; type locality: Pacific Ocean, be- tween 6° and 7° N. Lat., and 81° and 82° W. Long., in depths of 546 to 555 fathoms). Gar- man, S., The Plagiostomia, 1913:231 (synonymy, description). Beebe, W., “The Arcturus Adven- ture,” G. P. Putnam’s Sons, New York, 1926: 359 and 436 (color and food). Centroscyllium ruscosum, Gilbert, C. H., Bull. U. S. Fish Comm., 23, 1905: 580, fig. 230 (de- scription, figure, color). Discussion: Garman30 states of this species: “One of the more prominent differential charac- ters of the species described below, C. nigrum, is apparent in the five cusped teeth, the teeth of each of the other species being described as tricuspid. . . .” However, the specimens at 30 Garman, Mem. Mus. Comp. Zool., 24, 1899: 27. 1941] Beebe & Tee-Van: Sharks 121 hand from close to the type-locality and from the same depths as the type of C. nigrum, demon- strate considerable variation from this descrip- tion. Thus the lower jaw teeth possess, for the most part, but three cusps with indications of a single additional cusp on each side, making five in all as in the original description of the species. However, these additional cusps are very small and never as prominent as shown in the original figures of the species. Without careful observa- tion they might easily be missed. They can most easily be seen when the teeth are dried and viewed with transmitted light. In the description of the cotype of the nominal ruscosum, Gilbert found the same condition. Squalus Linnaeus, 1758. Gunther’s questionable Panama record of Acanthias vulgaris 31 cannot be referred to any definite known species, and it also may have come from the Atlantic. Comments on this record have been made by Gilbert & Starks,32 and Meek & Hildebrand.33 Beyond this note the genus is unrecorded from the eastern tropical Pacific, although Squalus suckleyi is known from immediately north of the region and is also reported from Chile, while Squalus lebruni and S. fernandinus are known from south of the region. Family Scymnorhinidae. Isistius Gill, 1864. Isistius brasiliensis (Quoy & Gaimard). Text-figure 33. Range: Known from many localities in tropical and temperate seas; in the eastern tropical Pacific known from Hawaii and in our region from a single specimen taken in open ocean 150 miles north of the Galiipagos Islands. Taken principally near the surface, doubtfully recorded from great depths. Field Characters: Small, large-eyed sharks with very small dorsal fins placed far back on the body, no anal fin, a dermal keel on each side of the tail, a short deep caudal fin. Brown, with a darker band around the neck across the gill- openings; lower surfaces lighter to white. (Illus- tration after Garman, 1899; 460 mm.) 31 Gunther, A., Trans. Zool. Soc. London, 7, 1868: 396. 32 Gilbert, C. If.. & Starks, E. C., Fishes of Panama Bay, 1904: 13. 33 Meek, S. E., & Hildebrand, S. F., Marine Fishes of Panama, 1, 1923: 64. Size: Grows to 19]Q> inches in length. Study Material: None. References: Scymnus brasiliensis, Quoy, J. R. C., & Gaimard, P., in Freycinet, L. de, Voyage autor du monde pendant 1817-1820, 1824: 198 (type locality: off Brazil). Isistius brasiliensis, Garman, S., Mem. Mus. Corny. Zool., 24, 1899: 34, Plate 1, fig. 1, Plate 2 and 3, Plate 49, fig. 2 (synonymy, description, skull, skeleton, internal organs, lateral line system, coloration, phosphorescence, historical notes; 1 specimen 18 inches long, taken 153 miles N by W of Albemarle Island, Galiipagos Islands, in beam trawl). Family Squatinidae. Squatina Dumeril, 1806. Squatina californica Ayres. Angel Shark. Range: South-eastern Alaska to San Diego with two indefinite records from Mexico, and ques- tioned records from Peru and Chile. Field Characters: A flattened ray-like shark with gill openings on the sides in a space between the head and the expanded pectoral fins; ventral fins expanded. Dark brownish or ashy, almost black, or reddish-brown above; white below. Size: Attains a length of five feet and a weight of 60 pounds. Study Material: None. References: Squatina californica, Ayres, W. O., Proc. Cal. Acad. Sci., 1859; 29; ibid., 1860: fig. 7 (original description, outline figures; type local- ity, San Francisco). Garman, S., The Plagio- stomia, 1913: 253, plate 16, figs. 1-4, plate 61, figs. 9-11 (description, figures; range: Cali- fornia and Mexico). Kumada, T., & Hiyama, Y., Marine Fishes of the Pacific Coast of Mexico, 1937: 17, plate 49 (plate and brief diagnosis of a specimen from somewhere along the west coast of Mexico). Squatina squatina, ? Evermann, B. W., & Radcliffe, L., U. S. Nat. Mus., Bull. 95, 1917: 11 (dimensions, short description, color; Lobos de Tierra, Peru). ? Pellegrin, J., Bull. Soc. Zool. Paris, 29, 1904 (Chile). 122 [XXVI: 15 Zoologica: New York Zoological Society Discussion: The status of specimens of the genus Squatina from localities in the eastern Pacific is by no means clear. Thus there exists a tendency to amalgamate the European, the western Atlantic and eastern Pacific angel- sharks under the all-inclusive name Squatina squatina (Linnaeus). In keeping with this, S. californicus Ayres, described from California, has been synonymized with S. squatina , and specimens from Peru and Chile have also been reported under the same name. Whether the angel-sharks from northern and southern localities in the eastern Pacific represent the same species and whether they should be considered as the same as the European squatina, is a question. Our assumption, based on obser- vations of other littoral sharks of the region, is that the eastern fish are not the same as the Atlantic or European form, and that a northern and southern species of the eastern Pacific fish could probably be differentiated. Thus we are uncertain as to the status of the 560 mm. fish recorded from Lobos de Tierra by Evermann & Radcliffe34 and the fish recorded by Pellegrin from Chile.36 Norman36 suggests that the former record may refer to S. armata (Philippi). Two additional species of Squatina have been recorded from the west coast of South America: S. armata Philippi37 described from Iquique, Chile, and redescribed by Norman38 from an Argentine specimen with a question as to its identity, and S. philippi Garman39 from Mexil- lones, Chile. References. (A few of the more commonly quoted papers have been referred to in this paper by name, rather than by reference to their publication place. The full references to these papers are given below.) Garman, S. The Plagiostomia, Mem. Mus. Comp. Zool., 36, 1913: xiii-515. Plates 1 to 75 in separate volume. Gilbert, C. H. & Starks, E. C. The Fishes of Panama Bay, Mem. Calif. Acad. Sci., 4, 1904: 1-304, Plates 1-33. Jordan, D. S. The Fishes of Sinaloa, Proc. Cal. Acad. Sci., (2), 5, 1895: 377-514 (Separate pagination 1-142), Plates 26-55. 39 Evermann, B. W., & Radcliffe, L., U. S. Nat. Mus., Bull. 95. 1917: 11. 35 Pellegrin, J., Bull. Soc. Zool. Paris, 29, 1904: 36 Norman, J. R., Discovery Reports, 16, 1937: 10, 11. 31 Philippi, R. A., Ann. XJniv. Chile, 71, 1887: 561, pi. 7, fig. 1. 3*NormaD, J. R., Discovery Reports, 15, 1937: 10. 39 Garman, S. The Plagiostomia, 1913: 254. Jordan, D. S. & Evermann, B. W. The Fishes of North and Middle America, Bull. U. S. Nat. Mus., 47, vol. 1, 1896: lx, 1-1240. Jordan, D. S. & Evermann, B. W. The Fishes of North and Middle America, Bull. U. S. Nat. Mus., 47, vol. 4, 1900: ci, 3137-3313, Plates 1-392. Kumada, T. & Hiyama, Y. Marine Fishes of the Pacific Coast of Mexico, Nissan Fisheries Institute and Co., Ltd., Odawara, Japan, 1937: 1-75, Plates 1-102. Meek, S. E. & Hildebrand, S. F. The Marine Fishes of Panama, Field Mus. Nat. Hist., Zool. Ser., 15, 1923: xi, 1-330, Plates 1-24. The following check-lists have been checked but not listed in the species references. Beltran, E. Lista de Peces Mexicanos, 13 mimeographed pages issued by Secretaria de Agricultura Y Fomento. Dirr. de Fomento Agricola Instituto Biotecnico, Mexico, D. F. Fowler, H. W. A List of the Sharks and Rays of the Pacific Ocean, Fourth Pac. Sci. Congr., Batavia, Java, 1929, Vol. 3, Biol. Pap.: 481-508. Jordan, D. S., Evermann, B. W. & Clark, H. W. Check List of the Fishes and Fishlike verte- brates of North and Middle America north of the northern boundary of Venezuela and Colombia. Rep. U. S. Fish Comm, for 1928, Part 2, (1930): 1-670. Terron, C. C. Lista de los Peces de la Costas de la Baja California. Anales del Inst. Biol., Univ. Nac. Auton. Mexico, III, 1932: 75-80. Ulrey, A. B. A Check List of the Fishes of southern Cali- fornia and Lower California. Journ. Pan-Pac. Res. Inst., 4 (4), 1929: 2-11. Plate I. Fig. 1. Rhineodon typus. 42-foot specimen from Arena Bank, Lower California. Fig. 2. Rhineodon typxis. Whale shark photo- graphed from motor boat while swimming close to surface, Gorda Banks, Lower Cali- fornia. View of part of head, body, gill- slits and pectoral fin. Two remoras are on the base of the pectoral fin. Plate II. Fig. 1. Triaenodon obesus. Specimen from Bahia Honda, Panama, 1,175 mm. total length. Fig. 2. Eulamia aethalorus. Embryo, 650 mm., taken from 2,070 mm. adult, Arena Bank, Lower California. Fig. 3. Eidamia velox. Specimen from San Lucas Bay, Lower California, 945 mm. total length. BEEBE & TEE-VAN PLATE I FIG. I. FIG. 2 FISHES OF THE TROPICAL EASTERN PACIFIC. PART 2. SHARKS. BEEBE & TEE-VAN PLATE II FIG. 1. FIG. 2. FIG. 3. FISHES OF THE TROPICAL EASTERN PACIFIC. PART 2. SHARKS. 1941] Breder cfe Gresser: Mexican Blind Characin 123 16. Correlations Between Structural Eye Defects and Behavior in the Mexican Blind Characin. Charles M. Breder, Jr. New York Aquarium , & Edward B. Gresser, New York University College of Medicine ( Department of Ophthalmology) . (Plates I-IV; Text-figures 1 & 2). Introduction. Hubbs & Innes (1936) reported the discovery of a new species of blind cave fish inhabiting the warm water caves in the vicinity of Pujal, San Luis Potosi, Mexico, and named it Anoptichthys jordani. This form is genetically continuous with Astyanax mexicanus (Fillipi) from which it is distinguishable only in the undeveloped eye and lack of pigmentation. The eye structure of these fish was found to be imperfectly developed and morphologically arrested at a low level of embryological life without powers of visual function, Gresser & Breder (1940). Especial interest in this species of fish lies in the fact that it is the only known cave species with clear genetic continuity to an outside visual type. Inhabiting warm water and being suitable for laboratory culture, it predicates an experi- mental approach to problems of blind cave fauna that have vexed biologists. The Aquarium Cave Expedition, Bridges (1940), established this genetic continuity and prepared the bases for further study. It was at first considered that the native habitat could best be utilized for the purpose with the establishment of controlled breeding tanks within the cave water system, but practical considerations caused an abandonment of such an approach. La Cueva Chica is roughly eighty miles due west of Tampico and some 350 miles northeast of Mexico City, close to the Pan-American Highway at the village of Pujal. Running nearly due south about 1,000 feet from its entrance, the cave ends in a large cul-de-sac five-eighths of a mile from the Rio Tampaon. The character of the cave water system, its flow and the presence of sink-holes in the probable line of direction between cave and river bespoke a close connec- tion, if not a direct one, between these two systems. Within the cave there were four major pools of water, the innermost of which at the time of our visit was isolated as far as could be determined. The remaining three pools were clearly connected by a small stream. See Bridges (1940) for a map of the cave. Our thanks are due Mr. James W. Atz for his patient assistance in making a considerable por- tion of the observations listed in Table III. Descriptive Anatomy. All the pools were seined to determine the character of the inhabitants. Grossly there were found three fairly distinct groups, one of which was divisible into two subdivisions. The somewhat arbitrary groups follow: (1) obviously blind with deformed, undeveloped eyes; (2) fish with sunken eyes, (a) covered by skin epithelium, (b) not covered by skin; (3) normal eyes. The sampling is given in Table I. Histo- logical studies were made by serial sections covering a considerable amount of the catch. Variations within the groups as to extent of de- velopment both grossly, histologically and to absolute size were numerous. Group 1 has been previously described in de- tail, Gresser & Breder (1940), and consists generally of an ocular sac connected to the skin surface by a connective tissue strand either solid or cystic and lined with epithelium. The ocular structures present are defective as to elements and form, the retinal vesicle usually being a lobulated sac without differentiation into iris, pupillary space or developing sensory elements. A crystalline lens is absent and the anterior chamber is absent or represented by an unlined space in a connective tissue stroma. The optic nerve is represented chiefly by a connective tissue cord. See Plate I. 124 Zoologica: New York Zoological Society [XXVI: 16 In group 2a the covered, sunken eye presents a better developed globe although malformed as to size and shape. The choroidal gland is usually well developed and the retina is either completely deficient in sensory elements and ganglion cells or large areas of retina are so deficient. A crys- talline lens when present is smaller than normal and is accompanied by an irregularly oval-shaped pupillary space; where absent the iris is a weak membrane forming a complete anterior wall to the retinal vesicle. In such conditions an ex- tremely shallow anterior chamber can be found. See Plate II. The uncovered, sunken eye, group 2b, varies greatly in detail but generally is of small size, deformed but with an anterior chamber, an iris quite well developed with a pupillary space. A lens is present but smaller in diameter in relation to globe diameter than normal. The retina is better stratified than in previous types but de- ficient irregularly in the rod-cone layer and the corresponding ganglion cell layer. The optic nerve reflects this deficiency in size and relative absence of nerve fibers. See Plate III. The orbital cavity in group 2 is well developed and where the globe is small or deformed there is a corresponding increase in orbital fat. The extraocular muscles are intact in all groups but apparently their functional use is limited in ac- cordance with the functional capacity of the eye. In the specimens examined, no symmetry of binocular development could be determined in individuals of groups 1 and 2; especially in group 2 asymmetry was marked as to morphological and functional development. Group 3 was in- distinguishable from the normal river fish but microphthalmia was frequent. The normal typically piscine eye structure is shown in Plate IV, based on a specimen from the Rio Tampaon. Obviously the pool formations present formed ideal and natural breeding tanks and presented in loco a highly suggestive explanation for the eye defects and probably as superior and factual experimental methods as human endeavor could devise and control. As was noted above, population studies in the most isolated pool No. 1 showed a nearly pure strain with completely ineffectual eye structures with an increase in percentage of better developed ocular organs as the probable connection with the outside stream was approached, so that in pool No. 4, closest to the Rio Tampaon, were con- tained a large percentage of structurally normal eyes. Most of the inhabitants of pool No. 1 con- formed to the ocular pattern first described and which originally were obtained from this pool. Inhabitants of pool No. 2 showed a small per- centage of intermediate eyes in which variations in size of the globe, pupillary openings and tissue structures were found but none of which pre- sented morphologically an image-forming organ. In pool No. 3 intermediate eyes were found in greater proportion. In this pool occasional in- habitants showed normally sized eyes, struc- turally well developed and which clearly per- mitted definite light behavior patterns. These fish also showed a greater amount of skin pigmen- tation. This data is detailed in Tables I and II. Of the inhabitants seined from pool No. 4, only occasional specimens were of least develop- ment with a preponderance of better-developed eyes and an increase in the proportion of normals both as to eye structure and pigmentation. Table I. Eye Condition and Pigmentation of Cave Characins. Expressed in % of catch. Based on 119 specimens. Eye Condition Sunken Eye (2) Covered Uncovered “Normal Location Blind (1) (a) (b) Eye" (3) Pool 2 85 6 — 9 Pool 3 16* 8 45 31 Pool 4 9 Pigmentation 9 82 Consid- Location None Little Moderate erable Full Pool 2 90 2 6 2 — Pool 3 34* 34 5 8 19 Pool 4 3 29 32 24 12 * Two specimens in this group blind on one side only. Table II. Association of Eye Condition and Extent of Pigmentation. Expressed in % of extent of pigmentation. Based on 119 specimens. Extent of Pigmentation in Per Cent. Eye Condition None Little Moderate Consid- erable Full Blind (1) 98* — 2 — — Sunken Eye 56 33 11 — — (covered) (2a) Sunken Eye(un- covered) (2b) — 62 15 15 8 “Normal” (3) 2 23 28 23 24 * Two specimens in this group blind on one side only. Behavior Experiments. Fishes representing the various types described in the preceding section were shipped alive to the New York Aquarium and form the bases on which the following experiments were made. In addition to these, specimens of the fully blind type which had been bred to the fifth generation by Mr. Albert Greenberg of Tampa, Florida, were used for comparative purposes. These had all been reared in brilliant light. Since Mr. Greenberg never obtained any but fully blind fish it is evident that this form breeds true to type. Even casual observation shows that these optically various fishes present different types of locomotor behavior. The fully blind individuals, both direct from the cave and Greenberg’s fifth generation fish, all keep up a continual swimming activity, constantly wandering in a seemingly 1941] Breder & Gresser: Mexican Blind Characin 125 aimless manner. Even at night, if a light were suddenly struck, they would be found to be moving about, giving the impression that they managed to keep on the move at all times. The normal river fish, on the other hand, unless feeding or engaged in some other activity, would remain quiescent in a school, only moving under some evident external stimulus. Those with intermediate eyes unable to form a retinal image but clearly able to receive an optical stimulus from a light source, behaved like the fully blind while those forming a defective image behaved like the normally eyed fish. Normally eyed fish brought from the cave at first acted nearly like the blind groups but later took on the habits of the typical river fish. It would thus appear that these fishes, unaccus- tomed to retinal images, took an appreciable time to be able to react to them in an appropriate fashion. In order better to understand the relation of eye defect to behavior as measured by their locomotor habits, troughs were established having a lighted and a dark end. These were so arranged that a given area was illuminated with a 60-watt bulb at a distance of three feet while the remainder was shadowed by a cover close to the surface of the water. Thus it was anticipated that by noting the positions of the fish in unit intervals their preferences could be determined in a quantitative manner. In the case of a ran- domly wandering fish the number of observations obtained over a given area of bottom should then be directly proportional to the whole area if there were no preference being expressed. Sig- nificant deviations from such a figure would then be a measure of preference. As a preliminary experiment a trough was set up 1' by 4' with water 6" deep. Three feet of its length was covered; the remaining area was ex- posed to light. In all cases the entire tanks were painted black in order to minimize reflections. Into this were placed four fully blind fish of the fifth generation. If these fish were moving strictly at random it then follows that the average of the unit observations which were made Table III. Results of Behavior Experiments. % Of % of % of Random % of Random Exp. Area No. of Fish Obs in Expec- Exp. Area No. of Fish Obs. in Expec- No. Lighted Obs. No. Type Light tation No. Lighted Obs. No Type Light tation 1 .25 100 4 5th Generation 73 73 42 .50 100 1 37 74 2 .25 100 4 111 111 43 .50 100 1 36 72 3 .25 100 4 88 88 44 .50 100 1 29 58 4 .25 100 4 54 54 45 .50 100 1 15 30 5 .25 100 4 96 96 46 .50 100 1 55 110 6 .25 100 4 81 81 47 .50 100 1 46 92 7 .25 100 4 45 45 48 .50 100 3 133 88.6 + 8 .25 100 4 55 55 49 .50 100 3 114 76 9 .25 100 4 85 85 50 .50 100 1 No Image Cave 43 86 10 .25 100 4 67 67 51 .50 100 1 42 84 11 .50 100 4 131 65.5 52 .50 100 1 11 22 12 .50 100 4 133 66.5 53 .50 100 1 26 52 13 .50 100 2 29 29 54 .50 100 1 21 42 14 .50 100 2 35 35 55 .50 100 1 11 22 15 .50 100 2 24 24 56 .50 100 1 73 146 16 .50 100 1 42 84 57 .50 100 1 87 174 17 .50 100 1 23 46 58 .50 100 1 7 14 18 .50 100 1 24 48 59 .50 100 1 12 24 19 .50 100 1 12 24 60 .50 100 1 21 42 20 .50 100 1 28 56 61 .50 100 1 21 42 21 .50 100 1 37 74 62 .50 100 1 31 62 22 .25 100 2 Blind Cave 42 84 63 .50 100 1 25 50 23 .25 100 2 61 122 64 .50 100 1 23 46 24 .50 75 1 30 80 65 .50 100 1 26 52 25 .50 100 1 Blind Cave 34 68 66 .50 100 1 Image Cave 48 96 26 .50 100 1 50 100 67 .50 100 1 3 6 27 .50 100 1 51 102 68 .50 100 1 3 6 28 .50 100 1 48 96 69 .50 100 1 0 0 29 .50 100 1 40 80 70 .50 100 1 0 0 30 .50 125 1 17 27.2 71 .50 100 1 0 0 31 .50 100 1 37 74 72 .50 100 1 0 0 32 .50 100 1 38 76 73 .50 100 1 0 0 33 .50 100 1 27 54 74 .50 100 1 0 0 34 .50 100 1 31 62 75 .50 100 1 *• ‘i* 9 18 35 .50 100 1 53 106 36 .50 100 1 21 42 Maximum 174 37 .50 100 1 44 88 Average 38 .50 100 1 43 86 61.4- 39 .50 100 1 45 90 Minimum 0 40 .50 100 1 51 102 41 .50 100 1 46 92 * Not the same specimen used in Experiments 66 to 74. 126 Zoologica: New York Zoological Society [XXVI: 16 at five second intervals should come to the figure “1.” That is, on strictly random distribution there should be three fish in the dark area and one in the light, or if the fish were schooling there should be four fish visible on the average every fourth observation. In a series of such tests, in which there were 2,100 such observa- tions, the percentage of expectation was 62.3 — % in light. This data is given in detail in Table III. Since experiments consist of 100 observations, the expectation from random movement should give 100%. Only one observation reached that figure. As the illuminated area of the trough was changed from time to time in order to avoid the inclusion of some other possible but unknown factor, it can only be concluded that by some means these fish were light sensitive to a slight degree and reacted negatively to such radiations or some secondary associated effect. Subsequent observations were made in smaller troughs l'x2'x 6" deep in which one-half was illuminated and one-half in darkness. Here fish in various smaller numbers were similarly checked and found to show like reactions, as is also set forth in Table III. Other fishes were then checked in a similar manner and these data are also given in Table III. They are summarized and presented in a con- densed form in Table IV. From these tables it is evident that the fully blind fish brought from La Cueva Chica reacted in a manner similar to those of the fifth generation reared in light: 80% avoidance as against 62%. This difference of 17% may actually be significant. The cave fish were larger than the others and possibly over- lying tissues of greater consequent thickness may account for the difference, or it may be that there is an increased avoidance to light in subsequent generations. Table IV. Averages and Ranges of Experiments. Exp. No. Type of Fish No. of Obs. % of Random Expectation Max. Mean Min. 1 to 21 5th Generation 2,100 Ill 62.3- 24 22 to 49 La Cueva Chica (Blind) 2.800 122 79.6+ 27 50 to 65 La Cueva Chica (No image) 1,600 154 58.7+ 14 66 to 74 La Cueva Chica (Image) 900 96 12.0 O 75 La Cueva Chica (Image) 100 18.0 — — River Fish (Normal) Fully unreadable Experiments 50 to 65 were performed on a fish that could distinguish light but not form an image. Fish of this group, too, avoided light, 59%, even more vigorously, but here was clear optical detection. Whether or not these fish form a useful retinal image is easy to detect. Fully blind ones, although feebly light-sensitive in the above sense, will not respond to a shadow intercepting the light falling on a tank. Those with full or partial vision react violently by dashing about when a shadow passes across them. The feeding reactions separate the latter two types. Those with image formation will come to the surface or near it, roll their eyes and make energetic strikes at the food particles as they sink through the water. The blind or merely light-detecting individuals react in no such fashion, merely cruising about mostly on the bottom and apparently finding their food by energetic random movement. One individual seemed to be just at the border line of image formation, normally cruising about like the blind but occasionally finding a sinking particle if it fell between the fish and the light source. Experiments 66 to 74 were performed on a fish with defective eyes but which could evidently form some sort of image. Experiment 75 was performed on another specimen of similar condi- tion. Here the percentage of expectation of the fish in the light was clearly much lower than in any of the others. Actually these fish spent most of their time just under the shelter of the shadow, obviously looking out into the lighted area, not wandering about but resting quietly. Table V. Aggregation Data on Blind Specimens. Nos. of observations of fish by groups Exp. No. in lighted area 0 1 2 3 4 1 48 35 14 2 1 2 34 31 26 8 1 3 38 37 24 1 0 4 52 42 6 0 0 5 32 46 17 4 1 6 43 37 16 4 0 7 62 31 7 O 0 8 59 28 12 1 0 9 35 48 14 3 0 10 44 44 11 1 0 Average U.7 37.9 14.7 2.4 0.3 11 18 46 24 11 1 12 17 41 34 8 0 Average 17 .5 43.5 29.0 9.5 0.5 13 73 24 3 14 71 23 6 15 79 18 3 Average 7 If . 0 27.0 4-0 22 65 28 7 23 48 43 9 Average 56.5 36.5 8.0 48 16 42 35 7 49 19 48 32 1 Average 17.5 45.0 33.5 4-0 Comparison with Theoretical Probability. 4 fish }4 area 31.6 42.2 21.1 4.7 0.4 Exp. 1-10 44.7 37.9 14.7 2.4 0.3 4 fish Y area 6.25 25.0 37.5 25.0 6.25 Exp. 11-12 17.5 43.5 29.0 9.0 0.5 3 fish Yi area 12.5 37.5 37.5 12.5 Exp. 48-49 17.5 45.0 33.5 4.0 2 fish Yi. area 25.0 50.0 25.0 Exp. 13-15 74.0 27.0 4.0 Exp. 22-23 56.5 36.5 8.0 FULLY LIGHT NEGATIVE 1941] Breder & Gresser: Mexican Blind Characin 127 Normal river fish acted much as the above, but stayed in the light continually unless dis- turbed by a slight noise or jar to the building. Consequently they were unreadable by this technique since their movements were controlled by external stimuli which were, so far as possible, prevented from reaching them during the course of the experiments. Since the normal river fish are a closely aggre- gating and to a considerable extent a schooling type, the annotations previously discussed were examined in reference to the appearance of groups of blind individuals where more than one fish was used in an experiment. These groups were used primarily to discover if any social influences were at work in regard to the light-avoiding re- action. Nothing of this sort appeared. Since it is evident that the seeing forms show a strong aggregating tendency, the data accumu- lated on the blind specimens were analyzed so as to show evidences of similar behavior in these fishes if such were present. The data are given in Table V and shown graphically in Text-figure 1 which indicates the results of these tests in reference to random expectation. As a definite number of fishes was used in each experiment and the lighted bottom area bore a definite relationship to the unlighted, 34 and 34> in the two types of tanks used, it follows that if the fish were moving at random without reference to light or each other their occurrence in the lighted area in ones, twos, threes, etc., should be predictable according to the binomial formula (p + q)n expanded for each term, in which n = number of fishes p = lighted area q = dark area (1) (2) npn~lq (3) n(w-1) p»- Y 1-2 1 1 (4) "(w-U(w-2) 303 1-2-3 (5) n(n-l)(n-2)(n-3) 4 4 1-2-3-4 These calculations are given in Table V for comparison with the observed figures for the several types of experiment. Some rather inter- esting features appear which are more clearly shown in Text-figure 2 which presents the com- parative data graphically. It may be seen in the < h 2 UJ a: z U 1- o u 2 Z u, o o UJ o a. L. CL UJ < 1 1 u. 2 a Q li. - z Z 5 O _l _» z Z CO CD Text-figure 1. Comparison of behavior in various experiments on light sensitivity in terms of expectancy of random movement. FULLY LIGHT POSITIVE 128 [XXVI: 16 Zoologica: New York Zoological Society data of Table V and in Text-figure 2 that in each case there were more observations of the lighted area without fish than random expectation calls for, which is another way of measuring these fishes’ light avoidance. In Text-figure 2, graph A, it is clear that the occurrence of the fish in the lighted area in groups follows closely that of the theoretical probability. All appear as less than the expected random value, which again is a measure of their individual light avoidance; the slopes of the connecting lines are in close agree- ment with random expectation. This is the measure of interlocked (aggregating) behavior and is clearly shown to be a good experimental zero. This graph is based on the largest number of experiments and shows closer agreement with the theoretical than the others based on smaller numbers. In graph B of Text-figure 2 both the observa- tions noting “no fish” and “one fish” are greater than random expectation while the rest are below it. This is also true for graph C. Except for this feature they are in agreement with graph A although based on much less data. Graph B represents the same fish (fifth genera- tion) as in A while C represents fish brought from the cave. Whatever the interpretation placed on this, these two groups were showing practically identical behavior. It is suspected that the small number of experiments, 3, give as good approx- imation as could be expected. This becomes more apparent when graph D is considered. Here two sets of different fish show close agree- ment with A, the “0” observations being above random expectation and the rest below with the trend of these lines being very nearly parallel. It can only be concluded that these fish were moving strictly at random in reference to each other but with a marked avoidance of the lighted area. Discussion. From the preceding descriptive histological data, the reactions of the various types with more or less defective vision, coupled with the experi- mental data on the avoidance of light and the social attitude of the blind examples, certain features become apparent that should have great significance in the development of cave-dwelling blind fish. If we operate from the assumption that blind and pigmentless animals are able to establish themselves in caves because therein both eyed and eyeless forms are on equal terms, we have two basic factors in effect and interacting; a genetic defect and an environmental peculiarity, the presence of the latter making a continuity and establishment of the former possible. In addition to these it is here experimentally shown that there is a third, a behavior factor, which operates in a positive way. Since the eyed forms use dark retreats only under the stimulus of fright and the blind ones avoid light, by a mechanism as yet not under- stood, we have a marked schism in the behavior of fishes otherwise identical and of genetic con- tinuity. Consequently it would seem fair to make the following assumptions on the operation of these factors in a state of nature. Given a population of river fish carrying one or more appropriately defective genes, individuals with slightly defective sight would seek the shelter of rocks and crevices more than their normally eyed fellows. A cave from which water leaves (as La Cueva Chica) would encourage entry, as these fish tend to work their way up small streams (contranatant) and hold their position in the upper reaches of the rivers by this means. Selection might enter at this point, i. e., fish with defective eyes that might happen to prefer the light would not stay or enter the cave and would either be lost or mixed with the river population, depending on the extent of the handi- cap of the visual defect. Those that happened to prefer the dark would stay there and transmit their defect, increasing and intensifying it by consanguinity. Once unable to form an image they would not seek light, and if they emerged at night would retreat with sunrise or be lost, as witness what happens in a lighted aquarium between image-forming and non-image forming cave individuals. It appears that the seeing fish attempting to school with the sightless cannot keep up with their continual aimless wandering. This leads to a “resentment” that frequently results in the eyed form attacking and killing the blind. There is no evidence of a substitute sensory mechanism having been developed in the blind fish, for the changing of the position of a rock in an aquarium will result in their swimming into it for a time. Apparently they merely learn where obstacles are by a trial and error develop- ment of some sense of space and distances. One specimen with image formation and miniature eyes was successfully kept with two fully blind individuals but took on their habits after about a month and showed no evidence of attack but spent much time following the blind ones. It is probably more than accidental that this fish exhibited great nervousness, dashing about wildly on approach to the aquarium. Returning to the discussion of natural conditions, it would seem that those which found their way out of the cave, and with good eyes, might stay out; how- ever, the current tends to keep them in, as these fish will not normally drop down stream and the period necessary for them to develop an under- standing of a retinal image might well be their undoing. Thus the river and cave populations would tend to stay fairly separate with perhaps the continual entry of defective-eyed fishes account- ing for the gradient of eyelessness within the cave. The question of whether the present population was once fully cut off geographically from the river fish is rather pointless in the light of the above. Certainly some such condition as men- tioned above was present at the time of original entry. If some geologic cataclysm cut off these fish for a time, they rejoined sufficiently soon to make a reestablishment of the original condition 1941] Breder & Gresser: Mexican Blind Characin 129 0 I 2 NO. OF FISH IN LIGHTED AREA Text-figure 2. Comparison of theoretical randomness with actual behavior showing lack of schooling behavior and negative reactions to light in fully blind fish. Solid line: mathematical probability. Dotted line: observed behavior. 130 Zoologica: New York Zoological Society possible, for otherwise the gradient would be quite inexplicable. If it is thought that some reestablishment of Lamarkian ideas are indicated, it may be pointed out that if the blindness were indeed induced by such a mechanism, the present data do nothing to establish it. Fish that have been forced to live in light for five generations are apparently even more “anxious” to get back into the cave than the present dwellers therein. If this means an increasing sensitivity to light (reestablishment of an eyed condition), it would defeat progress in that direction, for those in the process of regain- ing vision are the ones most anxious to avoid the stimulus inducing it. Such a mechanism should make for a status quo. A study of possible histological changes in the successive generations of those reared in light for a greater number of generations could help in this connection. Consequently we are forced to the conclusion that the origin of this blindness is rooted in some genetic defect that was able to find expression in a lightless environment and that the differential behavior of those individuals able to form an optical image as opposed to those merely light sensitive is the determining factor in keeping the population of the cave from a depleting voluntary exit and from too great an admixture of the oc- casionally entering outside river fish with good vision. The details of the genetic ratios thus become, as far as an explanation of this condition goes, merely of importance insofar as it concerns the speed of the process. Thus if there is a genetic state that produces large numbers of blind fish per generation it would move relatively fast, whereas if the blind individuals are genetically fairly rare it would move more slowly to reach the present state of affairs as found in La Cueva Chica. The continual locomotor activity and lack of any social grouping on the part of these blind fish would suggest that this feature of behavior has been obliterated, for insofar as experimental evi- dence goes, schooling is dependent on vision alone in other more or less similarly constituted fishes. Breder & Nigrelli (1935 and 1938) discuss the role of vision in schooling and give full bibli- ographies. Since thus far no substitute sensory mechan- ism has been demonstrated in this form and since nothing in their behavior indicates the presence of such, we infer that these fish are managing to survive in their specialized environment while laboring under a simple loss of function. This may mean that in an evolutionary sense this group is of relatively recent establishment as compared with other blind cave fish that possess elaborate non-visual sensory organs. Park, Roberts & Harris (1941) have shown that the crayfish of Mammoth Cave, Cambarus pellucidus Tellkampf, is light negative in a manner that seems to be fairly comparable to the present findings insofar as it is possible to [XXVI: 16 compare a sedentary invertebrate to an active fish. A study of the exact means whereby these fish are enabled to react to the presence of light or on associated phenomena is now being undertaken. SUMMARY. 1. Fully blind characins from La Cueva Chica are slightly negative to light or some associated effect, as are their fifth generation descendants reared in light. 2. There is a sharp break in the locomotive behavior pattern between those which can form a retinal image and those which cannot. 3. Individuals which form a retinal image have a strong schooling instinct and those which do not, evidence none. 4. Individuals with vision tend to rest quietly in a compact school while the blind continually wander about in an apparently aimless manner. 5. In light, attempts by seeing individuals to school with the blind often terminate by the former killing the latter. 6. Blind individuals in a newly rearranged aquarium will swim into obstacles but will avoid them after a time, apparently by some process of learning. 7. Apparently no substitute mechanism has been developed to function as a substitute for the lost vision. 8. The differential behavior existing between the blind and the seeing is apparently an addition- al positive factor in the establishment of this cave form. Bibliography. Breder, C. M., Jr. & Nigrelli, R. F. 1935. The influence of temperature and other factors on the winter aggregations of the sunfish, Lepomis aurilus, with critical re- marks on the social behavior of fishes. Ecology, 16 (1): 33-47. 1938. The significance of differential locomotor activity as an index to the mass physiology of fishes. Zoologica, 23 (1): 1-29. Bridges, W. 1940. The Blind Fish of La Cueva Chica. Bull. N. Y. Zool. Soc., 43 (3): 74-97, photos 23, map 1. Gresser, E. B. & Breder, C. M., Jr. 1940. The histology of the eye of the cave characin, Anoptichthys. Zoologica, 25 (10): 113-116, pi. III. Hubbs, C. L. & Innes, W. T. 1936. The first known blind fish of the family Characidae: a new genus from Mexico. Occ. Pap. Mus. Zool. Univ. Mich., (342): 1-7, pi. I. Park, 0., Roberts, T. W. & Harris, S. J. 1941. Preliminary analysis of activity of the cave crayfish, Cambarus pellucidus. Amer- ican Naturalist, 75: 154-171. 1941] Breder & Gresser: Mexican Blind Characin 131 EXPLANATION OF THE PLATES. All fish on the same scale. All eye sections on the same scale. Photographs of fish by S. C. Dunton. Photomicrographs by M. Sparago. Plate I. Fig. 1. Unpigmented specimen from La Cueva Chica, Pool No. 2. 53 mm. standard length. Fig. 2. Section of eye of the above specimen, show- ing sunken and overgrown eye with cystic development. (Group 1.) Plate II. Fig. 3. Scarcely pigmented specimen from La Cueva Chica, Pool No. 3. 47 mm. standard length. Fig. 4. Section of eye of the above specimen, show- ing no exterior chamber, lens or pupillary area development and retina partially developed. (Group 2a.) Plate III. Fig. 5. Partially pigmented specimen from La Cueva Chica, Pool No. 3. 58 mm. standard length. Fig. 6. Section of eye of the above specimen, show- ing uncovered sunken eye with well devel- oped form, iris poorly developed, retina, choroidal gland and optic nerve undevel- oped. (Group 2b.) Plate IV. Fig. 7. Normal river fish from the Rio Tampaon. 67 mm. standard length. Fig. 8. Section of eye of the above specimen, show- ing the normal optic constitution of these fishes. (Indistinguishable from Group 3.) BREDER & GRESSER. PLATE I. FIG i FIG. 2. CORRELATIONS BETWEEN STRUCTURAL EYE DEFECTS AND BEHAVIOR IN THE MEXICAN BLIND CHARACIN. BREDER & GRESSER PLATE II FIG, 3. FIG. 4 CORRELATIONS BETWEEN STRUCTURAL EYE DEFECTS AND BEHAVIOR IN THE MEXICAN BLIND CHARACIN. BREDER a GRESSER PLATE III. FIG. 5. FIG. 6. CORRELATIONS BETWEEN STRUCTURAL EYE DEFECTS AND BEHAVIOR IN THE MEXICAN BLIND CHARACIN BREDER & GRESSER. PLATE IV. FIG. 7. FIG. 8. CORRELATIONS BETWEEN STRUCTURAL EYE DEFECTS AND BEHAVIOR IN THE MEXICAN BLIND CHARACIN. 1941] Delacour: On the Species of Otus scops 133 17. On the Species of Otus scops. Jean Delacour. Owls have long been considered among the most difficult birds to understand, particularly in regard to systematics. Their secretive ways have made it difficult to collect suitable series, while the numerous plumage phases within the same form are often extremely puzzling. Among them, none are more confusing than those of the genus Otus, an ubiquitous group containing the small species known as “Scops owls” in Europe, and “Screech owls” in America. This could, of course, be expected, since nocturnal habits, color and geographical variations are exhibited in the extreme among them. As early as 1875, Sharpe, in the Catalogue of Birds in the British Museum (Yol. II. p. 44), wrote: “Difficult to understand as all owls are, the species of the genus Scops 1 are in every way the most difficult to identify. The impossibility of procuring series of some of the species to study at the same time; the absence of information as to the sequence of plumages from the young stage to that of the adult, and the puzzling way in which some species seem to possess rufous phases, while others do not — these are all prob- lems that only time can solve.” Time has certainly brought us a wider knowl- edge, but the classification of the Scops owls still leaves much to be desired. Authors differ widely, even today, as to the relationships and affinities of many forms. Particularly those of the smallest birds, related to the European Scops, seem to have remained in a state of con- fusion. The present paper is an attempt to clear it up as much as possible. The small Scops owls, resembling Otus scops scops, are widely distributed in Europe, Asia, Africa and, curiously enough, are represented in western North and Central America. They are absent from South America, the Malay and Australasian Islands. To the unprejudiced stu- dent, these birds are all obviously so closely re- lated that it is impossible to consider them but as mere subspecies of one species, since it is well known that none of them breed on the same ground. Whenever two forms have been found together, one has always proved to be a winter migrant. Scops owls, like many other avian types, are more or less migratory, according to distribution. The species is a dweller in temperate or tropical climates. Nowhere is it found far to the north. In the cooler regions of Europe, Asia and even perhaps northwestern America, it is only a sum- mer nesting visitor, migrating south as soon as the temperature becomes too low. Elsewhere, it does not wander much. It is found at both low and high altitudes, to the latter of which a few forms seem to be confined. As can be expected, the migratory forms have longer and more pointed wings than the resident races. Two facts have contributed to the obscuration of the status of these owls: first, the presence of different gray, brown and red color phases and their intermediates, which formerly caused much confusion but are now better understood; second, the unfortunate method of classifying them ac- cording to the wing formula, which has induced authors to consider as different species birds really hardly separated subspecifically. The rela- tive length of primaries is a late-acquired and relatively unimportant character; a more careful study shows that it is even quite variable in many forms. It certainly cannot stand as of specific value. The figures of wing contours, found in the course of this study, show some of the more typical formulas in different forms. But they represent only one or two of the more frequent formulas. In almost all subspecies there are numerous and important variations. It has been the mistake of many ornithologists of the last century to base generic and specific distinctions on easy, but often unimportant, characteristics such as wing formula, number of rectrices, etc. When undertaking a general study of the more difficult groups of birds, one is astounded to find that forms so closely related as to be obvious geographical races of the same species often are still widely separated in standard works. Unless a serious revision, based on actual series of specimens, be made, students certainly may be excused for not suspecting affinities which the literature does not suggest. 1 Now called Otus 134 Zoologica: New York Zoological Society [XXVI: 17 Such is the case of our Scops owls. In the Catalogue of Birds in the British Museum, Yol. II, 1875, cited above, Sharpe made excellent use of the materials and of information available in those days. Particularly he states (pp. 45 and 106) that Scops flammeolus is the American representative of Scops giu ( Otus scops scops) and that many forms should be considered as races of the European bird. He believes that Scops senegalensis is conspecific with Scops giu, being “indistinguishable in color,” and only smaller (pp. 51-52). He calls subspecies the following forms: capensis ( senegalensis ), pennatus ( sunia ), stictonotus, japonicus, malayanus and rufipennis. Further on (pp. 100-102) he recognizes that the two Andaman Scops owls, balli and modestus, belong to two different species, all of which is correct. In time, ornithologists arbitrarily came to separate specifically different forms on the basis of the wing formula. In the last published general list, J. L. Peters (Birds of the World, Vol. IV, 1940) rightly con- siders all Asiatic forms as subspecies of Otus scops, stating in a footnote: “Some authorities consider the Scops owls of Eastern and South Eastern Asia as specifically distinct by reason of their wing formula; other things considered, this difference hardly seems grounds for specific distinctness.” But he fails to recognize as such the African and American birds, although they are barely distinguishable in shape, color and wing formula. My attention was called to the unsatisfactory arrangement of these owls when H. Friedmann & H. G. Deignan ( Journal of the Washington Academy of Sciences, 20, No. 7, July 15, 1939, pp. 287-291) described a new race, distans, from two Siamese specimens, as a subspecies of Otus senegalensis to which it is truly very similar. That all the more sedentary tropical forms of Scops, small and round-winged, are very closely related cannot be denied, and the new distans, later on also found in South Annam, is curiously similar to the Angolan hendersoni, as was pointed out. But it is not very different either from the Indian sunia, particularly in its recently dis- covered extreme red phase, never found among African birds. The same sunia in its gray phase also closely resembles senegalensis from north- western Africa. It did not, therefore, seem acceptable to consider distans as a subspecies of senegalensis while sunia and other Asiatic forms were left within the species scops. Another interpretation had to be found. The study of materials now at my disposition in America has lead me to the following con- clusions : All the small Scops owls of Europe, Asia and western North and Central America, must be considered as geographical representatives — that is to say, as subspecies — of the species Otus scops. Differences in size, wing formula and coloration remain well within the normal range of racial variations, the northern migratory birds being duller and having longer and more pointed wings than the more sedentary tropical ones. None of the different Scops owls of this group breed in the same territory, and wherever they meet, it is only on the winter grounds of northern migrants. All these owls have similar habits, behavior and voice, varying only in a minor degree. The position of Otus brucei, from western Asia, is somewhat peculiar. It looks rather like a pale gray, lightly marked desert edition of Otus scops. However, it occupies large areas where the latter also lives and breeds. On close examination it proves to have longer tarsus and tail, and, according to Cheesman, its voice is completely different. No intermediates between these two Scops seem ever to have been found. It must be regarded as a different species. (See Dementiev, Syst. Av. Ross., Paris, 1935, p. 50.) Otus balli, from the Andaman Islands, where O. s. modestus also is found, stands as a separate species, being a larger, heavier bird, with a different color pattern. The very rare Otus leucopsis from S. Thom6 appears also to be a good species. The small Scops owls of the rutilus, umbra, manadensis and spilocephalus groups con- stitute also as many species, for the same reason. Furthermore, several races of 0. spilocephalus live on the same grounds as O. scops. I cannot decide whether or not the Philippine forms of Scops owls, as admitted by Peters, really belong to the species Otus scops ( calayensis , longicornis, mindorensis, romblonis, cuyensis). I could examine only two specimens of the latter, which seems to be too different in size and color pattern to be included, none of the others having been available. The same can be said of man- tananensis, from Mantanani Island, north of Borneo. In his “Birds of the Philippine Islands ” (London, 1934), Marquess Hachisuka does not consider any of the Philippine Scops owls as close- ly related to Otus scops. As none of the types and series preserved in European collections can at present be examined, I find it impossible to proceed now to a complete revision of the species Otus scops. I can only at- tempt here to establish a tentative list of its forms, leaving to the future further considerations and appreciations as to the validity of several sub- species. In such cases, I have provisionally accepted Peter’s conclusions in his “Birds of the World.” I am simply giving for each form its geographi- cal distribution, formula and dimensions of wings, while stressing interesting points on color phases and plumages from actual examination and also from literature whenever necessary. For general descriptions of the species, I refer my readers to those of Otus scops scops easily found in standard works, which can be used for comparison with other subspecies. All forms of Otus scops have the tarsus feath- ered to the base of the toes, except in a few cases which are mentioned below. Their eyes are usually yellow, but in some forms they vary to brown ( sunia and allies) , while in others they are always dark ( flammeolus ). 1941] Delacour: On the Species of Otus scops 135 In the present study, I have been greatly helped by my friends Dr. J. Chapin, Dr. E. Mayr and Mr. J. Zimmer, of the American Museum of Natural History, New York; Dr. H. Friedmann and Mr. H. G. Deignan, of the National Museum, Washington; Mr. R. M. de Schauensee, of the Academy of Natural Sciences, Philadelphia; and Mr. J. C. Greenway, of the Museum of Compara- tive Zoology, Cambridge, who kindly loaned me the specimens in their care. I have made particular use of the valuable in- formation found in the more recent works by Hartert, Mr. W. L. Sclater, Mr. E. C. Stuart Baker, Dr. D. A. Bannerman, Dr. J. Chapin, and I am much indebted to the researches of Dr. E. Stresemann (Mitt. Zool. Mus. Berlin, 12 (1) 1925, pp. 191-195); Dr. P. Dementiev (Syst. Av. Ross., Paris, 1935, pp. 49-51); Dr. H. Friedmann and Mr. H. G. Deignan, and of L. Peters (cited above). The drawings of wings were executed by Mr. Lloyd Sandford. I tender them all my grateful thanks. 1. Otus scops scops. Text-figure 1. Otus scops scops, o' Lanusci, Sardinia, 2-1-1901. (Amer. Mus. Nat. Hist., New York.) Strix scops Linnaeus. Syst. Nat., ed. 10, 1, 1758, p. 92. (Europe, restricted to Italy.) Distribution: Breeds in Central and Southern Europe, north to central France, northwestern Germany, the Alps, Austria, Hungary, south- western Russia, the Canary Islands, North Africa. Winters in West Africa, east to Ethiopia and south to Uganda. Wing Formula: 3rd primary longest, 2nd sub- equal; 1st primary long, equal to 6th, or between 5th and 6th, or 6th and 7th. Dimensions of Wings: c? 147-161 mm. $ MS- 166 mm. Color Phases: Amount of gray and reddish- brown variable, producing gray and brown phases, but no clear red phase. The markings vary much in intensity. 2. Otus scops cycladum. Pisorhina scops cycladum Tschusi. Orn. Jahrb., 15, 1904, p. 21. (Naxos). Distribution: Resident in the Cyclad Islands and Crete. (Not examined.) 3. Otus scops pulchellus. Stryx pulchella Pallas. Reise Versch. Prov. Russo. Reichs, 1, 1808, p. 456. (Siberia.) Distribution: Breeds in Russia, east of Longi- tude 35° East and north to Latitude 56°; Cau- casus, southwestern Asia north to Irtysh, east to Krasnoyarsk, south to Kirghiz Steppe, Ferghana, Tabargatai and the Altai. Winters in the upper Nile Valley, southwestern Asia and northwest India. Wing Formula: Same as in Otus scops scops. Dimensions of Wings: o’ 150-164 mm. 9 156- 163 mm. Color Phases: Very close to Otus scops scops, but brown phase almost absent ; only a few speci- mens have a little more rufescent tinge than others. General color more grayish and more uniform, particularly on the upper parts. (See P. Dementiev, Systema Avium Rossicarum, Paris, 1935, pp. 49-51.) 4. Otus scops turanicus. Pisorhina scops turanicus Loudon. Orn. Monatsb., 13, 1905, p. 129. (Ivara-Korum, Transcaspia). Distribution: Transcaspia, Bukhara, northern Persia, Armenia? Wing Formula: Like Otus scops pulchellus. Dimensions of Wings: Similar to precedent. Color Phases: According to Dementiev ( loc . cit.). This form is lighter still than pulchellus; pattern very fine, but well marked with more white markings. The one specimen in the Ameri- can Museum of Natural History, from Mt. Asilmadog, Transcaspia, answers well the above description. Spots and shades very clear on pale ground color. 136 [XXVI: 17 Zoologica: New York Zoological Society 5. Otus scops cyprius. Scops cypria Madarasz. Termes, 24, 1901, p. 272. (Livadia, Cyprus.) Distribution : Island of Cyprus. (Resident.) Wing Formula: Same as in Otus scops scops. Dimensions of Wings: VOLUME XXVI Part 3 Numbers 19-27 PUBLISHED BY THE SOCIETY THE ZOOLOGICAL PARK, NEW YORK October 31, 1941 CONTENTS PAGE 19. Eastern Pacific Expeditions of the New York Zoological Society. XXVI. Crabs of the Genus Uca from the West Coast of Cen- tral America. By Jocelyn Crane. (Plates I-IX; Text-figures 1-8) 145 20. Eastern Pacific Expeditions of the New York Zoological Society. XXVII. A Study of Young Sailfish (Istiophorus). By William Beebe. (Plates I-V ; Text-figures 1-9) 209 21. On the Reproduction of Opsanus beta Goode & Bean. By Charles M. Breder, Jr. (Plates I & II) 229 22. On the Reproductive Behavior of the Sponge Blenny, Paraclinus marmoratus (Steindachner) . By Charles M. Breder, Jr. (Plates I— III ; Text-figure 1) 233 23. The Chromatophores of Fundulus heteroclitus in Polarized Light. By A. M. Shanes & Ross F. Nigrelli. (Plates I— III) 237 24. New Races of Alaudidae and Timaliidae from Northern Thailand. By H. G. Deignan 241 25. Respiratory Behavior in Fishes Not Especially Modified for Breathing Air Under Conditions of Depleted Oxygen. By Charles M. Breder, Jr. (Plate I) 243 26. Eastern Pacific Expeditions of the New York Zoological Society. XXVIII. Fishes from the Tropical Eastern Pacific. [From Cedros Island, Lower California, South to the Galapagos Islands and Northern Peru.] Part 3. Rays, Mantas and Chimaeras. By William Beebe & John Tee-Van. (Plates I-IV; Text- figures 1-40) 245 27. Erotylidae of Kartabo, Bartica District, British Guiana. (Cole- optera.) By C. H. Curran. (Plate I; Text-figure 1) 281 1941] Crane: Crabs of the Genus Uca 145 19. Eastern Pacific Expeditions of the New York Zoological Society. XXVI. Crabs of the Genus Uca from the West Coast of Central America.1 Jocelyn Crane Technical Associate, Department of Tropical Research, New York Zoological Society. (Plates I-IX; Text-figures 1-8). [This is the twenty-sixth of a series of papers dealing with the collections of the Eastern Pacific Expeditions of the New York Zoological Society made under the direction of Dr. William Beebe. The present paper is concerned with specimens taken on the Ardurus Oceanographic Expedition (1925), on the Eastern Pacific Zaca Expedition (1937-1938) and on a special trip made to the Pacific shores of Panama by the author in January and February, 1941. For data on localities and dates of the Ardurus and Zaca Expeditions, refer to Zoologica, Vol. 8, No. 1, pp. 1-32, and Vol. 23, No. 14, pp. 278-298.] Contents. I. Introduction 145 II. Summary of Important Points 146 III. Material and Methods 148 IV. Ecology 149 A. Habitat 149 B. Burrows 150 C. Feeding 150 D. Fighting 151 E. Crippled Crabs 151 F. Display, Coloration and Shelter- Building : Their Relation to Courtship . . 151 Display Activity 152 Coloration 154 Copulation 156 Shelter Building 157 Behavior of Females During Court- ship 157 Origin of Display and Comparisons . . 158 Conclusions 158 G. Breeding and Growth 159 H. Individuality and Play 159 I. Enemies and Defense 160 Y. Phylogeny 161 Coast of America and in the Galapagos Islands 167 1 Contribution No. 628, Department of Tropical Re- search, New York Zoological Society. VII. Species of Uca Taken by the Eastern Pa- cific Expeditions of the New York Zoological Society 170 U. princeps (Smith) 170 U. heteropleura (Smith) 171 U. stylifera (Milne-Edwards) 171 U. insignis (Milne-Edwards) 173 U. pygmaea sp. nov 174 U. zacae sp. nov 175 U. galapagensis Rathbun 176 U. mordax (Smith) 176 U. brevifrons (Stimpson) 177 17. macrodactyla (Milne-Edwards & Lucas) 178 17. tomentosa sp. nov 179 17. umbratila sp. nov 181 17. argillicola sp. nov 183 U. oerstedi Rathbun 184 U. inaequalis Rathbun 185 17. tenuipedis sp. nov 186 U. batuenta sp. nov 187 17. saltitanta sp. nov 189 U. beebei sp. nov. 192 U. stenodactyla (Milne-Edwards & Lucas) . 195 17. helleri Rathbun 198 U. crenulata (Lockington) 198 17. limicola sp. nov 198 U. deichmanni Rathbun 199 U. latimanus (Rathbun) . . 201 U. terpsichores sp. nov 202 17. panamensis (Stimpson) 204 VIII. Bibliography 205 I. Introduction. The present study is based on three collections : It is the second paper in a series dealing with the brachyuran crabs of the Eastern Pacific Zaca Expedition.2 In addition, an account is included of specimens which were taken by the author during January and February, 1941, on the Pacific coasts of Panama and the Canal Zone. Finally, a discussion is included of specimens referred by Boone (1927) to Uca galapagensis and U. helleri, since a reexamination shows that other species and fresh points of interest are involved. 2 Previously published: Crane, 1940. 146 Zoologica: New York Zoological Society [XXVI: 19 The collections comprise a total of 1,093 specimens, distributed among 27 species, of which 11 are apparently new to science. This wealth of material and the ecological observations which I was privileged to make in the field on both the Zaca and Panama trips, are such that considerable additions are made in this paper to our knowledge of habits, habitats and the phylo- genetic relationships of the species. In the following pages all references to courtship, court- ship coloration, copulation and shelter building are based on observations made at La Boca, Balboa, Canal Zone, at the mouth of the Canal, during the recent Panama trip, while reports on habitat, general habits and everyday color in life are from notes made during both the latter trip and the Zaca Expedition, on specimens observed and captured along the west coast from Mexico to Panama. Certain hitherto disregarded physical char- acters have been found to be of taxonomic value, and while Miss Rathbun’s invaluable monograph (1917) remains, as before, the foundation of any work on these species, the number of new forms in the present collection necessitated the making of a new key. I wish to express my appreciation for aid in the preparation of this study to the following people: To Dr. William Beebe, Director of the Department of Tropical Research, for sugges- tions and criticism, and for granting me leave of absence to make special observations on Uca in Panama during the winter of 1941; to Mr. Templeton Crocker for the opportunity of col- lecting material while on a cruise of his yacht Zaca; to Dr. Waldo L. Schmitt of the United States National Museum for the loan of material, and for laboratory facilities during study trips to Washington; to Dr. Roy Miner of the Ameri- can Museum of Natural History for the loan of material; to Dr. Herbert C. Clark of the Gorgas Memorial Laboratory, for laboratory facilities during my trip to Panama and to Mr. H. H. Evans for most valuable suggestions in regard to collecting grounds in Panama and the Canal Zone. The drawings in the present paper are the painstaking work of Mr. James Butler (Text- figures 2-5) and Miss Janet Wilson (Text- figures 6-8). II. Summary of Important Points. 1. Courtship: The waving of the large claw by male fiddler crabs is without question primarily concerned with the attraction of females, at least during the breeding season, and only secondarily with the warning-off of crabs trespassing on a male’s feeding range. The waving is only one manifestation, or step, in a definite courtship display or dance which varies so greatly with the species that individuals can be recognized at a distance by their characteristic motions. The other portions of the displays include raising of the body and stretching of the legs, various steps to one side or the other, revolutions, and special motions with the minor cheliped. All of these motions show off the areas of most brilliant color to the best advantage: for example, the anterior (ventral) side of the merus of the first two or three pairs of ambulatories are usually brilliantly colored in courting males, contrasting to the colored areas elsewhere on the body; yet these areas show only when the chelipeds are outstretched in display; the same is true of the merus of the major cheliped. The display of twelve species was observed. 2. Color Change: Courting adult males, in con- trast to other adult males, and, of course, to females and young, change color daily upon exposure to sunlight within the space of a short time — from a few minutes to an hour or more being required. The crab does not display vigorously and completely until full courtship color is assumed, the courting dress being always brighter than any other phase found in the same species. That courtship coloration and display play a definite part in sexual recognition and attraction is certain; that they play one also in sexual selection is likely, but has not yet been proved by experiment. Courting males lose color rapidly when cap- tured, resuming more or less completely the shades and patterns characteristic of the females and young, so that accurate color notes on breed- ing coloration can only be made when the crab is free and undisturbed. The use of binoculars in such study is almost essential. 3. Mating: Copulation was observed above ground, for the first time in natural surroundings, in three species. In each case it was preceded first by elaborate display of the male, and then by mutual stroking. The major cheliped played no part in actual copulation. It is thought that in most cases mating normally takes place below ground, but that when a female becomes acqui- escent at the mouth of her burrow, which is usually too small for a male to enter, he induces her to mate there. When possible, however, he leads her, by means of display, to follow him down his hole. The latter procedure was ob- served in seven species. Even at the height of the breeding season, males are successful in only a minute fraction of their courtships, and then only after prolonged and much interrupted displaying which sometimes is protracted over a period of days. All the courtships were observed in Pacific Panama in late January and through- out February, in the midst of the dry season; ovigerous females were seen in all courting species. Whether breeding continues at other times of the year is not yet known. 4. Shelter Building: Shelters made of pellets of sand, roughly semispherical in shape, were built by courting males of certain species on certain days above the mouths of their burrows. The function of these structures remains as yet unex- plained. 5. Phylogeny: A tentative phylogenetic tree of the species of Uca from the eastern Pacific is presented. Basic characters, including especially the mouthparts and minor chelipeds, have proved to be of more value in tracing relationships than 1941] Crane: Crabs of the Genus Uca 147 SAN LUCAS B. AfTENA BANK GORDA BANKS "t MAZATLAN Bi CLARION ISL. BANDERAS B. CHAMELA b\o. TENACATITA B.LY MANZANILLO-^ SIHUATANEJO ACAPULCO DULCE R..4 PORT ANGELES PORT GUATULCO SANTA CRUZ B. TANGOLA-TANGOLA B. FONSECA CORINTO^fx SAN JUAN DEL SUR^h-l'T * PORT PARKER -< F'°5TA\ GULF OF DULGE MU RC I E L AGO B. \ ?'V,CA/x POTRERO GRANDE B PORT CULEBRA BR AX I LI TO B. PIEDRA BLANCA B. GULF OF NICOYA EASTERN E X P E D PACIFIC T I O N S NEW YORX ZOOLOG ICAL- SOCIETY SHORE COLLECTING STATIONS GA L APAGOS IS. Text-figure 1. Shore collecting stations of the Eastern Pacific Expeditions of the New York Zoological Society. the characters usually employed to differentiate species, such as the form of the major cheliped, width of the front and convexity of the carapace. Related species show fundamental similarities in courtship displays as well as in physical attri- butes. Specialization has taken place in a number of directions, but especially toward provision for withstanding dryness, in connection with the adoption of a truly littoral existence. Species living, or at least courting, on shores which are daily exposed by the tide for relatively long periods in general have the most highly developed courtships and the brightest colors, including dazzling white, in contrast to forms living exclu- sively on briefly exposed mud flats. 6. New Species: Of the 27 species in the present collection, 11 appear to be hitherto undescribed. This raises the total of apparently valid species known from the eastern Pacific from 22 to 33. 7. Taxonomic Notes: Miss Rathbun’s synony- my (1917) has been followed throughout. The following forms, described or recorded since that date, should apparently be synonymized as indicated: U. galapagensis, Boone, 1927 (not U. galapagensis Rathbun) part. = U. macrodactyla (Milne-Ed- wards & Lucas) ; part. = U. panamensis (Stimp- son) ; part, correctly identified. U. brevifrons var. delicata Maccagno, 1929. = U. brevifrons (Stimpson) . U. quayaquilensis Rathbun 1935. = U. festae Nobili 1902. 148 Zoologica: New York Zoological Society III. Material and Methods. The data in the present paper are the result of two separate methods of working. The first, con- sisting of intensive collecting and of habitat observations from southern Mexico to Panama, was undertaken chiefly on the Eastern Pacific Zaca Expedition, from December, 1937, through March, 1938. Because of the nature of the trip, during which a maximum of a week or ten days was spent in each bay, and all groups of crabs were studied and collected without special refer- ence to Uca, it was impossible to make prolonged and detailed observations of fiddler crab habits which by their nature require long daily periods of observation, preferably in the same locality. Therefore, during the five weeks spent in Panama during January and February of 1941, attention was directed primarily toward filling in the gaps in the Zaca habit notes, and only secondarily to collecting. Observation in Panama was largely confined to one small area, a slightly brackish cove at La Boca, Balboa, Canal Zone, at the very mouth of the Canal. This locality proved to be so rich and was so accessible that it seemed wise to concen- trate study here. However, colors and courtship displays were checked in other nearby places, including the mud flats at Bellavista, Panama City, and at Old Panama, a few miles south along the coast. It was found that when the same area was visited day after day, and hours spent in observa- tion, the activities of certain individual crabs could easily be watched for as much as four weeks at a time, individuals being differentiated by means of damaged or regenerated claws and legs, or scars on the carapace. Since many of the crabs at La Boca were, most fortunately, in the midst of courting at this time, and since my stay was definitely limited, it seemed best to make all observations on these crabs while they were in a completely natural condition, although there was a temptation to divide my time between that type of study and experiments in the field of sexual selection, to try to determine the actual extent, if any, of the effect of the male’s color and display in his attraction of a mate. However, since controlled experiments to be at all significant would have required a great deal of time, it seemed better to reserve them for a later study, and to limit observation to natural reactions, particularly since tins field is so little explored. On both the Zaca Expedition and on the more recent trip, a pair of Number 7 Zeiss binoculars was constantly used. Although it is true that most species will overcome their fear after a few minutes and resume their activities more or less normally if an observer remains perfectly quiet close by, still the slightest motion sends them into their burrows again. On the other hand, after crabs watched through a binoculars at a distance of 12 or 15 feet have once emerged and lost their fear, it is possible to move slowly, make notes, and switch observation from one individual to [XXVI: 19 another without alarming any over a period of hours. In photographing, a Leica camera was used with a 90 mm. telephoto lens at a distance of about 20 inches. The camera was attached to a focussing device, which in turn was screwed to a tilt-top tripod head. The latter was attached to a six-inch news-camera hand-holder. The handle was thrust through a square of cardboard as far as the tripod head, in order to protect the camera from sand, and then pushed into the muddy sand up to the cardboard. A pencil stuck into the ground beside the desired crab hole was used as a temporary focussing point. After the crab emerged it was often possible, if the first exposure was not made for several minutes, to release the shutter and wind the film for another picture without alarming the crab, pro- vided that the movements of the fingers were exceedingly slow, and hidden as much as possible by the camera. In bright sunshine, on the dark gray muddy sand of La Boca, an exposure of 1/60 of a second at F16 on Eastman Super-XX film was found to give the most satisfactory compromise between speed and depth, although even this combination gave, of course, very little depth and yet would not stop the motion of the large claw when on a downward or upward swing. A series of these photographs was found to be valuable in showing the relative position of the various appendages at a given point in the display in the various species, and proved a most useful addition and check to repeated visual observations. Methods of measurements and special terms used in the following pages should be defined thus: 1. Length. Measured from most anterior exten- sion of front to posterior margin of carapace, in the longitudinal median line. 2. Length of palm . In either major or minor cheliped, measured from its most proximal origin on external side to the gape between the chelae, midway between lower base of dactyl and upper base of pollex. 3. Base of palm to tip of pollex. Measured from most proximal point of base of palm on external side to tip of pollex. 4. Length of dactyl. Measured from its origin on dorsal profile to its most distal projection. 5. Width of front. Measured between posterior margins of bases of eyestalks, as they appear when the eyes are lying flat in their sockets. 6. Major side. The side giving rise to the large or major cheliped. 7. Minor side. The side giving rise to the small or minor cheliped. 8. Eyebrow. As defined by Miss Rathbun (1917, p. 375) ; the intervening space between the two margins of the upper part of the orbit ; it is usually broad, more or less inclined, roughly triangular. 9. Suborbital region. The area between the 1941] 149 Crane: Crabs of the Genus Uca lower border of the orbit and the pterygostomian region. 10. Grooving of ischium of third maxillipeds. This character is best examined with the append- age held out of liquid, and turned back and forth under a good lens in a strong, oblique light. 11. Number of spoon-tipped hairs on second maxilliped. Throughout the following study, where the number of spoon-tipped hairs is re- ferred to, the count given is of those on the merus of the second maxilliped only; that is, those which occur on the tip of the palp are excluded. In those forms with most of the internal (upper) surface of the merus lined with these hairs, only those individual hairs are counted which project beyond the inner margin; the number of rows of hairs given, however, is always the total number. In all counts, hairs with rudimentary spooned tips are disregarded. 12. Display. The unit of behavior consisting of a single wave or gesture with the major cheliped along with its associated activities (stretching, running, “dancing,” rapping, etc.). A number of displays following quickly upon one another, with little or no pause between, is termed a “series.” IV. Ecology. A. Habitat. In general, the habitat of individual species is much restricted; some forms are found only in open sun on brackish salt mudflats, others in the deep shade among mangrove roots, still others on open muddy sand beaches, and so on. A given type of terrain may be expected to yield certain species; a similar but slightly different type only a few yards away, supports other forms. Some species, however, may inhabit more than one kind of environment. The great majority live in mud, either on protected tidal flats close to the mouths of streams, so that the water is somewhat brackish, or on the banks of sluggish streams of brackish water. Probably species originally migrated from this relatively stable, quiet en- vironment to fresh water streams on the one hand and to open marine beaches on the other. The habitats of species in the present collection are as follows : 1. Muddy banks of fresh water streams (mostly shaded). pygmaea zacae brevifrons mordax latimanus 2. Muddy banks of brackish streams (mostly shaded). mordax brevifrons limicola latimanus 3. White clay banks of brackish streams (mostly shaded). argillicola 4. Mud, among mangroves (completely or partly shaded). zacae mordax brevifrons tomentosa umbratila inaequalis tenuipedis batuenta crenulata 5. Mud, among unshaded mangrove shoots. insignis oerstedi batuenta 6. Tidal mud flats (unshaded). princeps heteropleura stylifera insignis macrodactyla oerstedi inaequalis (rarely) saltitanta beebei (near beach) galapagensis (?) heller i (?) 7. Protected muddy-sand flats or beaches (unshaded). princeps heteropleura stylifera insignis beebei stenodactyla deichmanni latimanus terpsichores 8. Marine sandy beaches, among stones. panamensis There is evidence (see page 1 7 Off. ) for the belief that U. princeps, heteropleura and stylifera spend most of their time on mud-flats, but come to adjacent sandy-mud beaches to court, since the added firmness of the ground would probably be a distinct aid to display. La Boca: As an example of a particularly and surprisingly rich locality for the study of fiddler crabs, a small cove on the left bank of the Pacific mouth of the Canal is unsurpassed. This cove, lying between the Balboa docks and the La Boca ferry, and surrounded by piers, motor roads and the continuous traffic of the Canal, yielded 15 species of Uca in January and Febru- ary, 1941, in an area not more than 600 feet square. The total number of individuals in this space must have run into the hundreds of thousands. Of these species, 12 were actively courting. The cove is entirely empty of water at low tide, exposing a mud fiat and, bounding it on two sides, a narrow beach of muddy sand. At the inner end, the muddy sand is mixed with gravel, and adjoins a small clump of mangroves. 150 [XXVI: 19 Zoologica: New York Zoological Society The third side, parallel to the Canal, is bounded by Pier No. 4, and the fourth side, facing the canal mouth and the ferry, is open. The water is slightly brackish, because of the overflow from the canal locks. The exceptional richness is probably due chiefly to the fact that a large sewer opens practically into the cove, and that garbage from ships is frequently washed up. The dissolved organic detritus from these sources must greatly enrich the tidal deposits on the mud. The fact that these 15 species taken included 5 hitherto undescribed (although all of them had been taken also on the Zaca Expedition, farther up the coast) shows how many valuable studies are waiting to be made in the tropics, even in the most accessible, thickly populated and apparent- ly unlikely areas. The extent to which more than one species occupies a given stretch of ground seems to depend entirely on the available space and the number of individuals it can support, provided of course that the terrain is equally suitable for the various species. At La Boca, for example, in one part of the beach stenodactyla was dominant, but mingled with numerous beebei and rare stylifera; farther down toward the mud, where the ground stayed more moist, were found deichmanni, hetero-pleura, and, dominantly, beebei. In the most protected part of the cove, most subject to dryness, behind the mangroves, latimanus was dominant, with only very few terpsichores intermingled; in the gravelly sand area, beebei was dominant, mingled with uncom- mon terpsichores. In less rich areas, the species tend to keep separate, there being definite boundaries, for example at Corinto, Nicaragua, between a large colony of stenodactyla and an equally large one of stylifera close by, on ap- parently identical terrain. In crowded colonies, with the feeding ranges of individuals much restricted, there are of course many more provocations for fighting than in uncrowded areas, but it also seems that far more toleration has been developed in these individu- als. Sex obviously is an important if not vital factor in argument. Adult males of different species and the same or different sizes will tolerate each other’s burrows exceedingly close together: beebei and stenodactyla have been seen living day after day only one and a half inches apart, whereas adult males of the same species will tolerate each other’s burrows not less than three inches apart and generally more. Females and young crowd closely, likewise, with practi- cally no argument. B. Burrows. No special studies were made on the subject of burrow digging, except to determine average depths and forms for the various species. As was to be expected, species on yielding mud had the shallowest holes, and large species high up on relatively dry muddy sand beaches, or along the banks of drying streams, dug deepest. The burrows of individuals of the same species varied considerably, depending on their location and, of course, on the size of the crab. The instinct of burrow-making seems least well developed in panamensis, which lives on the stone-strewn ends of sandy beaches. Very young crabs of all species do not dig, but run freely in and out of the burrows of adults, both of their own and different species, which pay no attention to the small ones whatever. For detailed accounts of burrow digging, consult especially Pearse (1912), Dem- bowski (1926) and Verwey (1930). An interesting point is that stylifera and beebei, at least, often occupy their burrows for days and even weeks, without changing the location of the mouth by so much as half an inch. Contrary to Dembowski’s observations on pugilator in cap- tivity, in all the species observed the crab emerged after high tide, leaving the entire length of the burrow free, and frequently did not repair or change it at all from one day to the next. The crab must simply push its way through the sand in emerging, instead of digging himself out in such a way that a hole only the size of himself is left at the top, as seen by Dembowski. Burrows are usually, but not always, plugged up before the tide covers them. Females and young, especially of latimanus, which lives at La Boca high on the shore, often plug their holes with a dome of pellets brought up from below for two or three hours around noon on especially hot, bright days; later, if the tide is still far out, they reemerge and resume feeding. C. Feeding. The general process of scooping up mud or sand with the spooned minor chelipeds, carrying it to the mouth, separating the organic particles by poorly-understood actions of the mouthparts and passing the remaining detritus out at the posterior end of the buccal cavity in the form of a pellet, has been too well stated previously to need repetition. The observations of Pearse (1912), Monod, and Verwey are the most detailed. The method by which the actual sifting is done remains as mysterious as ever. Every species observed on the west coast definitely wiped or clipped off the pellet which formed at the posterior part of the buccal cavity after between 6 and 16 cheliped-fuls of sand or mud had been conveyed to the mouth, and carefully placed the pellet in front or to one side of the crab; the only exceptions were a few indi- viduals observed starting to feed while there was still some water on a mudflat as the tide was going out; in these cases the water dissolved and carried away the rejected mud. Matthew’s statement that U. leptodactyla on the coast of Brazil picked up separate organic particles from the ground is most interesting; it seems likely that there was an error made in observation ; the same is almost certainly true, as Verwey has contended, of Symon’s statement that the re- jected mud passed out at the front of the buccal cavity. Altogether, I have watched at least 20 species of Uca actually feeding, and in all the process is identical, save for three exceptions. The first, that of Uca’s feeding with the mud 1941] Crane: Crabs of the Genus Uca 151 slightly underwater, has been noted; the second was that of four individuals of brevifrons seen feeding on mammalian excrement on the bank of a fresh-water stream, several miles from the coast, at Port Parker, Costa Rica; the third is the practice of panamensis of frequently climbing upon rocks and scraping off the algae for food with its small cheliped, just as do Pachygrapsus transversus, Grapsus grapsus, and others. Both brevifrons and panamensis, however, also feed in the usual fashion of fiddler crabs. Although there are doubtless exceptions, like the brevifrons mentioned above, and those noted by Pearse (1912), fiddlers as a general rule are certainly not scavengers: a number of times I have seen garbage, dead fish, or a dead bird in all stages of freshness and decomposition washed up in a colony of fiddlers and lying untouched in their midst, although normally-nocturnal hermit crabs were swarming over it. The roles played by the various types of minor chelae and the various numbers of spooned and wooly hairs on the second maxillipeds are un- known; they will be further discussed on pages 161 to 165. D. Fighting. The conclusions of most other observers were verified that, although brief duels, in which the large chelae are interlocked, are frequent, injury is exceedingly rare. I have only once seen any dismemberment actually take place, and this was the extreme tip of a dactyl. However, the frequent sight of crabs with similarly or further damaged chelae indicates that such mishaps do occur now and then, although, of course, it is possible that they can also occur in other ways, such as by being broken on a stone against the side of the burrow, or in escaping from an enemy. A definite duelling ritual was followed in all except cases of extreme provocation, such as pursuit of the same female in U. stenodactyla, in which there was no time for preliminaries. Usually, however, a duel proceeded as follows: Only infrequently was a duel preceded by display, — i. e., by rhythmic series of beckonings — and then never by vigorous display such as was used in courting a female. Instead, the combatants first prance toward each other, stiff-legged, pat- ting the ground with the bottom of the great palms and claws, as though in challenge. Then both lunge and feint a few times, with the back and sides of the semi-flexed chelipeds, which meet in audible clicks. At last, after several or half a dozen such parries, they proceed to the last step and interlock claws. This last movement is always undertaken warily, since there is always danger that the nipper may be damaged or wrenched off. Then, the claws locked, the crabs lunge in turn, pushing each other back and forth, first one and then the other sinking down and back until his shell often actually touches the ground. Usually this continues until the weaker breaks away and runs for his hole; occasionally he is somersaulted backward with a flip of his opponent’s claw; often the larger simply stops fighting and moves off without there being any decision. Most duels last only a few seconds; the record was 25 minutes (see p. 160). Sometimes, as has been observed by others, the weaker is pushed down his own hole, where he uses his large claw to stop up the opening, for a few minutes, apparently to prevent the entry of his antagonist. The provocation for a fight is usually either poaching on the territory of another male, or courting of the same female. Sometimes, how- ever, there is no stimulus apparent for the most spirited encounters ; until some better explanation is found, these must be laid to sheer excess energy, probably associated with the condition of the glands in the breeding season, or may simply be termed sport (see p. 159). E. Crippled Crabs. Several instances have been observed where crabs with badly damaged chelae, with the cheliped completely missing, or with more than two ambulatories missing, were definitely bullied by other crabs. In each of these cases the cripple was repeatedly driven down his hole by one of his neighbors. Once a normal male U. beebei kept watch for at least several hours over an injured neighbor, the burrow of which was a foot away, well beyond the feeding ground of the normal crab. Every time the cripple emerged the neighbor would stop his own feeding or courting, run over and struggle until the cripple retired to his hole, after which the normal crab would push sand in after him and stamp it down until no trace of the burrow was left. Each time the cripple emerged the routine was repeated. No attempt appeared to be made actually to harm the crab or to follow him down his hole. Many crabs with claws in various stages of regeneration, however, seemed to carry on com- pletely normal lives. Twice I have seen crabs with the minor cheliped missing which were eating awkwardly, but with apparent success, with the major cheliped. I have not yet observed a male without the major cheliped temporarily mistaken for a female as described by Verwey. For a case of a male courting with a half-regen- erated claw, see p. 155. F. Display, Coloration and Shelter-Building: Their Relation to Courtship. For more than a hundred years the sexual dimorphism of fiddler crabs, the function of the waving of the large claw in the air, the frequently brilliant color of the large claw and the possible significance of all these factors in regard to sexual selection, have become matters of increasing interest and controversy to students both of Crustacea and of general evolution. The opinions of the principal workers in the field may be briefly summarized as follows: The earliest — Muller (1869, 1881), Darwin (1871) and Alcock (1S92, 1902) — were convinced that both waving and bright colors played a definite part in court- ship, that females recognized males in this fashion 152 Zoologica: New York Zoological Society [XXVI: 19 and appreciated the display, which gave to color and activity a vital role in sexual selection. Pearse (1912, 1914.1, 1914.2) accepted the con- clusion that the males danced around the females in order to attract their attention, but was not convinced that the females were thus attracted, or that the colors had anything to do with sexual selection. Symons (1920), Johnson & Snook (1927), Beebe (1928) and Matthews (1930) added brief observations on various species which showed that males were sometimes definitely stimulated by females to increased waving activity and special behavior. And, finally, Verwey (1930) and Hediger (1933, 1934) denied that the waving had anything whatever to do with courtship, declaring that it was carried on instead solely to designate the possession of a hole, and limitation of the surrounding feeding area. After concentrated observation in Panama during the breeding season of a dozen species living in a single restricted area, I am convinced that the truth lies somewhere near the middle of these extremes of conflicting opinions. Waving, as in other crabs, is certainly carried on some of the time as a warning to other males and to de- limit territory in some (but not in all) species of Uca. On the other hand, in many, if not in all species, waving definitely plays a large part in courtship and, in one species at least ( latimanus ), is apparently carried on only by courting males. In regard to the importance of color in waving, I cannot draw definite conclusions until further studies have been carried out. These must in- clude experiments in artificial coloration as well as many more observations on the natural be- havior of the crabs. However, it seems certain that color and motion are correlated in display, although color is perhaps the less important of the two aspects. This subject will be further discussed later (pp. 154 to 159). Display Activity: A comparison of the observa- tions of the authorities listed above shows that among different species there are certain varia- tions in the activities accompanying waving. Studies on Philippine fiddlers by Pearse (1912) show that in one or more species the males dance around the female, showing only their backs to her, so that the great chelae could not be seen by the females, which usually paid no attention to the males in any case. Sometimes the males held statuesque poses, the chelipeds upraised or outspread for minutes at a time; the females often were, surprisingly, more brightly colored than males of their own species. Pearse states, however, that his observations were not made during the breeding season, and season should doubtless be held a most important factor in all studies of display and coloration among these crabs. Alcock, reporting on annulipes in Ceylon, Symons on an unidentified Uca on the same island, Johnson & Snook on crenulata in Cali- fornia and Matthews on leptodactyla in Brazil, all reported that the males of these particular species were galvanized to more active waving by the appearance or presence of females. Johnson & Snook stated definitely that this reaction took place during the breeding season, and that crabs did not wave in captivity, al- though they carried on all other normal daily activities. Beebe, observing mordax in Haiti, found that certain females stimulated the males to intensified waving, which consisted of “a beckoning in five jerks, the last of which almost threw the crab over on its back; the difference between this gesture of the right hand of passionate fellowship and that of shaking the fist in the face of any passing male was hardly to be discerned. In the case of courtship the fiddler would often freeze into a statuesque pose for three or four minutes at a time.” Swartze & Safir observing pugilator in Massa- chusetts, and Hediger studying tangeri in Moroc- co, agreed that the crabs waved at the rate of once every two seconds; pugilator varied simple waving with statuesque poses according to Pearse (1914.1) such as those in mordax and some of the Philippine crabs (Pearse, 1912). Hediger mentions no posing in tangeri, however, but refers to the elevation of the whole body which sometimes accompanies beckoning. Verwey, in his study of Uca in Java, reported that signata in waving, stretched high up and sank back, sometimes shaking all over. This species, according to the same author, when captured as well as when in the field, waved threateningly at both males and females, though apparently not so forcefully at the latter sex. Although Verwey witnessed no copulations, he saw males approach and cover females from behind, giving the females no opportunity to recognize, much less be attracted, by the male in question. He was apparently present during the breeding season, since he has a number of records concerning the abundance of ovigerous females on various dates. The observations summarized above show that waving, whether or not the observer admits its being used in courtship, is often accompanied or varied by the following activities: (1) by statu- esque posing with the arm outstretched or up- raised, (2) by shaking of the entire body, (3) by elevation of the body with each beckoning gesture, (4) by “dancing” around the females, and (5) by increase of tempo in the presence of females. My own observations show that the last characteristic of accelerated waving is the only one which is characteristic of most species in display. In Panama, each of the species studied combined with waving variations of one or more of these five and other activities. In fact, each species proved to have a definite, individual display, differing so markedly from that of every other species observed, that closely related species could be recognized at a distance merely by the form of the display. Furthermore, related species had funda- mental similarities of display in common, and series of species, showing progressive specialization 1941] Crane: Crabs of the Genus Uca 153 of structure, in general showed similar progression in display. At La Boca, Balboa, Canal Zone, the court- ships of twelve species were studied in consider- able detail. These twelve are divided into three groups, each composed of more or less closely related species. The first (“Group 1,” p. 165) consists of princeps , heteropleura and stylifera; the second (“Group 4,” p. 166) of oerstedi, inaequalis, batuenta and saltitanta; the third (“Group 5,” p. 166) of beebei, stenodactyla, deichmanni, latimanus and terpsichores. In this last group the first two and the last three form definite subgroups. The display of a thirteenth species, panamensis, was seen only once, and on this occasion the crabs were apparently not displaying fully; hence it is omitted from this summary, but described on p. 204. The basic element in all fiddler display is, of course, “waving” or “beckoning” with the major cheliped. In its simplest form this con- sists of unflexing the manus and chelae from the resting position in front of the mouth, by move- ments of the ischium, merus, and, perhaps, carpus, which elevate the distal elements diagonally over the head; they are usually lowered at once, without a pause, often with a jerk, in the same plane to the original position. The variations of this simple beckoning gesture and its accompaniments in the twelve species listed above are as follows: 1. Beckoning or Waving: a. Three of the four crabs of the second group hold in common a special sort of cheliped motion: After the usual beckoning, the cheliped is brought to the ground, flexed, more or less in front of its usual rest position, and is then bounced back into place with three or four raps of the ground; this additional action is added to the regular display occasionally in inaequalis, often in batuenta and always in saltitanta. In the latter species the cheliped is bounced vigorously almost in place, since only rarely does it fall in front of its normal rest position. In physical characteristics also these three species show progressively greater specialization. b. U. deichmanni holds the cheliped for an instant at the highest point of its reach, then lowers it into position and raises it again without a pause, so that the accent comes at the peak of its stretch, instead of in the flexed rest position. c. U. latimanus makes a somewhat circular gesture in beckoning. To a lesser degree this is occasionally true of other species. d. U. terpsichores starts display with the manus and dactyl of the large cheliped half unflexed, pointing straight out in front of the crab. e. The rate of display when the crab is not specially excited varies among species from two or more gestures to the second with no pause be- tween them ( saltitanta , beebei, deichmanni and terpsichores) to one and a half to three seconds being required for every display, including both the gesture and the pause following it ( princeps , oerstedi, inaequalis). Both slow and fast displays occur in all groups, without relation to the degree of specialization. Usually a more or less definite number of dis- plays is made in a series when a given species is displaying fully, each series being followed by a rest of seconds or minutes. The most tireless so far observed is saltitanta, which frequently dis- plays, at the rate of two gestures to the second, for upwards of one hundred at a time without an instant’s pause. After a rest of several seconds another equally long series will be commenced, and the procedure may continue, without inter- ruption for true rest or feeding, for at least an hour. On the other hand, in every species display is often casual, half-hearted and punctuated by feeding. 2. Elevation of Body during Beckoning: As in tangeri, the body is elevated and depressed during each beckoning by princeps, heteropleura, styli- fera (sometimes), inaequalis, batuenta, saltitanta and deichmanni. The other five species hold it consistently high during a series of displays. In elevation, the crab stretches to tip-toe with the raising of the cheliped, and sinks into position when it is lowered. U. heteropleura carries this habit to the greatest extreme, stretching so high that only the tips of the two middle pairs of ambulatories remain on the ground. U. stylifera sometimes stretches the two front ambulatories, elevating the anterior part of the carapace, and simultaneously flexing the posterior legs, so that the rear part of the body is lowered during a display; this variation does not depend on the added stimulus of a female, but may be inserted in the midst of a series of displays, or an entire inter-tidal period may be devoted to this variety. Species which elevate and depress the carapace during each display occur in all three groups. 3. Position of Chelae: The chelae of both major and minor chelipeds are usually held slightly open and parallel during a series of displays; some- times, however (especially in batuenta and salti- tanta) they are opened and closed with every display — opening wide on the upward swing of the large cheliped and closing on its descent. 4. Motion of Minor Cheliped: Often the minor cheliped is moved in a feeble imitation of the beckoning or outward spreading of that of the major. This is generally true of princeps, hetero- pleura, and terpsichores, and true under condi- tions of excitement in other species. 5. Some species remain in one spot during display, notably heteropleura, which stretches up on its two middle pairs of ambulatories. More, however, usually take several steps to one side or the other with each elevation of the cheliped; in a given series of displays, they usually move several times in the same direction, then back to the starting point beside the hole; or with one gesture they may move to the right, with the next to the left. There is considerable leeway of behavior within a species in this activity, and individuals show similar variation in their be- havior from hour to hour or day to day. U. oerstedi tends to move right around its hole, facing always outward, when the attention of a 154 Zoologica: New York Zoological Society [XXVI: 19 female is not involved, during a series of displays. In the highly specialized courtship of stenodactyla, the males chase the females with the arm out- stretched, motionless, or held overhead. If they can manage to approach a female closely enough, they surround her loosely with the cheiiped, and race with her over the beach, apparently trying to maneuver her to their holes (for details see p. 196). 6. Revolution: The onty instances I saw of a crab dancing with his back to a female were all performed by individuals of a single species, beebei. In this crab the anterior part of the carapace is a brilliant iridescent green, the cheii- ped chiefly rose, ochre and plum-colored, the anterior parts of the ambulatory meri magenta. Males of this species in a number of cases almost always revolved before females wdiose attention they had captured. The color of the carapace in this relatively dull form was certainly to human eyes at least as striking as that of the cheiiped, and the combination most effective. 7. Special Effects Reserved for Later Stages of Courtship: The usual acceleration in tempo of display during active courtship of interested or potentially interested females has already been noted. In addition, special actions, highly characteristic specifically, may be observed in the later stages of courtship. Probably many more will be added to the following list when the complete courtship of each species is known. a. U. oerstedi, when actively courting a female, vibrates the brilliant blue anterior ambulatories when the chelipeds are outstretched. b. U. saltitanta, when convinced that a female is ready to follow him down his burrow, pauses halfway down its mouth, extends the ambula- tories of the major side rigidly in the air and vibrates them rapidly. c. U. latimanus vibrates the minor cheiiped rapidly when displaying before an interested female. U. stenodactyla, when pursuing a female, extends this appendage stiffly outward, the chelae wide open, corresponding roughly to the gesture made with the large cheiiped at this time. d. U. terpsichores sometimes adopts the statu- esque pose previously noted in Philippine species (Pearse, 1912), in pugilator (Pearse, 1914.1) and mordax (Beebe, 1928), standing before a female for seconds at a time with the major cheiiped spread out rigidly sideways; this pose in the present species is never held longer than a minute. U. stenodactyla extends the major cheiiped simi- larly, either sideways or upward, when chasing females. e. Special steps with the ambulatories, giving the effect of a definite dance, are taken in the last stages of courtship by two quite unrelated species, stylifera and terpsichores. f. Mutual stroking of legs and carapaces with the ambulatories by both male and female took place in the final part of courtship in all three species in which copulation was observed, namely stylifera, beebei and stenodactyla. In each case the male instigated the stroking and was by far the more active partner. 8. Use of Special Display Ground: All species, when courting, in general keep the ground sur- rounding their holes well packed down and free of feeding pellets, but saltitanta, which lives and displays on the stickiest, dampest mud flats, whenever possible mounts to the summit of a nearby elevation, sometimes a dozen times his own height, in order to display. 9. Although particular attention was given, no hint was seen of the use of antennae, ocular stylets (in heteropleura and stylifera) or of stridu- lating ridges (in terpsichores and allies) in display. Coloration: As has been repeatedly noticed, the adult males in the genus Uca are frequently brilliantly colored, especially in regard to their large cheiiped. The females and young on the other hand are relatively dull, being usually brown or gray, often spotted or mottled with darker or lighter. Pearse (1912) alone has re- marked that “at Manila the female fiddlers often were, to the human eye, more brightly colored than males of their own species, and the female’s bright colors were on her back and legs so that they could readily be seen by a male dancing behind her, but she did no dancing.” As the author remarked, his observations were not made during the breeding season. Muller, observing a species of Uca in Brazil, was apparently the first to notice color change in the field, of which no further accounts seem to have appeared until the present study.3 His summary (1881, p. 472) is as follows: “When it (the fiddler) runs from its moist burrow into the sunlight the entire splendor of its nuptial clothes develops; as soon as one catches it, the pure white, the light green, which decorate its claws, begin to lose their luster and change in a few minutes into uniform gray.” The present author, unfamiliar with the above reference when making observations in Panama, found exactly similar 3 The interesting observations made on Atlantic fiddler crabs from which the eyestalks have been removed are not directly relevant to the present study, since all were made under extremely unnatural laboratory con- ditions. The most recent summary (1940) of Abramo- witz & Abramowitz indicates that eyestalk removal brings about loss of pigment, accelerates moulting, increases the death-rate in moulting and results in gigantism. The authors also insert the following remarks (p. 187) without further elaboration, regarding the breeding of U. pugilator in captivity: “ Uca breeds during September, as indicated by the appearance of large masses of eggs, copulation and final shedding of the eggs. Animals without eyestalks have been observed to copu- late and shed their eggs. However, such am'mals were blinded for only a few weeks before the onset of the breeding season, and thus sufficient time may not have elapsed for any effect on reproduction to take place. This is worth investigation, however, for as yet no endo- crine influence on the reproductive system of crustaceans has been demonstrated.” Brown concluded (1940) that the source of the chromatophorotrophic substance of the crustacean eyestalk, including that of Uca, is the sinus gland. Kleinholz '& Bourquin (1941), however, state that not all of the conclusions of the above-men- tioned investigators are yet proved, due to various details of the technique employed in the experiments, and of laboratory conditions involved. 1941] Crane: Crabs of the Genus Uca 155 conditions in all of the crabs which were known to be courting. Two of the most striking illustra- tions will be summarized here and are typical of the rest. Only adult males and females of stylifera were found on the relatively firm muddy sand shore above the mud flats at La Boca. They occurred there in increasing numbers throughout Febru- ary, and it seemed likely that they migrated from the soft, damp flats to court on the dryer ground. ( U . saltitanta and others, however, manage to keep clean and carry on a strenuous display in the midst of the mud-flats.) Courting males of stylifera, after emergence in the morning, changed from dull gray with the major cheliped chiefly brownish or yellowish, to pure white with the major cheliped orange, yellow, pink and white and the ambulatories bright purple. There was a brief intermediate phase where the carapace was brilliant yellow instead of white. It appeared that for the first one to three days an individual male was on the beach, and before he was dis- playing fully, the carapace brightened daily only to this yellow phase. A male stylifera which coaxed a female, after prolonged courtship, to follow him down his hole, promptly enlarged it, then stopped it up with both of them inside, and remained with her there until the following low tide. The next day the male was in poor coloration and did not display. The female had vanished. The courtship of a male of this species with the major cheliped in process of regeneration and only half size was observed throughout three weeks. Although otherwise normally colored — with dazzling white carapace and the legs purple — the regenerated member changed daily from drab gray brown to perfectly white, like the carapace, instead of yellow, orange and pink. This individual courted two females impartially for days until one moved away; he then paid all his attention to the other; I never saw copulation take place, nor saw either female attracted to his hole, but both often allowed him to approach and stroke them at the edges of their holes. I never, however, saw them stroke him in return, as in a consummated courtship between a normal male and female. For further details concerning this species, see p. 171. The other most striking example of color change occurred in latimanus. The females, young and non-courting males, no matter how large, as well as males newly emerged from their holes after high tide, were similarly colored, the carapace being brown with gold spots and the ambulatories brown banded with dark; in the case of the males the large cheliped was dull chestnut brown. In these crabs it was a stricter rule than in any other species that only displaying males changed color daily, and conversely that all males which changed color displayed; finally, all of these and only these, built shelters (see p. 157). In the displaying males, the carapace each day, after feeding, during shelter building, and before display started, became pure white, the major cheliped chiefly bright orange, and the anterior sides of the ambulatory meri plum red. In this species the display-color-change-build- ing cycle was definitely under tidal influence. When first observed, on the last three days of January and the first two of February, during the spring tides of new moon, practically all the males high on the beach, with holes covered with water only during the highest tides, were dis- playing, while none of those lower down were do- ing anything save feeding and repairing their holes. Among this latter group not so much as a single waving movement was seen, although full grown males were as numerous as near the high water line. Similarly, in contrast to the upper group, males and females paid no attention to each other and I did not see a single fight go be- yond the stage where the owner of a hole moved menacingly toward a trespasser, his cheliped thrust forward, without waving; in each of these few cases the threatened crab moved off promptly. In the following days of neap tides, when the water did not reach the upper part of the beach, the members of the upper colony remained in their burrows, six inches to a foot underground, for eight days (February 3-11). When dug up they appeared to be in a partial coma, and did not try to escape for half an hour or more. During the succeeding spring tides of full moon these upper colony crabs reemerged, repaired their burrows and fed, but only a single individual on a single day displayed, changed color and built a shelter during the entire period. They again remained in their burrows during neap tide, reemerging during new moon around Feb- ruary 25. Again there was no display. On the lower part of the beach, at this new moon period, however, a wave of display, color change, shelter building, mutual interest between the sexes and duelling, swept over the colony, ex- actly as it had four weeks previously higher on the beach. Observation unfortunately had to stop while this second courting wave was at its height. No actual copulations were seen, but several times females were observed to follow males down their holes, after watching long dis- plays, and many more abortive courtships were observed, in which the female lost interest and moved away. (See also p. 157.) In both these species, stylifera and latimanus, as well as in the others which changed less spectacularly, the display coloration was swiftly lost when the crabs were captured. The time required daily for the assumption of courtship coloration varied in different individuals and, to a lesser extent, in the same individuals on differ- ent days. One of the swiftest to change after emergence in the morning was stylifera which, on brilliant days, became fully colored within 15 minutes or less ; one of the slowest was latimanus, which sometimes required two hours or more. The slowness is probably explained by the fact that latimanus always built a shelter before dis- playing, which necessitated its coming in contact with the damp sand and, apparently, dryness and sunlight are almost equally important in effecting color change; however, no experiments have as yet been conducted on this subject. Females, young, and non-courting males also brighten somewhat after emergence, but the / 156 Zoologica: New York Zoological Society brightening consists only of emphasis of the prevailing drab colors, and is probably due as much to drying off as to exposure to light. One of the most interesting aspects of display is the fact that ten of the twelve species studied show brilliant colors on the anterior side of the merus of the first three pairs of ambulatories, areas which can be seen only when the chelipeds are widespread in display. Another point has already been mentioned, namely, that in beebei, a relatively dull-colored form, almost the brightest portion is its iridescent green carapace: this species alone was observed to revolve in front of the female before which it was displaying; it is noteworthy that the ambulatory meri were as brilliant posteriorly as anteriorly. The display coloration of the males of various species as far as known may be simplified and summarized in the following table. For the species in each group refer to pp. 165ff. [XXYI: 19 manus and terpsichores — have the most arresting coloration, in which dazzling white plays an important part: in all except stenodadyla, which is brilliant blue in front, white or pinkish-white behind, the carapace is completely white, terpsi- chores is white except for the cheliped, and saltitanta is altogether white. (It is in the two latter species alone that the anterior ambulatory meri are not specially colored.) In all of the above species, too, the displays are more highly organized. Copulation: Actual copulation was observed in stylifera, beebei and stenodachyla, a total of five times. In each case it was similar, taking place at the mouth of the female’s hole after a more or less prolonged and intense display which some- times, apparently, lasted for days between the same two individuals, subject to innumerable interruptions by the hesitation and frequent withdrawal of the females as much as by alarms Group 1 Gray or white Courtship Color Carapace Cheliped Ambulatories Anterior side of ambu- latory meri Buccal & pterygosto- mian regions Purple, orange, yellow and white in each Major side salmon or- ange, minor side white Group 4 Purple, brown-and-white or white Purple-and-blue, brown- and-white, or white Purple-and-blue, brown- and-white or white Peacock blue, purple, pinkish, white Turquoise, brown, white Group 5 Iridescent green, or iri- descent blue and white Gray or white Purple-and ochre or pink-and-white Each many shaded, orange through magenta Gray or red Gray or white a. Purple-and white, or scarlet b. Yellowish, plum red, white Green-brown, white Blue-and-green or white White Gray or purple a. b. a. b. a. b. a. b. Group 1 is, in regard to color as in physical form, the most homogeneous; the other two vary considerably in the different species. No group is characterized by colors specially confined to them. One of the most outstanding traits of coloration is the prevalence of white carapaces, which are found in all three groups, in a total of six species, although one of these six has the anterior portion bright blue. Often these species go through a phase of bright or dull yellow, or of ochre-streaks, before assuming the pure white. The phase before the yellow is usually dull gray. The gray carapaces found in princeps and deichmanni, the close relations of which display in white, may be evidence of relatively low development of color in these two species, cor- responding to their relative lack of specialization within their respective groups. Possibly, how- ever, none of the numerous specimens observed displaying was in full coloration: the examples of princeps were all small, around 15 mm. in length, although they acted in every way like adults. In the twelve species studied it is interesting that the most highly specialized forms in each group — e. g. stylifera, saltitanta, stenodadyla, lati- from birds. Each copulation witnessed was pre- ceded by mutual stroking of the legs and carapace with the ambulatories. The position was exactly similar to that illustrated by Pearse (1914.1) for pugilator in the laboratory. The two crabs faced each other, sternum to sternum, abdomens bent back, that of the male inside that of the female, the anterior ambulatories of each grasping each other around the body, the female with her posterior three pairs on the ground, supporting most of the weight of both. The chelipeds of the male rested above the female, the minor lying lightly on her carapace behind, or sometimes on top of, her eye, the major always clearing carapace and eyes, held flexed and perfectly quiet. The two remained to- gether, quivering at first then quiet, for a very short time, up to three minutes; then the female would very gently disengage herself and slip down her hole, which she soon closed up, regardless of the position of the tide. In at least one instance, the male definitely stood guard over it afterward and warned off other males, without waving the cheliped rhythmically. In heteropleura, stylifera, saltitanta, beebei, stenodadyla, latimanus and terpsichores, females 1941] Crane: Crabs of the Genus Uca 157 were seen to follow males, after vigorous display, down the holes of the males, remaining for any- where from a few seconds to at least until the next low tide, since they did not emerge before the tide covered the holes. Always, during courtship, the male appeared to employ inex- haustible patience and gentleness except in two cases involving princeps, when two males of this species, on different dates, displayed briefly, then crept up on indifferent females from behind and tried, unsuccessfully, to drag them over to their (the males’) holes. It is probable that copulation usually takes place in the burrow of the male, but that when the female has been brought to the proper pitch of excitement at the mouth of her own hole, copulation takes place there, since her burrow is usually too small in diameter to receive the male. Provision for safety from birds and lack of inter- ruption would seem to encourage copulation in the hole of the male, but the difficulty experienced by the males in arousing the females would explain the occasional surface pairings. Pearse’s observation that copulating females are hard-shelled was supported by my own observations. For further details of pairing, see below and pp. 173, 193, 196. Shelter Building: In 1873 Verrill reported that specimens of minax built “ovens” above their holes. Matthews, working in Brazil on leptodac- tyla, made a similar observation. He described the method of building very accurately, his obser- vations checking with my own made in Panama on beebei, latimanus and terpsichores. He attributed to the “hoods,” however, the function of keeping the burrows from drying out in the hot sun. What- ever their function — which I have not yet dis- covered— it cannot be this, for it is only courting males that build shelters ; the shorter burrows of the smaller females and young surely need protec- tion far more than those of large males, yet this device is never used by them; instead, when the sun becomes too hot for them they use a simple plug to close their burrows, or push up material from below. Of the three Pacific forms which were seen to build shelters, the instinct is least developed in beebei, best in terpsichores. In the first case the form of the shelters varies from little more than a pillar of sand beside the hole to a fairly well arched hood, little higher than the builder; also, it is not built by all displaying males, and is not necessary to win a female, since in the case of one of the two copulations witnessed in this species, the male had no shelter; in the other there was a well built shelter which played no apparent part in courtship. In latimanus a well formed shelter is always built by every displaying male. In terpsichores on the other hand the shelter is the best formed of all, and relatively the largest, but it is not built by every displaying male; however, the breeding season was waxing in this species when I left, and increasing numbers of shelters were being built every day. Incidentally, terpsichores and latimanus are closely related, and beebei more distantly, but all belong to the same group of highly specialized crabs, all of which are adapted for life on relatively firm, dry ground. Only beebei occurs on mud-flats as well as on muddy-sand beaches; in the mud-living individuals, the building of shelters is more spasmodic and unsuccessful than ever; the very nature of the semi-liquid mud would make the erection of a well-arched shelter almost an impossibility. This mysterious habit of building is probably a special development of the practice of stopping up the hole with a plug before the tide covers it. (Incidentally, this latter practice in the Pacific species studied is very casual, being rather frequently omitted by individuals in each species without any apparent reason). In making the shelter, as in plugging up the hole, the legs of the major side are used in scraping up and carrying the sand, and patting it into place, the crab always working from the underside of the growing half-dome. This method is in direct contrast to that employed in hole digging, when the minor side is invariably employed. It would seem that there is a distinct waste of effort when the crab, as often happens, first repairs his burrow, dump- ing the loads of sand several inches from the hole, and then builds the shelter, scraping the sand from a similar distance, but never using excavated, already loosened, damp sand for this purpose. The separate origin of the two activities explains this lack of correlation. For further details on the building activities of the several species see pp. 194, 203. On p. 196 is an account of equally inexplicable wall-building in mature females of stenodactyla. Behavior of Females during Courtship: As has been said, the role of the female, until the final stages of courtship, is one of complete indifference or definite retreat. Each male must display for long periods daily in the hope of attracting the attention of a single female sufficiently to make her pause in eating or passing, and watch him. Once this is accomplished, courtship may pro- ceed a dozen times a day to the point where copulation is about to take place, only to have the approach of a bird, or, even more frequently, the recurrent withdrawal of the female — often accompanied by the abnormally early plugging of her hole for the day — interrupt the courtship. The latter is resumed only after a repetition of lengthy preliminaries. In the cases observed where females either followed males down their holes, or mated with them on the surface, the successful male was sometimes only moderately large and brightly colored, compared with the most spectacular in size and color of the same species. How females, all so similarly formed and in- conspicuously colored, are recognized by males of their own species in mixed and crowded colonies remains a major mystery. Beebe ob- served a phenomenon in mordax which was also frequently apparent in various species in Panama, namely that a particular female often had the power to stimulate any number of males to violent display, merely by ceasing for an instant 158 Zoologica: New York Zoological Society to feed, or by emerging from her hole, or wander- ing a few inches from her usual position, whereas other individuals, equally large and of identical coloration, attracted little or no attention from the same males at the same time. For the behavior of stenodactyla when pursued by males, see p. 196. Origin of Display and Comparison: Little work has been done on courtship and mating in other groups of crabs. The principal study is that by Chidester (1911), who found that sex discrimina- tion is tactual in Callinedes, Cancer, Carcinus and Platyonychus. In Uca it is certainly not tactual, but instead largely or completely visual. As in Chidester’s examples, however, once a female has permitted a male to touch her — except in the case of females pursued and encircled by male stenodactyla — the female is not passive in the movements preceding copulation. Hediger (1934), although he does not think that waving of the large cheliped has anything to do with courtship, nevertheless presents a most sensible theory of its origin: “It is clear that the lifting of the chelae originated as a preparatory fighting motion (compare with other crabs), which as in many other cases, became only a threatening gesture, then was developed and transformed in this group into the signalling motion.” From my own observations, it appears that waving may be carried on throughout the year by many species of fiddlers, especially, perhaps, by the less highly specialized forms, as a purely threatening gesture, warning encroaching crabs away from a chosen feeding ground surrounding the hole. This type of display is relatively phlegmatic— unless an actual fight is imminent — and is often carried on with a perfunctory air while the minor cheliped is engaged in feeding. During the breeding season, however, the tempo of waving is greatly accelerated, and the activity is now devoted chiefly to the end of attracting females. Slow-waving-cwm-feeding periods still occur at this time, a habit which obviously enables a crab to advertise himself to females which might pass him unnoticing, if he did not wave while he ate. The function of waving in warning off males at this time also seems definite- ly connected with courtship in most cases, since the females in a given territory are perhaps more or less consciously preempted by the male. At least during the breeding season, when most of my observations were made, no male was ever seen to attack any except adult or nearly adult males: females and young came and went as they pleased, and there was far greater toleration between males of different species than between those of the same kind. Also, in a number of forms — especially in stenodactyla— females have been seen very definitely to be the cause of duels between males. The conclusions of Yerwey and Hediger, that waving has nothing to do with courtship, may be explained by the fact that courtship display may prove to be poorly developed in signata and tangeri, while their property sense is strong. On [XXVI: 19 the other hand, it may be that neither of these observers witnessed actual courtships in mid- breeding season. The latter explanation is per- fectly possible, even though Verwey, at least, must have spent a great deal of time in observa- tion: in Panama I had to wait days to see proof in certain species that males were actually stimulated to energetic display by interested or potentially interested females, and that the females without question paid attention and were influenced to receptivity by the display activities, marked by waving, of the males. In other sections of the animal kingdom, notably among birds, a number of cases are known where the courtship display or behavior is scarcely different from, or is identical with, warning and threat. Among the best known are the displays of pheasants, grouse, and certain sandpipers and other wading birds. The same is true of many songs: the male sings both to attract the female and to let possible rivals know, preferably without a fight, that he is in possession of a nesting site and will challenge interlopers. It is interesting to remember here that where, in the lyrebird and bowerbird the display has outgrown its function as a specialized part of the mating cycle and, as has been suggested, perhaps “become very largely of an almost recreational nature” (Stoner, 1940, p. 98), carried on through- out the year, the opposite seems to have hap- pened in the case of fiddler crabs, where courtship display probably arose from year-round threat activities. In latimanus, at least, it seems now to be restricted to the breeding season. The marvelous correlation between display and color found in the various species of birds of paradise is found to a lesser extent, but no less unquestionably, in fiddler crabs. Since courtship dances are well known in a number of insects and spiders, there seems to be no reason to deny the existence of such displays among crabs on the basis of their being invertebrates. Conclusions: From the foregoing data on Uca, a few conclusions may be drawn concerning the functions of display which are applicable to the species studied at Panama, all of which appear to be relatively highly specialized forms. It is important to remember that these conclusions are not necessarily true of the entire genus. 1. Waving and its accompanying behavior in fiddler crabs form, in the various species, dis- tinctive displays which are complemented and supplemented by temporary coloration patterns. These colors are shown to the best advantage only when the crab is in the midst of display. 2. No evidence at all has been found of the influence of sexual selection in the old-fashioned sense — that is, of a female’s deliberately choosing a brightly colored or especially active crab for a mate in preference to one which was duller or slower. Nevertheless, display coloration is so closely linked with the characteristic display in the various species, that it seems very unlikely that coloration is only an accidental and useless result, or a mere waste product, of glandular 1941] 159 Crane: Crabs of the Genus Uca or other physiological activity, or that display is simply the result of excess energy. Instead, it seems certain that females do distinguish and recognize males of their own kind by both motion and color — as well, perhaps, as by scent or some other means — and that they are eventually attracted and stimulated to mate by the display of persistent males, which are usually among the most brilliant or acrobatic. 3. At the same time, display doubtless serves also as a warning to rival males to keep away from a chosen feeding and display territory sur- rounding the hole. This function also is per- formed in other seasons of the year in some, but not in all, species. 4. Associated with display in certain species is the erection of shelters of muddy sand above the hole. No function can be attributed to these yet, except that they probably serve as further advertisement of the presence of a male in breeding condition. G. Breeding and Growth. The observations at Panama were all made between January and March; the ovigerous females in the Zaca collection from the west coast of Central America were also taken at this season, but since these dates coincided with the dates of both trips, no real conclusions can be drawn. However, since sunlight and dryness are needed to bring out the display colors of males, and since these months are the height of the dry season, it seems reasonable to assume that fiddlers have at least a major breeding season at that time. Evidences of sexual rhythms were seen in the species studied, some being apparently near the close of the breeding season while others were obviously entering it. Some (e. g., panamensis and umbratila) were not displaying, except for a few abortive displays in the former species, nor were ovigerous females seen. An account of tidal rhythm in the display of latimanus has already been given (p. 155). Ovigerous females appeared to move around less than non-ovigerous ones, but were frequently seen feeding in full sunlight. They were never the object of display by a male. It seems likely that ovigerous females of the stylifera group, at least, come ashore to copulate, but go down to the mud flats to carry their eggs (p. 172). No fiddlers were seen at night, even in full moonlight, on the two evenings when observations were made (c. f. Pearse, 1912). The eggs, counted in eleven species, numbered between 500 and 15,000, relatively low totals compared with those of cancroid crabs, for example. Similarly, the eggs are relatively larger, and of remarkably similar sizes, ranging between .24 and .27 mm. in diameter. No special studies have been made on develop- ment, but from time to time in the following pages notes are included on the general growth trends and characteristics of young crabs of various species, which will be useful only in identification and as hints on intrageneric rela- tionships. Detailed work must wait for the future. Apparently no holes are dug until at least several crab instars have been passed, and the crab reached a length of 3 mm. or more. Before this time the crabs run freely in and out of the burrows of larger crabs of both their own and other species. Young crabs in their physical characteristics, as is to be expected, frequently give clues to their relation to other species. For example, the young of macrodactyla can with difficulty be distin- guished from adults of zacae. As is well known, the young lack the characteristic ridges inside the palm, the whole cheliped gains its size and elongate chelae only gradually, and often the orbits are more oblique than in the adult. Often, too, there is more pile on the carapace, or pile which is quite absent in the adult (as in umbratila and oerstedi). In stenodactyla, latimanus and their allies, however, with the carapaces semi- cylindrical and the orbits scarcely oblique, the orbits even of crabs less than 2 mm. long are similarly almost straight, although the carapace is relatively flat. In the young of crabs of the very narrow-fronted group containing princeps and allies, the front is wider than in the adult, but still so narrow that there is no danger of con- fusion with other groups of species. Spoon-tipped hairs on the second maxillipeds are fewer in young than in adults (see p. 161). Precocity: Adolescent males, distinguishable by their smaller size and short-fingered chelipeds, sometimes may be observed apparently in the midst of learning to court and build shelters. These individuals have not attained full court- ship coloration, and their movements during display show various stages of practice. The display usually is erratic and casual. Similarly, shelters may be started but not completed, or the crab may spend an entire morning building a structure which is small and badly made. Once I saw such a shelter toppled over by a strong gust of wind. Special examples of precocious behavior were noticed in beebei, where an obviously immature male, without a shelter and in poor coloration, induced a female to follow him down his hole, after an energetic display. Once she was down there, however, he became frantically restless and popped out and in again every few minutes, displaying vigorously in the direction of other passing females. Another young male, this time a stenodactyla, after fruitlessly courting a female who paid no attention whatever, and who finally vanished down her hole around which she had just finished building a high wall (see p. 196), deliberately walked over to her wall and pulled it down with his ambulatories, trampling it into the ground until no sign of it was left. He then re- turned to normal feeding, without display, beside his own hole. H. Individuality and Play. A strong spirit of individuality was observed in the fiddler crabs, and I agree with Pearse (1912, 160 Zoologica: New York Zoological Society 1914. 2) that some of their behavior can only be interpreted as sheer play. Several adjacent males of similar size, belonging to the same species, on the same day would show definite traits of individuality. One would be especially belliger- ent, seeking every excuse for a duel; another would build a shelter and display strenuously all day, scarcely stopping to feed; the third, although he had spent most of the preceding day fighting and courting, might on this morning feed con- tinuously and enlarge his burrow, punctuating this activity with only a few half-hearted dis- plays. Similarly, some females of various species were much given to wandering about, peering down the holes of adult males, hurriedly retreating, paying brief and successive attention to the dis- plays of a number of neighboring males, and altogether behaving in a manner which in higher animals would certainly be termed coy and flirtatious. I have used these ultra-anthropo- morphic terms advisedly, because I have been unable to find any other words in the language which so exactly define the actions of these individual females. Others, of the same species, spent hours feeding quietly on several square inches of ground. In the end members of the latter group proved just as susceptible as the wandering individuals to the advances of dis- playing males. Two large male stylifera furnished a good example of a social relationship of sorts which continued for at least a week. Their burrows were a yard apart, in an uncrowded portion of the beach, although other males of their own species were close by. Every day they followed an invariable routine, consisting of emergence, cleaning, feeding, accompanied by change to dis- play coloration, and then — without a sign of preliminary waving or warning or argument of any kind — they would meet on the invisible boundary line between their burrows and fight. The duel always ended several minutes later in identical fashion, the smaller being somersaulted backward by the larger. The vanquished would then pick himself up and retreat hastily to his burrow, while the winner resumed feeding with- out another glance. After a half hour or so more both would begin to display, without taking any further notice of each other. I never saw two fights in one day, and there was never any female in their vicinity. Finally both moved away and I lost track of them. The apparently sporting aspect of part of the courtship activities of stenodadyla is described on p. 196. One of the most individualistic, inexplicable performances I saw was that of a moderate-sized but apparently adult male terpsichores. His dis- play coloration was not well developed on the day in question, his usually white carapace being heavily streaked with dull yellow and his cheliped scarcely pink. He did not build a hood or display, but enlarged his burrow and fed energetically. Then, suddenly, he went straight over to the newly erected shelter of a neighbor fully eighteen [XXVI: 19 inches away. Without any provocation or pre- liminaries he undermined the shelter from the rear and pushed it down on top of its owner; the two crabs then spent 15 minutes fighting, in the course of which both darkened rapidly, losing all trace of display coloration, and the shelter owner lost the tip of his pollex. Finally, the aggressor let the owner go, then went directly to the next hood, six inches from the first, and repeated the episode exactly. In this case, too, the owner was powerless and was constantly thrust down his own hole, although he put up a good fight. At last, after another 25 minutes of uninterrupted struggle, the aggressor released this crab also, and returned, without any hesitation, to his first vic- tim, who by now was cleaning himself up and had regained most of his display coloration. At the approach of his former antagonist, the victim tried to flee down his hole, but was seized from behind. Another duel, lasting no more than several minutes this time, followed, and ended as on the first two occasions by the aggressor’s abruptly releasing his victim. This time the former returned slowly but directly to his own hole, cleaned himself, and began to feed. Neither of the two victims rebuilt their shelters on that day, although the tide was only slightly past dead low at the time. The general conclusion to be drawn from all this variability of action is that fiddler crabs, nervously the most highly organized of all Crustacea, show a truly remarkable latitude of behavior. This is especially striking when fiddlers are compared with ants and bees, which are tied down to severely patterned behavior by the hyper-development of their social organi- zation. I. Enemies and Defense. Along the coast of Central America the greatest enemies of fiddler crabs are shore birds At La Boca alone five species of herons, snowy and American egrets, curlews, sandpipers, herring gulls and great-tailed grackles hunted fiddlers daily at low tide. At Port Parker, Costa Rica, numerous sandpipers were hunting them. Every- where, too, raccoon tracks were found among the fiddler holes, especially on mornings when the low tide came very early. Fiddlers were also taken from the stomach of a lizard, Ctenosaura similis, at Culebra, C. R., and from that of another, Basiliscus galeritus, on Gorgona Island, Colombia. The crabs are adept at escaping all these ene- mies, and yet in wasting as little time under cover as possible. Each colony, whether mixed or formed of a single species, has worked out a scale of alarms based on the movement of sus- picious objects. At La Boca noises, ranging from the cries of their bird enemies to the shouts of human beings, whistles, cannon-fire and dyna- mite, had no meaning for them. Neither did the passing of butterflies and wind-blown leaves within an inch or two of their eyes. But a bird flying over within twenty-five feet or a plane within, say, two hundred, was the “alert” which sent all the crabs scurrying to the mouths of 161 Crane: Crabs of the Genus Uca 1941] their holes, where they froze, poised for instant flight within. The “take cover” signal was the approach of a bird either on foot or wing within ten to twenty feet, depending on both bird and crab, and the approach of a human being within, on the average, thirty feet. This distinction among “no cause for alarm,” “on the alert,” and “take cover” must save them a great deal of time for feeding and courting during every low tide. A most interesting point is that brightly colored adult males are the first to enter their burrows and the last to leave. Y. Phylogeny. The study of the present collection of eastern Pacific fiddler crabs has shown the importance of a number of physical characteristics to which little attention has previously been paid. These, combined with the occurrence of eleven species hitherto undescribed, have clarified considerably the relations of the various species to one another. In spite of the fact that no complete picture can be formed until the genus as a whole has been similarly and more exhaustively studied, it seems worthwhile to present some tentative con- clusions in regard to these relationships. (Text- figure 5). The first necessity was to discover the primi- tive forms of the various physical characters, in order to decide which species were as a whole the least specialized. It appears that in a theoretical, primitive Uca the carapace is moderately arched, strongly narrowed behind, the orbits strongly oblique, the front moderately narrow, the major cheliped relatively small with short fingers and lacking tuberculated ridges across the palm, the minor cheliped with well developed teeth and a slight gape, the third maxilliped with a rudi- mentary median groove traceable only anteriorly, the second maxilliped with a moderate number of spoon-tipped hairs, and the abdominal ap- pendage stout with a thick arm near its tip. It would presumably live in a fairly stable environ- ment, not subject periodically to great dryness, and would have neither coloration nor display highly developed. A few species, especially pygmaea, argillicola and helleri, fulfill the majority of these conditions, although each of these three has already started to specialize in various divergent fashions. Unfortunately, the display of none of these has been observed. From this theoretically primitive form, special- ization has proceeded in a number of directions. These may be divided into ecological and structur- al, the latter being closely dependent on the for- mer. For this reason, the former will be listed first. 1. Ecological Specialization. a. Moving from stable, damp habitat, such as briefly exposed mud-flats, or mangrove marshes, to a periodically dry habitat, such as sloping shores or the banks of small fresh water streams. b. Color change associated with courting. c. Shelter-building associated with courting. d. Display dances associated with courting. From available material, it appears that b, c and d are most highly developed in those crabs which have been most modified for a truly littoral life. 2. Structural Specializations. a. Increase in Thickness, through Great Arching of Carapace and Underparts, Straightening of the Orbits, Divergence of Carapace Sides Posteriorly, Fusion of Abdominal Segments: These modifica- tions serve the two-fold purpose of guarding against dessication, and of giving more room in the branchial chambers for the exposure of blood- vessels, an arrangement which functions as a primitive lung when the crab remains out of water for so long that the gills become tempora- rily useless. This increase in thickness, accom- plished by most or all of the means listed above, becomes evident in end-species of Groups 2, 4 and 5, forming a good example of convergent evolution. It is most highly developed in Group 5, in which the end-forms all live a strictly littoral existence and spend hours daily exposed to the sun and air. The bottoms of the burrows of latimanus, the thickest of all, may not be covered by the tide for more than a week at a time. Although the species in Group 1, char- acterized chiefly by very narrow fronts, are in an isolated series, they are in many particulars highly specialized. They are all fairly fiat, and there is some evidence to show that they come ashore only to court, but otherwise live on briefly exposed mud-flats. U . panamensis, the specialized crab forming Group 6, is noticeably flattened; this is probably associated with its unfiddler-like habit of hiding under stones. b. Specializations of Mouthparts: The signifi- cance of most of these adaptations cannot even be guessed at. They include: i. Grooving or smoothing of ischium of third maxilliped (PI. VII). In the least specialized forms, and continuing through all groups except 2, 3 and end-species in Group 5, the ischium is moderately flat with a well developed inner groove and a central groove represented only by an anterior (distal) median depression. In the end-forms of Group 2 the median groove is pro- gressively better developed, extending posteriorly (basally) and swerving inward to fuse almost or completely with the basal end of the inner groove. There seems to be no practical significance to this trend, except to show a fundamental rela- tionship. In Groups 5 and 6 the ischium is practically smooth except for an inner groove, being flattened in Group 6, as is the rest of the crab, and broad and swollen in Group 5, the swelling being obviously merely a continuation of the general tendency to increase bulk in order to mitigate dryness. ii. “Spooning” of hairs on merus of second maxilliped (Text-figs. 2, 3). The most usual state is for the inner edge of the anterior half of the merus and the tip of the palp to have many or all of the hairs terminating in concave, pectinated 162 Zoologica: New York Zoological Society [XXVI: 19 £ c 3> Text-figure 2. Inner (dorsal) view of merus of second maxilliped in Uca. A, princeps; B, mordax; C, oerstedi; D, latimanus. expansions, termed “spoon-shaped” for the sake of brevity. Departure from this norm extends in both directions, toward total reduction and toward tremendous increase, both of these de- partures taking place only in end-species. In Group 1 they are more than moderately numer- ous, and in addition have characteristic spines at the base of the shallow spoons, the pectina- tions of the latter being rudimentary. Through- out Group 5, where it reaches its highest develop- ment, spooning is further increased. On the other hand, in Groups 2 and 4 the spoons become progressively fewer, until in end-forms they are almost or completely lacking. Since they are best developed in species living at least part of their lives on muddy sand shores, as opposed to semi- 1941] Crane: Crabs of the Genus Uca 163 Text-figure 3. Typical spoon-tipped hairs from merus of second maxilliped in Uca, front and lateral views. A, princeps; B, mordax ; C, oerstedi; D, latimanus. liquid mud, it is probable that their function is iii. Development of “woolly” hairs on second concerned with the sifting of organic detritus maxillipeds. These structures have not been from this relatively dry medium. studied at all in detail, and they are included here 164 Zoologica: New York Zoological Society [XXVI: 19 Minor chelipeds in Uca. A, heteropleura; B, pygmaea; C, zacae; D, galapagensis; E, mordax; F, brevifrons; G, macrodactyla; H, tomentosa ; I, umbratila; J, argillicola; K, oerstedi; L, inaequalis; M, tenuipedis; N, batuenta; O, saltitanta ( Published with permission of the Secretary of the Smithsonian Institution. Garrulax chinensis lochmius, subspecies nova, with the type an adult male, U. S. N. M. no. 336663, collected at Chiengsen Kao, North Thailand, January 15, 1937, by H. G. Deignan. Of lochmius, I have sixteen specimens from the provinces of Me Hong Son, Chiengmai, Chiengrai, and Nan. Probably may be placed here also Salvadori’s example from Yado, Karenni, which “somiglia piu ad un esemplare di Hong-Kong e forse appartiene ad una forma distinta, intermedia fra la cinese e quella del Tenasserim” (Ann. Mus. Civ. St. Nat. Genova, ser. 3, vol. 6, 1914, p. 6). III. A representative of Pomatorhinus erythrogenys, recorded in Thailand only from Doi Chiengdao, between 4,500 and 6,800 feet, has proved to be, not imberbis (Karenni), as hitherto believed, but a quite different race, which may be known as Pomatorhinus erythrogenys celatus, subspecies nova, with the type an adult male, U. S. N. M. no. 336873, collected on Doi Chiengdao, North Thailand, March 20, 1937, by H. G. Deignan. From imberbis it is readily distinguished by having the frontal feathers, sides of throat, neck, breast, and belly, the flanks, thighs, and under tail-coverts distinctly paler — orange-rufous, in- stead of chestnut-rufous; the feathers of the lores dark gray, not grayish-white. From erythrogenys, it differs in its smaller size, more vivid color along the sides of the body, and dark gray (not grayish-white) lores. According to Stuart Baker (Fauna Brit. India, Birds, 2nd ed., vol. 1, 1922, pp. 220-222), all races of erythrogenys have the “iris light greenish white, yellowish white or pale bright yellow; legs and feet fleshy or fleshy-brown; bill yellowish-horny.” The male of celatus has the irides red; the orbital skin plumbeous; the bill gray, with basal half dark slate; the feet, toes, and claws horny-brown. With the new form I place skins from Kalaw and Taunggyi, in the Southern Shan States. Imberbis will, perhaps, like Pomatorhinus sch. nuchalis (a similarly saturate race), prove to be restricted to the Karen Hills. 1941] Breder: Respiratory Behavior in Fishes 243 25. Respiratory Behavior in Fishes Not Especially Modified for Breathing Air Under Conditions of Depleted Oxygen. Charles M. Breder, Jr. New York Aquarium. (Plate I). Among fishes living in environments which are usually or periodically depleted of dissolved oxygen are many with specialized respiratory accessories which permit them to make use of atmospheric oxygen direct or indirectly. Fishes without equipment capable of coping with situa- tions of depleted oxygen must perish when con- fronted with them unless escape is possible. When simple escape by flight is impossible, at least certain of such fishes will make attempts to seek out methods of survival which are certainly not part of the normal scheme of their activities. Two cases, indicative of such efforts at survival, seem to be rather suggestive of a point of approach to the whole problem of fish behavior in adjusting to changing environmental conditions. One case involved a large female Lebistes reticulatus (Peters) in the laboratory of the New York Aquarium which was being used for some experimental work involving other matters but which incidentally concerned activities under restricted oxygen supply. In the course of this experiment small measured blocks were floated on the surface as a means of restricting gaseous interchange between the air and water. In the one referred to, the surface was nearly covered with these floats with one Lebistes in the water below. Following its natural response to rise to the surface and take advantage of the greater amount of oxygen near the surface film, which by virtue of peculiarly specialized mouths the Micro- cyprini are able to do efficiently, it found normal behavior under such conditions impossible be- cause of the crowded floating blocks. After some little effort it managed to wriggle up between two of the blocks and bear its weight on them in such a fashion that the blocks tilted and pre- sented a sloping surface. On this support it perched itself with only its tail immersed. Here it would rest for long periods of time, now and again submerging presumably because of the drying effects of the air. It looked very much like the normal performance of a Periopthalmus. This is shown in Plate I, Fig. 1. Such behavior went on for six days until the experiment was discontinued.1 One is forced to wonder about the origin of habits in fishes in which they volun- tarily leave the water, for certainly Lebistes does not come out of water under normal conditions. Some of the Microcyprini voluntarily leave the water even when there is no immediate suffocation, as anyone familiar with Rivulus can attest. These fish may be sometimes found flipping their way along in damp jungles, Breder (1927), or sometimes buried in the damp debris of the jungle floor. Their method of emergence and subsequent behavior is entirely different from that of Lebistes and apparently there is no connection between the two. The second case involves a species of bottom fish, Achirus lineatus (Linnaeus). One such specimen, 46 mm. standard length, kept in a small aquarium at the Florida Field Laboratory of the New York Aquarium, was noticed per- forming in an unusual manner when, because of the death of some tank mates, the aquarium became foul. It would swim up to the surface and there flutter its long dorsal and anal in such a manner as to impel backwards the water above it, while in the meantime it would cup its body in such a fashion as to be fairly dish-shaped. By this action it cleared its upper concave surface of water and rested floating on the surface film by virtue of the water displaced. It is shown floating in this manner in Plate I, Fig. 2. This means of flotation was not entirely perfect, and it was forced intermittently to keep up the activity to prevent itself from sinking. By this performance, however, it apparently was able to take advantage of the greater concentration of oxygen at the surface in a manner similar to that of some of the Microcyprini with their flattened heads and upturned snouts. This fish finally succumbed after several days so it cannot be said whether its method was poor or merely 1 Physical data on experiment (Nov. 6, 1933) Temp. 25.5° C.; Vol. H2O 1488cc.; Surface area 3.5 C2; Free CO2 0.44 mM. For this experiment a mason jar was used as an aquarium. In another experiment fish died at a concentration of 0.46 mM. CO2. 244 Zoologica: New York Zoological Society inadequate to the circumstances of this particular aquarium. The only reference that we have been able to find with a bearing on this item is in Beebe & Tee-Van (1928). They report that this species performed in the following manner in Port-au- Prince Bay, Haiti. “On several nights I caught young soles of this species, near the surface, at our submerged light. They swam slowly along and when at the surface elevated the encircling ring of vertical fins, and depressed the body, and in this cupped shape floated with no apparent movement of fins or body. The tips of all the rays could be seen breaking the surface film, but I could see no difference in the level of the en- closed water and that outside. These specimens measured from 17.5 to 25 mm.” Their fish were only about half as large as the present and were certainly under no suffocating influence where taken. It is thus evident that there is a back- ground for this behavior in the activity of the smaller sizes. Apparently it is normally given up before the size of our specimen is reached. The present behavior then might be considered as a return to more juvenile activity on the incidence of adverse circumstances. At the Field Laboratory many were kept under conditions far from ideal, but the present case is the only one which displayed this habit. Al- though submerged night lights were used con- tinually they failed to attract this species. Most of our specimens were considerably larger than those of Beebe & Tee-Van. [XXVI: 25 The two cases mentioned are the only examples known to the writer who, because of circum- stances, has had unusual opportunity both in the field and at home for a long period to note such items of behavior. How many Lebistes have been handled in that time would be hard to estimate but one item of behavior in this as well as related poeciliids stands out prominently. These fishes, when in pools with gently sloping edges, feed freely in very shallow water but are notably careful to avoid being “stranded.” Consequently it must be assumed that the individual herewith discussed overcame this tendency in response to the greater pressure of suffocation. During this same period several hundred achirids of various species and under varying conditions have never shown any disposition to the floating habit herein described. Surely here is a waiting field of investigation in fish behavior, touching perhaps on the farthest reaches of the mental life of fishes. Such ques- tions arise as to how instinctive are such acts and how widespread in an individual species. References. Beebe, W. & Tee-Van, J. 1928. The Fishes of Port-au-Prince Bay, Haiti, with a summary of the known species of marine fish of the Island of Haiti and Santo Domingo. Zoologica, 10 (1): 1-279. Breder, C. M., Jr. 1927. The Fishes of the Rio Chucunaque Drainage, Eastern Panama. Bull. Amer. Mus. Nai. Hist., 57 (3): 97-176. EXPLANATION OF THE PLATE. Plate I. Fig. 1. Lebistes reticulatus resting out of water after the manner of a Periopthalmus. Fig. 2. Achirus lineatus floating on the surface film of water depleted in oxygen. BREDER PLATE I FIG 2. RESPIRATORY BEHAVIOR IN FISHES NOT ESPECIALLY MODIFIED FOR BREATHING AIR UNDER CONDITIONS OF DEPLETED OXYGEN. 1941] Beebe & Tee-Van: Rays, Manias and Chimaeras 245 26. Eastern Pacific Expeditions of the New York Zoological Society. XXVIII. Fishes from the Tropical Eastern Pacific. [From Cedros Island, Lower California, South to the Galapagos Islands and Northern Peru.] Part 3. Rays, Mantas and Chimaeras.1 William Beebe & John Tee-Van Department of Tropical Research, New York Zoological Society. (Plates I-IV, Text-figs. 1-40). [This is the 28th of a series of papers dealing with the collections of the Eastern Pacific Expeditions of the New York Zoological Society made under the direction of Dr. William Beebe. The present paper is concerned with specimens taken on the Noma Expedition (1923), the Arcturus Oceanographic Expedition (1925), the Antares Expedition (1933), the Templeton Crocker Expedition (1936), and the Eastern Pacific Zaca Expedition (1937-1938). For data on localities, dates, dredges, nets, etc., of the second, fourth and fifth of these expeditions, refer to Zoologica, Vol. VIII, No. 1: 1-45 ( Arc- turus), Zoologica, Vol. XXII: 33-46 (Templeton Crocker), and Zoologica, Vol. XXIII: 278-298 (Eastern Pacific Zaca).] Introduction 245 Key to tropical eastern Pacific rays and mantas. . . . 247 Superorder Platosomeae (rays, mantas) Order Narcobatea Suborder Narcobatida Family Torpedinidae Narcine entemedor 247 Narcine ommata 248 Narcine vermiculata 248 Discopyge ommata 249 Discopyge tschudi 250 Order Batea Suborder Batida Superfamily Rhinobatoidea Family Rhinobatidae Rhinobatus planiceps 251 Rhinobatus productus 251 Rhinobatus glaucostigma 251 Rhinobatus leucorhynchus 252 Zapteryx exasperatus 252 1 Contribution No. 630, Department of Tropical Re- search, New York Zoological Society. Part 1, Lancelets and Hag-fishes, was published in Zoologica , Vol. XXVI, pp. 89-92; Part 2, Sharks, in Zoologica, Vol. XXVI, pp. 93-122. Family Pristidae Pristis zephyreus 253 Pristis pectinatus 253 Superfamily Rajoidea Family Rajidae Raja aguja 254 Raja badia 254 Raja ecuadoriensis 255 Raja equatorialis 256 Raja inornata 256 Raja sp. A 257 Raja sp. B 257 Psammobatis scobina 258 Psammobatis brevicaudatus 259 Psammobatis lima 259 Psammobatis spinosissimus 259 Superfamily Dasybatoidea. Family Dasybatidae. Dasyatis brevis 260 Dasyatis longus 261 Dasyatis pacificus 262 Gymnura marmorata 263 Urotrygon aspidurus 264 Urotrygon asterias 265 Urotrygon binghami 266 Urotrygon chilensis 267 Urotrygon goodei 267 Urotrygon mundus 268 Urobatis concentricus 268 Urobatis halleri 269 Urobatis maculatus 270 Family Aetobatidae Aetobatus calif ornicus 271 Aetobatus peruvianus 271 Pteromylaeus asperrimus 272 Stoasodon narinari 272 Family Rhinopteridae Rhinoptera steindachneri 273 Family Mobulidae Mobula lucasana 273 Manta hamiltoni 274 Subclass Hoiocephali (chimaeras) Superorder Chismopneae Order Chimaerea Suborder Chimaerida Superfamily Chimaeroidea 246 Zoologica: New York Zoological Society [XXVI: 26 CLARION TENACATITA B. MANZANILLO SIHUATANEJO ACAPULCO'' DULCE R..V PORT ANGELES-G PORT GUATULCO- SANTA CRUZ B." TANGOLA-TANGOLA B. GU LF CORINTO SAN JUAN DEL SURjT&^V s PORT PARKER MURCIELAGO B/ POTRERO GRANDE PORT CULEBRA- BRAXILITO B. PIEDRA BLANCA B.O ,GULF OF DULCE (c GULF OF (.LC.T..Z ,SL, NICOYA 1 COCOS O ISL. EASTERN PACI FIC EXPEDITIONS NEW Y O R X ZOOLOGICAL SOC E T Y SHORE COLLECTING STATIONS UVITA 1 '■ Vi <- PEARL 1 B j ; f 2 \ > O PAR IDA J '"BAHIA l > o ISL. / HONDA \ 1 r. COI BA ISL. j BANK) ( K GORGONA ISL.-/ 7 03 > GA L APAGOS IS. ECUA DOR 4/ PERU Text-figure 1. Principal localities in the tropical eastern Pacific where collections were made by the Department of Tropical Research of the New York Zoological Society. Family Rhinochimaeridae Harriotta curtiss-jamesi 278 Family Callorhinchidae Callorhinchus callorhynchus 279 References 279 Introduction. This paper deals with rays and mantas col- lected in tropical eastern Pacific waters on five expeditions of the Department of Tropical Research of the New York Zoological Society under the direction of Dr. William Beebe. As an aid to future students we have included references to all species recorded from the tropical eastern Pacific. The geographical boundaries of the region under consideration in this paper, and which we call the “tropical eastern Pacific,” are as follows: The coastal waters of North and South America from Cedros Island, off the coast of Lower Cali- fornia, and the Gulf of California, southward to northern Peru, including off-shore islands such as the GaMpagos and Revillagigedo groups, Clipperton, Cocos and Malpelo Islands. As far as references are concerned, we have listed the original reference with type locality, and references referring to the region under discussion. Additional references have been added, however, whenever their inclusion was felt necessary. Some of the more commonly 1941] 247 Beebe & Tee-Van: Rays, Mantas and Chimaeras cited papers have been referred to by name and not by publication, serial and page numbers; the full references will be found on page 279. Syno- nyms of nominal forms described from the region are included, but not necessarily those from extra-limital localities. Forty-three rays and two chimaeras are reported from the waters of the tropical eastern Pacific. We are indebted to Miss Janet B. Wilson for inking the drawings. Key to the Families of Rays and Mantas of the Tropical Eastern Pacific. 1. Snout long, slender and flat, armed with strong teeth laterally (sawfishes) . . Pristidae, p. 253 Snout not armed with teeth laterally 2 2. Tail comparatively thick, bearing two dorsal fins; with or without a well-developed caudal fin; no serrated caudal spine 3 Tail slender, with one or no dorsal fins; bearing a filamentous caudal fin in the Dasyatidae, otherwise without a caudal fin 5 3. Caudal fin well developed; ventral fins not notched on outer edge (do not mistake notch formed by clasper of males for a notched fin) 4 Caudal fin absent, or represented only by a slight fold of skin; ventral fins notched on outer edge (do not mistake the notch formed by the clasper of the male for a notched fin); skin rough with scattered spines Rajidae, p. 254 4. Skin rough with scattered spines; electrical organs not developed .... Rkinobatidae, p. 250 Skin everywhere perfectly smooth; electrical organs developed Torpedinidae, p. 247 5. Dorsal fins absent; eyes not at edge of head Dasyatidae, p. 260 A single dorsal fin present; eyes at edge of head 6 6. Head with a pair of horn-like arms projecting forward Mobulidai , p. 273 No horn-like arms projecting forward 7 7. Snout bilobed Rhinopteridae, p. 273 Snout single lobed Aetobatidae, p. 271 Family Torpedinidae. Key to tropical eastern Pacific genera. la. Pelvic fins distinct from body posteriorly, not joined together to form a disk Narcine. lb. Pelvic fins united to body along their entire length, joined together to form a disk Discopyge. Narcine Henle, 1834. Key to tropical eastern Pacific species. la. Disk with a single large ocellus in the center. ommata. lb. No ocellus in center of disk. 2a. Disk with a few inconspicuous rings in the young, becoming uniform in color in the adult cntemedor. 2b. Disk with numerous white vermiculations. vermiculatus. Narcine entemedor Jordan & Starks. Ocellated Electric Ray. Range: Gulf of California to Panama (Mexico : San Felipe Bay, Gonzaga Bay, San Francisquito Bay, Mulege, Ballenas Bay, Carmen Island, Agua Verde Bay and La Paz, all in the Gulf of California, Mazatlan; Costa Rica: Gulf of Nicoya; Panama: Panama Bay). Field Characters: Tail thick with two dorsals and a caudal fin well developed; pelvics not notched on outer edge; skin smooth everywhere. The ray is provided with electric organs which generate sufficient power for adequate protection. Pelvic fins separate. Young with several faint ocelli, disappearing in adult. (Illustration from Specimen 24,996; 385 mm.) Color: The adult is uniform putty brown with indistinct spots of dusky on the body, both dorsals and caudal fin. In young individuals there are about four pairs of large ocelli on the upper surface, dark-centered with a pale outer ring of yellowish-tan. These markings fade with age so that in some specimens they are almost or quite absent. Size: The largest recorded specimen is 762 mm. in length. Local Distribution: We took this species only in Inez Bay, on the west coast of the Gulf of Cali- fornia. In mid-April they were found in shallow water at night. Abundance: Records in literature are of single specimens, except at La Paz where they were said to be common. We found them abundant at night in Inez Bay. Food: Specimen No. 25,249 had in its stomach a pure culture of red polychaete worms, armed with strong spicules. No. 24,996 had eaten six polychaete worms and one ascidian. Study Material: 13 specimens. Mexico: Inez Bay; 12 rays seen, 6 collected (24,996, 25,235, 25,249); Photo. 7565; length 330 to 385 mm., April 13 to 16, 1936. Speared at night in shallow water in the southern part of Inez Bay. References: Narcine entemedor, Jordan, D. S., Fishes of Sinaloa, 1895: 387, 508 (original description, color; type locality, Mazatlan, Mexico; La Paz). Gilbert, C. H., & Starks, E. C., Fishes of Panama Bay, 1904: 15, 207 (description, Gulf of California, Panama). Osburn, R. C., & Nichols, J. T., Bull. Amer. Mus. Nat. Hist., 35, 248 Zoologica: New York Zoological Society [XXVI: 26 1916: 144 (Mexico: Agua Verde Bay, Mulege). Meek, S. E., & Hildebrand, S. F., Marine Fish of Panama, 1, 1923; 74 (comparison with Narcine brasiliensis) . Breder, C. M., Jr., Bull. Bingham Oceanogr. Coll., 2 (1), 1928: 5, fig. 2 (Mexico: San Francisquito, Gonzago Bay, San Filipe Bay; Ballenas Bay on Carmen Island; color of young and adult, figure) . Narcine omtnata Clark. Range: Pacific Coast of Central America; exact locality unknown. Field Characters: Tail thick with two dorsals and a caudal fin well developed; pelvics not notched on outer edge; skin smooth everywhere; electric organs present; pelvic fins separate; a brilliant ocellus in center of disk. Color mark- ings strikingly similar to those of Discopyge ommata Jordan & Gilbert ; but the species differs in not having the ventral fins united into a con- tinuous disk as in Discopyge. Study Material: None. References: Narcine ommata, Clark, H. W., Proc. Cal. Acad. Sci., 21, 1936: 383 (original description; type locality. Pacific coast of Central America, exact location unknown; type No. 5444, Mus. Cal. Acad. Sci., Ichthy.). Narcine vermiculatus Breder. Vermiculated Electric Ray. (Plate I, Fig. 1). Range: Mexico, El Salvador. (Mexico: “west coast of Mexico,” 15 miles west of San Benito, South Mexico, at 14° 40' 20" N. and 92° 40' 30" W.; El Salvador, off La Libertad, 13 fathoms.) Field Characters: Tail thick, with two dorsals, caudal fin well developed ; pelvics not notched on the outer edge; skin smooth everywhere; electric organs present; pelvic fins separate from each other; color as below. (Illustrations after Breder, 1928; 204 mm.). Color: Deep chocolate brown above with numerous spots and vermiculations of white; ventral surface white except posterior edges of ventrals and pectorals which are slightly dusky; posterior edge of both dorsals and caudal edged with white ; a single white spot a little longer than spiracle at the center of each dorsal; two irregular white bands across caudal; line bounding upper and lower coloration on peduncle, above dermal fold, wavy. In general the light marks tend to run transversely over body. Size: Kumada records a two-foot specimen. Study Material: 1 specimen, El Salvador: off La Libertad. Young male, (27,523), 58 mm., Dec. 16, 1937, dredged at station 198: D-l, 13 fathoms. References: Narcine vermiculatus, Breder, C. M., Jr., Bull. Bingham Oceanogr. Coll., 2 (1) 1928: 6, figs. 3 and 4 (original description; type locality, 14° 40' 20" N., and 92° 40' 30" W., 15 miles west of San Benito, South Mexico; type No. 1143, Bingham Oceanogr. Coll., Pea- body Museum, Yale Univ.). Narcine sp., Kumada T., & Hiyama, Y., Marine Fish Pac. Coast Mexico, 1937; 21, Plate 54, fig. B (short description, color, figure). Discussion: Four specimens of this beautiful ray are now known, three of which are males. The nearest related species is Narcine brasiliensis, but differing from this, to mention only one of several characters, the Pacific ray has the pre- orbital snout in the disk length, 3 to 3.6 times, while in the Atlantic fish the proportion is 4.4 times. The variation in color and pattern in brasiliensis is extreme, and is apparently con- cerned with and influenced by locality, individu- ality and development. This prepares us for acceptance of the corre- sponding but less extreme variation in the four known specimens of vermiculatus , as intraspecific. In our 58 mm. ray the pattern is in the form of large, well-defined white spots on the body and central disk and pelvics, which become short, irregular bands on the disk margin, numbering 11 or 12 altogether. On the upper tail the white is in the form of 3 complete bands, and a 4th across the caudal fin. The 2nd and 3rd tail bands are enforced by a large spot on each dorsal. Both the dark color and several of the white bands overlap on the under side of the posterior half of both pectorals and pelvics. The next ray in size is Breder’s, which is three and a half times larger (204 mm.). Here the pattern deserves the specific name of vermicula- tus, the white being reduced, and divided into small irregular spots and lines. In Kumada & Hiyama’s plate of a ray of the same size (210 1941] 249 Beebe & Tee-Van: Rays, Manias and Chimaeras mm.), the pattern is again quite distinct. The white is still more reduced, and is faint, and in the form of a few meandering, irregular lines, most of which are on the outer margin of the disk and radiate outward. The tail bands are almost absent, but each unpaired fin still has a whitish line across it. The 4th individual we know only from Kuma- da’s brief mention of “other large specimen (2 feet).” In both Breder’s and our specimens the rims of the spiracles are smooth, and while Kumada & Hiyama mention “small tubercles” in their description, yet their plate, which is drawn in great detail, shows them smooth. Discopyge Tschudi, 1846. Key to tropical eastern Pacific species, la. A strongly marked, brilliant ocellus in the center of the disk ommata. lb. No ocellus in center of the disk tschudii. Discopyge ommata Jordan & Gilbert. Ocellated Electric Ray. (Plate I, Fig. 2). Range: Gulf of California to Panama (Mexico : Gonzaga Ba}q Puerto Refugio, Inez Bay, Santa Cruz Island, San Francisco Island, Arena Point; Costa Rica: Port Culebra; Panama: 10 miles south of Pearl Islands, Bay of Panama). Field Characters: Tail thick with two dorsals and a caudal fin; pelvics not notched on the outer edges; skin smooth everywhere. Provided with electric organs which generate sufficient power for adequate protection. Pelvic fins adnate to body. Color as below. (Illustration from speci- men 25,236, 142 mm.) Color: The color variations of six specimens are so extreme that no single description can cover all. This is not dependent on age, sex or locality and change of individual pattern can play but a slight part. The commonest pattern is a light brown dorsal background, covered thickly with small, well-defined black dots. In the center of the back there is always a strongly-marked ocellus. This usually has a black or yellow ochre or rufous core, surrounded by a succession of solid or broken rings of alternating black and pale, brown. Other ocelli up to fourteen are scattered about on the upper surface, some like the spots of a jaguar, or again they may be almost solid. The most common situations are at the posterior inner angle of the pectorals, a pair close together between the central ocellus and the first dorsal fin, and one at the base of each dorsal fin. The preorbital area is free from dots and is marked with black and yellow ochre in the form of W’s, M’s and Y’s. Below immaculate. Our Arena specimens are wholly free from dots, the skin being faintly marbled, while the central ocellus has a small, bright rufous center, then a wide ring of black, outside of which is a narrow pale one. This concentration of pigment is evident in the post-pectoral and dorsal fin ocelli. There, extreme patterns are exhibited in specimens no more than 105 and 142 mm. apart in length. A 65 mm. ray is thus described by Seale: “Mottled and marbled with brown over the back, with a distinct, round white spot sur- rounded by rings of black and white on the center of the back.” Breder has found the same ex- tremes of variation in specimens from the Gulf of California. Size: The largest recorded specimen is our ray from Inez which measures 185 mm. in total length. Local Distribution: These little rays have been taken not far from shore in 1 to 35 fathoms. Abundance: Common at night in shallow water in Inez Bay. Elsewhere it has been recorded singly or in pairs. Food: We found amphopods, small shrimps and worms in the stomachs. Also considerable quantities of very fine bits of quartz. Breeding: In our largest specimen, a female of 185 mm. length, taken on April 10, the organs were considerably enlarged. 250 [XXVI: 26 Zoologica: New York Zoological Society General Habits: At night, near shore, in the light of our flashes, these rays were clearly visible as they swam through mid-water or close to the bottom. Unharmed specimens, when picked up from the dredge hauls, gave forth an electric shock, especially if the fish was seized on eacli side of the ocellus, in the middle line of the body. The shock was sufficiently strong, if unantici- pated, to make one drop the fish. It is said to generate as much as twenty volts. Three shocks in succession were noticed, the third weaker than the others. The conspicuousness of the pattern suggests the possibility of its serving a function of warning to vertebrate enemies. Study Material: 11 specimens: Mexico: Inez Bay, 3 males, 2 females (25,774) 116 to 174 mm. April 13, 1936, speared at night in shallow water (one saved) ; Inez Bay, 1 male, 1 female (24,932) Col. Plate 31,168 and 185 mm. April 10, 1936, dredged, Station 141:D-1, in 7 to 9 fathoms; Inez Island, 1 male, (25,236) 142 mm. April 9, 1936, dredged in shallow water; Arena Bank, 1 male, 2 females (25,338) 105, 120, 155 mm. April 19, 1936, male dredged at Station 136:D-5 in 33 fathoms, females Station 136:D-6 in 35 fathoms. References: Discopyge ommata, Jordan, D. S., & Gilbert, C. H., in Jordan, D. S. & Bollman, C. H„ Proc. U. S. Nat. Mus., 12, 1890; 151 (original description, color; type locality, 10 miles south of Pearl Islands, Bay of Panama). Jordan, D. S., & Evermann, B. W., Fishes of North and Middle America, 1896; 78 (short description). Osburn, R. C., & Nichols, J. T., Bull. Amer. Mus. Nat. Hist., 35, 1916: 144 (Santa Cruz Island, Gulf of California). Meek, S. E., & Hildebrand, S. F., Marine Fishes of Panama, 1923: 74 (short description). Breder, C. M., Jr., Bull. Bingham Oceano. Coll., 2 (1), 1928: 8 (color variation, Gulf of California; Angel Island, Gonzago Bay, San Francisco Island). Beebe, W., Bull. N. Y. Zool. Soc., 39, 1936: 236 (figure). Beebe, W., "Zaca Venture,” 1938: 123, 300, fig. 10 (note on electric shocks, figure). Seale, A., Allen Hancock, Pac. Expecl., 9, 1940: 2 (color; Costa Rica: Port Culebra). Discopyge tschudii Heckel. Range: Reaches the middle of Peru at the edge of our province. Said to extend around Patagonia as far north in the Atlantic as the Rio Plata. (Peru: Between Huaclio and Chancay.) Description: Disk circular; tail less than half the total length. Mouth, eyes and spiracle in the anterior seventh of the total. Mouth small, protractile; teeth bands narrow; teeth minute, flat, inner margins acute angled. Eyes small. Spiracles much larger than the eyes, and a short distance behind; margins without papillae or projections. Gill apertures small, hindmost two behind the middle of the disk. Dorsals small, rounded, subequal; origin of first dorsal slightly in front of the hind margin of the pelvics; second dorsal reaching little behind the origin of the caudal. Pelvics broad, rounded, united behind the vent. Tail small, depressed, dermal folds prominent, posterior margin of caudal con- vex, supra-caudal portion of fin longer. (Illus- tration after Gunther, 1898, 143 mm.) Color: Dusky reddish-brown above, darker over the middle; dull whitish below. Size: The Peruvian specimen was 143 mm. in length. A male from the Atlantic measures 410 mm. Study Material: None. References: Discopyge tschudii, Heckel, J. T., in Tschudi, J. J., Fauna Peruana, Pisces, 1845: 32, PI. 6 (original description, type-locality between Huacho and Chancay), Steindachner, F., Zool. Jahrb., Suppl., IV, 1898: 332, Plate 21, fig. 15 (description, figure). Abbott, J. F., Proc. Acad. Nat. Sci. Phila., 1899: 329 (copy of original description). Garman, S., Plagiostomia, 1913: 303 (shortened translation of original description). Norman, J. R., Discovery Reports, Coast Fishes, 2, 1937: 11, fig. 3 (Record of several from Atlantic, Gulf of St. George, Argentina, figure). Family Rhinobatidae. Key to tropical eastern Pacific genera. la. Width of disk l}i to 1}4 times into the length of the disk; nostrils oblique; snout pointed, its angle 60° Rhinobatus. lb. Width of disk equal to its length; nostrils trans- verse; snout shorter, obtusely pointed, its angle 85° Zapteryx. Rhinobatus Linck, 1790. Key to tropical eastern Pacific species. la. A single rudimentary spiracular fold ..planiceps. lb. Two spiracular folds. 2a. Rostral ridges close together for their anterior halves productus. 2b. Rostral ridges separated for their whole length. 3a. Rostral ridges rather broad; back with regularly arranged slate-colored blotches glaucostigma. 3b. Rostral ridges narrow; back without blotches leucorhynchus. 1941] 251 Beebe & Tee-Van: Rays, Mantas and Chimaeras Rhinobatus planiceps Garman. Flat-headed Guitarfish. Range: Galapagos Islands and Peru. (Peru: Payta, Callao, Pacosmayo, Lobos de Tierra La Punta; Galdpagos Islands.) Field Characters: Flattened shark-like rays; snout narrow and elongate; tail thick and with two dorsals and a well-developed caudal fin; pelvics unnotched on outer edge; skin rough with scattered spines; nostrils oblique; a single, rudi- mentary spiracular fold. (Illustration after Garman, 1913, 448 mm.) Color: Light olive green above, with numerous symmetrically placed dark blotches, with rather vague outlines. The dorsals, caudal, and outer parts of paired fins with slight rufous tinge; white below. Size: The largest recorded specimen is 763 mm. in total length. Abundance: Apparently common along the Peruvian coast. Study Material: None. References: Rhinobatus planiceps, Garman, S., Bull. Mus. Comp. Zool., 6, 1879-1880: 168 (original descrip- tion, 21 specimens from Payta, Callao, and Galapagos Islands). Garman, S„ Proc. U. S. Nat. Mus., 3, 1880 (1881): 520 (description); Garman, S., Bull. Mus. Comp. Zool., 17, 1888-1889: 89, Plate 24 (description of lateral- line canal system). Jordan, D. S., & Evermann, B. W., Fish North and Middle America, 1, 1896: 64 (descrip- tion). Garman, S., The Plagiostomia, 1913: 283, Plate 17a (description, figure; Lobos de Tierra, Peru). Nichols, J. T., & Murphy, R. C., Bull. Amer. Mus. Nat. Hist.. 46, 1922: 504 (Pascasmayo, Peru). Anonymous, Bol. Mus. Hist. Nat. "Javier Prado," No. 9, 1939: 122, 125 (La Punta, Peru). Rhinobatus productus Ayres. Northern Guitarfish; Shovel-nosed Shark. Range: San Francisco south to Agua Verde Bay, Gulf of California. (Mexico: Cedros Island, Port San Bartholome, Turtle Bay, Magdalena Bay, Agua Verde Bay.) Field Characters: Flattened shark-like rays, snout narrow and elongate; a stout tail furnished with two dorsal fins and a caudal fin. Skin covered with fine shagreen, with rows of hooked spines down the back and tail, and a small cluster at the shoulder; nostrils oblique; two spiracular folds; rostral ridges approximated along anterior half. Brownish-gray above, with a series of round, slaty spots. (Illustration after Kumada & Hiyama, 1937, 464 mm.) Size: Reaches a length of over four feet. Abundance: Reported as common in Turtle Bay. Study Material: None. References: Rhinobatus producta, Ayres, W. O., MS., Girard, C. F„ Proc. Acad. Nat. Sci. Phila., 7, 1856: 196 (original description, Monterey, California). Rhinobatis productus, Osborn, R. C., & Nichols, J. T., Bull. Amer. Mus. Nat. Hist., 35, 1916: 142 (Mexico: Cerros I., Port San Bartholome and Agua Verde Bay). Rhinobatos productus, Wales, J. H., Copeia, 1932: 163 (Mexico: Turtle Bay, Magdalena Bay). Norman, J. R., Proc. Zool. Soc. London, 1926: 973, fig. 26 (description, figure, synonymy). Rhinobatus productus, Kumada, T., & Hiyama, Y., Marine Fish Pacific Coast Mexico, 1937: 18, Plate 50 (description, figure). Rhinobatus glaucostigma Jordan & Gilbert. Slaty-spotted Guitarfish. Range: Lower California to Ecuador (Mexico: San Bartolome Bay, Guaymas, Mazatlan; Ecua- dor: Bay of St. Helene). Field Characters: Flattened, sharlc-like rays; snout narrow and elongate; tail thick with two dorsals and a well-developed caudal fin; pelvics not notched on outer edge; skin rough with scattered spines; nostrils oblique; two spiracular folds; rostral ridges widely separated and rather broad; back with regularly arranged, slate- colored spots. (Illustration after Kumada & Hiyama, 1937, 382 mm.) Color: Brownish-gray above, with a series of round, slate-colored spots symmetrically ar- ranged. Pectorals and pelvic fins with pale margins; a dark blotch and some irregular dark patches below the snout. (Norman, Proc. Zool. Soc. London, 1926: 970.) Size: The largest published record is of a fish 762 mm. in length. 252 Zoologica: New York Zoological Society [XXVI: 26 Local Distribution: Sandy bottoms in shallow water. Abundance: There are exceedingly few records of this ray; at Mazatlan it is said to be very common on sandy bottoms. Study Material: None. References: Rhinobatus glaucostigma, Jordan, D. S., & Gilbert, C. H., Proc. U. S. Nat. Mus., 6, 1884: 210 (original description, color, comparison with R. productus and R. leucorhynchus, Mazatlan); Evermann, B. W., & Jenkins, O. P., Proc. U. S. Nat. Mus., 14, 1892: 132 (Guaymas, Mexico); Jordan, D. S., Fishes of Sinaloa, 1895: 387 (Mazatlan) ; Jordan, D. S. & Evermann, B. W., Fishes of North and Middle America, 1, 1896: 62 (de- scription, color, Gulf of California) ; Boulenger, G. A., Boll. Museiidella R. TJniversita. di Torino, 13, No. 329, 1898: 1 (Baie de St. Helene, Ecuador) ; Gilbert, C. H., & Starks, E. C.. Fishes of Panama Bay, 1904; 14 (comparison with leucorhynchus)', Garman, S., Plagiostomia, 1913: 282 (description, color, Gulf of California); Norman, J. R. Proc. Zool. Soc. London, 1926: 970, figure 23 (description, color, range, Mazatlan, figure); Breder, C. M., Jr., Bull. Bingham Oceano. Coll., 2, (1) 1928: 5 (comparison with leucorhynchus). Rhinobatus productus: Streets, T. H., Bull. U. S. Nat. Mus., 7, 1877: 55 (San Bartolome Bay, Lower Cali- fornia). Rhinobatus leucorhynchus: Jordan, D. S., & Gilbert, C. H„ Bull. U. S. Fish Comm., 2, 1883: 105 (name only, Mazatlan). Rhinobatus sp., Kumada, T., & Hiyama, Y., Marine Fish West Coast Mexico, 1937 : 18, Plate 50, B (short description, figure, west coast Mexico). Discussion: Kumada & Hiyama’s Rhinobatus sp. has been placed in the synonymy of this species, although there are points of disagreement in the descriptions. Rhinobatus leucorhynchus Gunther. White-snouted Guitarfish; Fiddlefish. Range: Mazatlan, Mexico, south to Guayaquil, Ecuador. (Mexico: Mazatlan, Tenacatita Bay, San Benito; Costa Rica: Gulf of Nicoya; Panama: Bahia Honda, Panama Bay; Colombia: Tumaco; Ecuador: Guayaquil). Field Characters: Flattened shark-like rays; snout narrow and elongate; tail thick, with two dorsals and a well-developed caudal fin; pelvics unnotched on outer edge; skin rough with scat- tered spines; nostrils oblique; two spiracular folds; rostral ridges rather narrow; back without spots. Color: Dark gray above, the snout and outer parts of disk paler, no spots or blotches. Lower parts pale, the distal part of snout dusky. Another fully adult specimen with ten, irregular, roundish spots scattered at random on the back. Younger individuals are said to be generally lighter in color, with a few pale spots scattered over the upper parts. Size: A female of 625 mm. has been recorded. Local Distribution: Sandy bottoms in shallow water. Breeding: Two young, about to be born, were taken from an adult ray, on April 9, in thirty fathoms, off San Benito in southern Mexico. Each measured 165 mm. in length. A ray 140 mm. long was taken in Tenacatita Bay, Mexico, on December 10. Study Material: Not taken by us. A single ray of this species which had been caught in the Gulf of Nicoya, was identified in the collection of the museum at San Jose, Costa Rica. Discussion: There seems considerable likeli- hood of identity between leucorhynchus and glaucostigma , as indicated by the following: Gilbert & Starks admit very close relationship between the two forms, but give three apparent differences: pattern, rostral ridges, shape of snout and size of posterior gill-slit. Stark says that the Ecuadorian specimen of leucorhynchus in snout, rostral ridges and disk outline is intermediate between a Panama speci- men and a Mazatlan specimen of glaucostigma. In leucorhynchus the dorsals are thought to be darker and the shagreen is rougher and coarser. Breder notes the irregularity of the pattern in his individual of leucorhynchus and adds, “This suggests the possibility of R. glaucostigma Jordan and Gilbert being identical with it or a race.” This can be satisfactorily settled only by a study of a reasonably adequate series of indi- vidual rays. References: Rhinobates leucorhynchus: Gunther, A., Proc. Zool. Soc. London, 1866; 604 (original description, Panama) . Rhinobatus leucorhynchus: Gunther, A., Trans. Zool. Soc. London, 6, 1869: 396, 490. (description of type. Pacific coast of Panama) ; Gunther, A., Cat. Fish. Brit. Mus., 8, 1870: 444 (description and illustration of head) ; Garman, S., Proc. U. S. Nat. Mus., 3, 1881: 517 (descrip- tion, Panama); Jordan, D. S., & Gilbert, C. H., Bull. U. S. Fish Comm., 2, 1882: 105 (Mazatlan, Mexico); Jordan, D. S., Proc. U. S. Nat. Mus., 8, 1886: 364 (Panama); Jordan, D. S., & Evermann, B. W., Fishes North and Middle America, 1896: 62 (description, color, Panama and vicinity); Gilbert, C. H., & Starks, E. C., Fishes Panama Bav, 1904: 14 (comparison with glaucostigma)-, Starks, E. C„ Proc. U. S. Nat. Mus., 30, 1906: 762, 763 (comparison with glaucostigma, Gulf of California, Panama, Ecuador); Garman, S., The Plagiostomia, 1913: 282 (description, color); Wilson, C., Ann. Carnegie Mus., 10, 1916: 58 (Tumaco, Colombia); Meek, S. E., & Hilde- brand, S. F., Marine Fishes Panama, 1923: 68 (descrip- tion, color, Mazatlan to Ecuador) ; Norman, J. R., Proc. Zool. Soc. London, 1926; 971 (description, illus. of head, type length); Breder, C. M., Jr., Bull. Bingham Oceano. Coll. 2 (1), 1928: 4 (color, comparison with glaucostigma; San Benito, Mexico; Bahia Honda, Panama). Rhincobatos leucorhynchus: Seale, A., Allan Hancock Pacific Exped., 9, No. 1, 1940: 2 (Tenacatita Bay, Mexico). Zapteryx Jordan & Gilbert, 1880. Zapteryx exasperata (Jordan & Gilbert). Rough-skinned Guitarfish. Range: San Diego and the Gulf of California south to Panama. (Mexico: Gonzago Bay, San Felipe Bay, Espiritu Santos Island; Panama: Panama Bay.) 1941] 253 Beebe & Tee-Van: Rays, Mantas and Chimaeras Field Characters: Flattened, shark-like rays; tail thick with two dorsals and a well-developed caudal fin; no notch on outer edge of pelvics; skin rough with scattered spines; color not uni- formly black; nostrils transverse; disk dark, with black-edged yellow spots, or strong transverse bands. (Illustration after Kumada & Hiyama, 1937, 485 mm.) Color: The two extremes of pattern and color are as follows; grayish-brown above, a band of dark brown near the ends of the nostril ridges; between this and another dark band which crosses the bases of the ridges, there is a light band; a band across the head between the eyes is somewhat confluent with the band in front of it, except for a dark spot on the posterior angle of each pectoral. In the second type of extreme pattern, the disk has several rough, yellowish spots as large as the pupil, each spot occellated with black. Size: Reaches a length of about three feet. Local Distribution: Shallow waters near shore. Study Material: None. Discussion: The variation in pattern and color removes all possibility of these being specific characters, and in all other respects xyster ap- pears to be indistinguishable from exasperata. References: Platyrhina exasperata, Jordan, D. S., & Gilbert, C. H., Proc. U. S. Nat. Mus., 3, 1881 : 32 (original description; type locality, San Diego, California). Syrrhina exasperata, Jordan, D. S., & Gilbert, C. H., Proc. U. S. Nat. Mus., 5, 1882 (1883): 621 (color, Pana- ma); Breder, C. M. Jr., Bull. Bingham Oceanogr. Coll., 2, 1928: 5 (Mexico: San Francisquito Bay, Gonzago Bay, San Felipe). Rhinobatos exasperatus, Jordan, D. S., Proc. XJ. S. Nat. Mus., 8, 1886: 364 (Panama). Zapteryx xyster, Jordan, D. S., & Evermann, B. W., Fishes North and Middle America, 1, 1896: 65 (original description, type locality, Panama) ; Norman, J. R., Proc. Zool. Soc. London, 1926: 980 (comparison with exasperatus). Brock, V., Copeia, 1938: 130 (Espiritu Santo Island, comparison with exasperata). Xapteryx xyster, Kumada, T., & Hiyama, Y., Marine Fish Pacific Coast Mexico, 1937: 19, Plate 52. Family Pristidae. Pristis, Klein, 1779. Key to tropical eastern Pacific species. la. Origin of first dorsal fin well in advance of the origin of the pectorals; lower lobe of caudal fin present, but small; 14 to 23 teeth along each edge of the rostrum zephyreus. lb. 24 to 32 teeth along each edge of the rostrum. ? pectinatus. Pristis zephyreus Jordan & Starks 1895. Sawfish. Range: Mexico to Ecuador. (Mexico: Mazat- lan; Guatemala: Chiapam; Costa Rica: San Juan del Sur; Panama: Balboa, Rio Grande at Mira- flores, Rio Chucanaque; Colombia: Rio San Juan; Ecuador: Guayaquil.) Field Characters: Shark-like rays with an elongate, depressed body; snout produced into a flat, very long rostrum, armed along each lateral edge with a series of 17 to 23 large teeth; lower lobe of caudal small; origin of first dorsal in advance of the pelvics. (Illustration from figure of closely related P. perotteti, after Day, 1888.) Study Material: No living individuals were observed or captured. A large rostrum (No. 28,723) of this species was obtained in San Juan del Sur, Costa Rica, Jan. 10, 1938, from a fisher- man, who had taken the sawfish himself. The snout measures 900 mm. from the tip to the origin of the proximal, lateral pair of teeth; at the latter place the width of the snout is 175 mm. and the length of the longest tooth is 45 mm. References: Pristis zephyreus, Jordan, D. S., & Starks, E. C., in Jordan, D. S., Fishes of Sinaloa, 1895: 383 (original description; comparisons of various descriptions of various authors; type locality, Mazatlan, Mexico; type in Stanford University). Gilbert, C. H., & Starks, E. C., Fishes of Panama Bay, 1904: 14 (amended de- scription, specimens from Rio Grande at Miraflores, Panama). Pristis microdon. Meek, S. E., & Hildebrand, S. F., Marine Fishes of Panama, 1, 1923: 66 (tidal streams near Balboa, Panama, 22 specimens, 800-1075 mm., descrip- tion). Breder, C. M., Jr., Bull. Amer. Mus. Nat. Hist., 57, 1927: 99 (Rio Chucanaque, Panama, 5 specimens, 770- 965 mm., notes on ecology). Breder, C. M., Jr., Bull. Bingham Oceano. Coll., 2 (1), 1928: 4 (refers to last mentioned reference, note on habits; under P. pectinatus). Pristis perotteti, Gunther, A., Cat. Fishes Brit. Mus., 8, 1870: 437 (Chiapam, Guatemala). Steindachner, F., Denkschr. Akad. Wiss. Wien., 42, 1880: 102 (fresh and brackish water around Guayaquil, Ecuador; two speci- mens, 790 and 870 nun.). Jordan, D. S., & Gilbert, C. H., Bull. U. S. Fish Comm., 2, 1882 (1883): 105 (Mazat- lan, Mexico). Wilson C., Annals Cam. Mus. Pittsburgh, 10, 1916: 58 (Guayaquil, Ecuador). Eigenmann, C. H., Indiana Univ. Studies, 46, 1920: 10 (Rio San Juan, Colombia). Eigenmann, C. H., Memoirs Cam. Mus., Pittsburgh, 9, 1922: 25 (Rio San Juan, Colombia). Pristis antiquourum, Gunther, A., Trans. Zool. Soc. London, 6 (7), 1868: 397 (Chiapam, Guatemala). Discussion: This species has been merged with P. microdon, and is so considered in many publications. Until good comparisons are made with materials from other oceans we prefer to maintain the Pacific coastal form as a separate species. ? Pristis pectinatus Latham. Sawfish. Range: Acapulco, Mexico is the only definite locality record within our area on the Pacific coast. 254 [XXVI: 26 Zoologica: New York Zoological Society Field Characters: Shark-like rays with an elongate, depressed body; snout produced into a flat, very long rostrum, armed on each side with 24 to 32 teeth. Study Material: No live fish observed or col- lected; two small rostrums were purchased at Acapulco, Mexico, Nov. 26, 1937. They measure 172 and 175 mm. from the tips to the proximal pair of teeth; width at this base 28 mm.; number of teeth 27 right, 27 left, and 26 right, 28 left respectively (Nos. 28,724, 28,725). References: 1 Pristis pectinatus, Latham, J., Trans. Linn. Soc. London, 2, 1794: 278 (“in the ocean.”). Breder, C. M., Jr., Bull. Bingham Oceanogr. Coll., 2 (1), 1928: 4 (no definite locality, west coast of Central America or Mexico; 2 specimens, 710 and 724 mm. and rostrum of a third specimen). Discussion: Further study may show these Pacific sawfish to be another species. Family Rajidae. Key to tropical eastern Pacific genera. la. Rostral cartilage produced forward from the skull, stiffening the rostrum Raja. lb. Rostrum soft, flexible, lacking the rostral pro- longation of the cranium Psammobatus. Raja Linnaeus, 1758. 2 * * * & Key to tropical eastern Pacific species. la. No spines present about the eye; 33 rows of teeth aguja. lb. Spines present about the eye; 37 to 44 rows of teeth. 2a. Three large spines about the eye, one in front and two behind (see illustration) badia. 2b. Usually more than three spines about the eye, the spines smaller in size. 3a. Three or four rows of spines between the eye and the edge of the disk; a spine on each shoulder; an obscure darkish patch near the base of the pectoral fin on the dorsal surfaces equatorialis. 3b. No rows of spines below the eyes; no tubercles or spines on the shoulders. 4a. A series of spines or prickles along the middle of the back, a few of the ante- rior ones being quite large; a dark ring on the upper surface near the base of the pectoral fin (Cedros Island, north- ward) inornata. 4b. A single median spine on the back; no dark ring near the base of the pectoral fin on the upper surface (Ecuador) ecuadoriensis. 2 Tills key should be used with great caution, as our knowledge of the tropical eastern Pacific forms of this genus is fragmentary. Two species of the genus Raja have been reported by Kumada & Hiyama from the “west coast of Mexico.” Whether these fish belong to the tropical eastern Pacific fauna or to that of the colder waters further north we do not know. Because of scanty material and lack of familiarity with the group, we do not attempt to give these specimens a name or to place them in the synonymy of some other form. We have included copies of Kumada & Hiyama’s descriptions under the heading of “Raja sp. A." and “ Raja sp. B.” Raja aguja Kendall & Radcliffe. Peruvian Ray. Range: Near Aguja Point, Peru (Lat. 5° 47' S., Long. 81° 24' W.), in 536 fathoms. Field Characters: Caudal fin absent or only a slight fold of skin; two dorsal fins crowded together near tail tip; pelvics notched on outer edge; skin rough; thirty-three rows of teeth. Color, purplish-brown, with or without scattered pale, good-sized spots. (Illustration after Ken- dall & Radcliffe, 1912, 480 mm.) Size: Grows to 480 mm. Study Material: None. References: Raja aguja, Kendall, W. C., & Radcliffe, L., Mem. Mus. Comp. Zool., 35, 1912: 78, 167, Plate 1, figs. 1 and 2 (original description, color; type locality: near Aguja Point, Peru). Raia aguja, Garman, S., The Plagiostomia, 1913: 358 (re-description of one of the types). Discussion: The female specimen in the original description appears as if it might be assigned to Psammobatis. Raja badia Garman. Range: Gulf of Panama. (Thirty miles east of Capa Mala, Lat. 7° 5' 30" North, Long. 79° 40' West), in 1270 fathoms. Field Characters: Ray with caudal fin absent, a low dermal keel on upper and lower sides of tip of tail; two dorsal fins crowded together near tip 255 Beebe & Tee-Van: Rays, Mantas and Chimaeras 1941] of tail; pelvic fin notched on outer edge; skin rough, with a single median row of large tuber- cles down back and tail, conspicuous, large spines about the eye. (Illustration after Garman, 1899, 257 mm.) Color: Chocolate brown above and below; narrow white areas about the mouth, and a white spot immediately behind the middle of the shoulder girdle. Size: Length of only known specimen, 257 mm. Study Material: None. Text-figure 13. Breeding: Two egg-cases which, as Garman writes, are "probably R. badia,” were taken in two hauls by the Albatross, on Feb. 24, at almost the same locality, about fifty miles south-west of Mariata Point, Panama, in 465 and 782 fathoms respectively. Garman figures one of them and describes them as follows: “The egg case itself, without the tendrils, is 64 by 90 mm. The horns are mutilated, evidently they were of consider- able length; their bases are stout and thick. Over the entire surface the case is covered with fine villi or pile, in longitudinal series, which though harsh to the touch, gives the appearance of a soft rich, black velvet. The second speci- men has a longer, closer pile.” References: Raja badia, Garman, S., Mem. Mus. Comp. Zool. 24, 1899: 22. Plate 6, figs. 1-3 (original description, color, egg cases: type locality, Panama Bay). Raia badia, Garman, S., The Plagiostomia, 1913: 357 (description, color). Raja ecuadoriensis nom. nov.3 Range: San Helena Bay, Ecuador. Description: The figure adjacent and the follow- ing translation are from Tortonese’s description and account. 3 Nom. nov. for Raja steindachneri Tortonese, 1939, not of Delftn, 1901. “ Raja steindachneri. Dell'. (Tav. 1). “Length 388 mm., disk length 224 mm. at the widest place, which is one-third wider than long. The line which connects the tip of the snout with the center of that which goes from one tip of the pectorals to the other is somewhat less than half of this latter. The line from the tip of the snout to that of one pectoral intersects the margin of the body slightly near the tip of the pectorals, and comprises one and a half times the length of the disk. The pectorals have convex angles, almost acute. The posterior ones are rounded. The anterior margin is slightly sinuous, forming on each side two slightly marked re-entrances which render it prominent in relation to the eyes. The pectorals reach to the middle of the ventrals, which are clearly bilobar and measure three- quarters of the length of the snout. The latter is long and acute, much straitened toward the tip which is rounded; the longitudinal diameter of the eye is comprised five and a half times in its length, the interorbital space three times. Rostrate angle of about 45°. The snout is a little less than a third of the length of the disk. The rostrate cartilage is narrow and robust, united in its distal three-fifths, forked at its base. The interorbital space is markedly concave. The orbits have small spiracula. The mouth is a little curved; its width is comprised one and three-fifths times in the pre-nasal part of the snout. 37/37 teeth, with a central triangular point. The nasal valves are fringed on the margin turned toward the mouth. Small pteri- gopodia; the free part measures 8 mm. The tail is depressed, with a longitudinal cutaneous fold on each side extending to the caudal one. The length of the approximately equal dorsal fins is a third of the snout ; they are separated by a space rather smaller than half the base of the first dorsal fin; the second is united with the caudal one, which is placed low, and double the length of the eye. 256 [XXYI: 26 Zoologica: New York Zoological Society “Nearly all of the body is smooth. Two spines, with some other minute ones, stand before the orbit, two placed right above it, and some very small ones above the spiracula. A little behind these, and in the middle of the back stands a single strong isolated spine. Near the outer angle of the pectorals run 8 unequal series of small spines, arched more or less parallel to one another, and parallel also to the margin of the body, extending farther backward than forward. The dorsal median line of the tail is occupied by a series of 17 robust spines inserted into elliptical and laterally compressed scutelli that are not very much unequal in their dimen- sions; two other spines are placed between the dorsals. A fine band of tiny spinules follows the ventral margin, characteristics which distinguish the male: its total length is 393 mm., of which 178 make up the tail, but the pterigopodia measure 118 mm. This fact, combined with other small morphological differences might lead to the supposition that the Ecuador race represents a heretofore unknown species. It is however, im- possible for me to ascertain this now, and since I am acquainted with R. steindachneri only through the treatise cited, I feel justified in the assumption that the specimen studied belongs to this species, observed until now near the coasts of Chile (type of Iquique) and Peru, but spreading also farther North. “The races, limited perhaps to a small number of forms, which live in the western regions of South America, are not yet well known; concern- ing them comparative studies of a certain breadth are still lacking.” Reference: Raja steindachneri (not of Delfln) Tortonese, E., Boll. Musei Zool. Anat. comp. Torino, 1939 (3) 47, no. 89: 3-5, Plate 1 (description, figure, Ecuador). Discussion: Tortonese’s description and figure are based, according to his text, upon Evermann & Kendall’s description of a ray4 that the latter assigned to Raja steindachneri Delfin, which in turn was based upon Steindachner’s Raja chilensis, 1896. A comparison of Tortonese’s figure with the original plate and description of Raja chilensis shows that the latter is quite properly placed in the synonymy of Psammobatus lima Poeppig, that Tortonese’s fish with its prominent rostral cartilage was correctly assigned to the genus Raja and that it bears no resem- blance to Steindachner’s chilensis. It is therefore necessary to give another name to Tortonese’s ray. Raja equatorialis Jordan & Bollman. Range: Gulf of Panama in 33 fathoms, 7 miles south of Pearl Islands. (8° 6' 30” N., 78° 51' W.) Field Characters: Rays without distinct caudal fin; two dorsal fins near tip of tail; pelvic fin notched at outer end; large spines present around eye; three prominent rows of spines or tubercles on the tail. Color: Light brown, spotted with paler, back 4 Evermann, B. W., & Kendall, L., Bull. U. S. Nat. Mus., 95, 19X7: 14. with obscure reticulations of the ground color, forming honey-comb-like markings surrounding paler markings; an obscure, roundish dusky blotch at middle of base of pectorals, and a darker one near their posterior base; edges of ventral and pectoral fins and the snout pale ; dark markings on interorbital area and below eye. No markings below. Size: The only known specimen is 14 inches long. Study Material: None. References: Raja equatorialis, Jordan, D. S., & Bollman, C. H., Proc. U. S. Nat. Mus., 12, 1890: 150 (original description, color; type locality, Gulf of Panama; type No. 41,132, U. S. Nat. Mus.). Meek, S. E., & Hildebrand, S. F., Marine Fishes of Panama, 1, 1923: 71 (description, color). Gilbert, C. H., Proc. U. S. Nat. Mus., 48, 1915: 308 (re-examination and description of the type). Discussion: This ray is related to Raja inornata, from which it is apparently distinguished by having four rows of spines below the eyes, a series of stout spines on each side of the tail in the male, and no prickles on the back except the median series and the spine on each shoulder. Considering the variability found in R. inornata, a comparison between the type of equatorialis and specimens of inornata would be advisable. However, Gilbert (1. c.) states in his reexamina- tion of the type of equatorialis that “R. equatori- alis is not very closely related to any other American species.” Raja inornata Jordan & Gilbert. California Skate. Text-figure 15. Range: Straits of San Juan de Fuca south to Cedros Island, Lower California. Field Characters: Disk broad, diamond-shaped, anterior margins undulating. Caudal absent, two dorsals near tail tip; pelvics notched; small 1941] Beebe I- < o u z I- I o U) > H l/) O OL Text-figure 1. Comparison of behavior between experimental and control specimen (shown in solid lines) together with the data of Breder & Gresser (1941) (shown in dotted lines) in terms of expectancy of random movement. 292 Zoological New York Zoological Society [XXVI: 28 SECTION A- A' Text-figure 2. Construction and details of experimental “cave.” A. Plan of pool and cave as constructed. B. Section through pool and cave. C. Light trap as seen from inside cave. D. Plan of light trap. a. Loose stones and gravel to form light trap and block fish exit at point of inflow to pool, b. Trap door to cave. c. Corked hole in trap door for feeding and temperature reading, d. Roof of cave composed of flooring on stringers covered with soil and planted with vines, e. Large Cyperus plant, f. Nyrnphaea. g. Nitella. panions. This plus the large areas involved apparently prevented the results disastrous to the blind individuals noted by Breder & Gresser (1941). At least at no time was anything seen that could have been interpreted as an attack on the blind by the normal eyed fishes. At times the addition of water would cause all to rush into the outflowing stream and enter the cave, where presumably they sought the blocked inflow from the other pool. These effects were variable and at first puzzling but it soon appeared that the causes lie in temperature differentials between the cave and the open pool. For example there was a marked tendency for all the fishes to be out in the open in the daytime and to retreat to the cave at dusk, both eyed and blind alike. During the daytime the open pool would warm under the influence of solar radiation and rise several degrees above the cave temperature while at night it would drop below that of the cave which because of its cover did not lose its temperature nearly so fast. A study of Table II will indicate how these varying seasonal and diurnal tem- perature changes effected the relationships of the water temperature in the cave to that in the open pool. The results may be summarized as follows: 1. If the temperatures were identical, or nearly so, the fish would move in and out freely, usually with the majority of blind fish out of sight and the eyed specimens in a compact school in the lighted pool. 2. A difference of as little as 1.5° F. between the cave and the open pool would cause prac- tically all to be in the warmer section, independ- ent of time of day or brightness of light. 3. Water entering from the other pool would have various effects dependent on whether it was warmer than that in the cave or not. If the cave was notably cooler than either pool the water first passing into the lighted portion and cooling it would first prevent the fish from enter- ing the cave and later as the cave warmed from the new water and the open pool cooled they would all enter, stimulated both by flow and temperature. Many of both types of fish were evidently gravid during the middle of summer but it is doubtful whether any spawning took place as the 293 1941] Breder & Gresser: Light Sensitivity of Mexican Blind Characin TABLE II. Temperatures and related data. Water temperatures °F. Date Hour Cave Pool Max. Min. May 23 9:00 p.m. — 73 — — 2 cave fish in pool 24 7:00 p.m. — — 70 64 25 7:00 p.m. — — 65 58 26 10:30 a.m. 62 69 — — 4:00 p.m. 60 67 — — 7:00 p.m. — — 69 57 29 8:00 p.m. — — 77 66 30 3:00 p.m. 66 67 72 62 10 cave fish in pool 31 8:00 p.m. — — 74 60 June 1 4:00 p.m. 63 62 — — 6:00 p.m. — — 66 61 2 8:00 p.m. 65 65 72 61 9 river fish in pool 3 8:00 p.m. — — 75 61 4 6:30 p.m. — — 70 61 July 14 3:00 p.m. 69 73 — — 18 8:00 p.m. — — 75 67 20 8:30 p.m. — — 78 63 22 7:45 a.m. 69 67 — — 8:00 p.m. 69 73 78 62 23 7:45 a.m. 70 69 — — 8:00 p.m. 74.5 77 82 67 24 7:45 a.m. 73 71 — — 8:00 p.m. 78 76 81 68 26 8:30 a.m. 75.5 75 — — 8:00 p.m. 75.5 78 82 71 27 10:30 a.m. 74 76 — — 8:00 p.m. 77 78 82 67 30 7 :30 p.m. 72 72 78 70 31 6:30 p.m. 72 72 71 70 Aug. 2 9:30 a.m. 73.5 73.5 77 69 7:00 p.m. 74 75 79 75 3 8:00 p.m. 72.5 73.5 79 66 16 9:00 a.m. 66 67 83 53 18 5:00 p.m. 64 66 73 53 28 2:00 p.m. 65 66 78 58 Sept. 5 9:00 a.m. 67.5 66 79 53 8 10:00 a.m. 66 63 70 56 1:30 p.m. 65 66.5 — — 4:00 p.m. 67 66 — — 13 2:00 p.m. 62 60 — — Removed fish, 1 cave, 4 river. Mean 69.2 + 70.1 75.3 + 62.9 + 69.1 + (average of all min. and max.) Maximum 78 78 83 75 Minimum 60 60 65 53 Note. Columns marked “Max. and Min.” read on a minimax thermometer at hour noted and reset for next reading. temperatures were probably below their spawning threshold. In any event there were no young evident, although either eggs or very young fish may have been eaten by the adults or by aquatic insects if such were actually produced. The experiment was terminated with the ad- vent of cool weather as it was evident that soon their lower threshold would be passed. As it was, some of the minimum temperatures reached and survived by these fish are rather remarkable, especially when it is borne in mind that the home waters of these fish hover about 80° F. Since the fish were variously in and out of the cave it was impossible to keep a close check on the actual numbers present. When the fish were removed on September 13 it was found that only five remained, one blind and four eyed. Less than two weeks previously many more than this number were seen when a warm day drew them out of the cave. It is suspected that with the cooling water and a failure of their alertness that they fell prey to frogs, as the most likely predator normally about these pools. No matter what the cause of their destruction, under the conditions of natural enemies, fluctuating temperatures, et cetera, it is clear that a marked differential of survival is present. Reduced to terms of per- centage, 10% of the blind individuals survived while 44+% of the eyed ones were still there when the experiment was terminated. Although it must be admitted that the total numbers are small, those that were recovered were in an excellent state of vigor and are now in aquaria. We have every reason to believe that the differ- ence in numbers has bearing on both the blindness and light color of the cave fish. Incidentally it seemed odd that the frogs did not catch the blind 294 Zoological New York Zoological Society fish early in the experiment, for although they knew their way around the pool they could not tell where a large frog might be floating and if both were present would sooner or later bungle into the frog’s dangling hind legs. At first this unaccustomed accident would cause the frog to give a typical fright reaction but soon the frogs became accustomed to such incidents and on occasion they would turn and act as though to pursue the blind fish. The safety of these fish seemed to reside in the fact that customarily when they do strike into something they ordi- narily charge off into another direction at top speed, and then not infrequently seem to be confused and are likely to strike the bottom or something else with which they are fully familiar and take some little time to quiet down again. Discussion. Since it has been shown that reactions to temperature gradients override reactions to light in both blind and visual types it may well follow that temperature differentials in a state of nature form an important influence on the entry and further penetration of these caves. There would thus seem to be at least three factors in the be- havior of these fish leading to cave entry. 1. Negative phototaxis on the part of blind fish and a tendency for visual types to hide in dark places and peer out. 2. Positive rheotaxis tending to cause these fishes to swim upstream. 3. Positive thermotaxis tending to cause these fishes to move into warmer waters. The first two requirements of the environment are met by La Cueva Chica and while at the time of our visit the subterranean waters were prac- tically of the same temperatures as surface waters, surely at other seasons the ground waters are warmer than those of surface drainage. During the dry season these ground waters should be cooler than these of the surface if they were not heated from below, a geologic feature of this region with its magnetic layers close to the surface and with an abundance of warm springs. Although the temperatures encountered in the field were considerably above those obtainable in the pool experiments there is no reason to suppose that the fishes’ behavior would be any different in regard to thermal preferences. Incidental to this is the fact shown by Doudoroff (1938) that a variety of marine species will select water warmer than that of their normal habitat, if given the opportunity. This finding of a temperature gradient reaction helps to account for the presence of fishes as remote from the cave’s exit as is evidently possible for them to reach, for presumably there is such a gradient from these cave springs, where they rise from the depths, to the juncture at the river, during part of the year at least, although in the short section of the cave available to human exploration no such gradient was detected. [XXVI: 28 The cave water temperatures taken in a wide variety of places varied scarcely at all from 80° and this quite erratically. In aquaria, the same fish appeared to be not in the least inconveni- enced by temperatures in the low ninety degrees although eggs laid in water of such temperatures failed to develop normally. Temperatures as low as 65° and for short intervals dropping to 53° have been survived. It thus appears that these fish have a rather wide temperature tolerance. Nevertheless, normally eyed river fish, hatched in an unfurnished, rectangular and well lighted aquarium with no opportunity to hide at all, voluntarily entered a simulated cave where their eyes were useless, on the stimulation of slight flow or a slight thermal difference. Blind fish of the fifth generation raised in light and without previous experience behaved the same way. Given fishes with three such items in their normal behavior, all that would seem to be necessary, in addition, to establish the observed condition in La Cueva Chica would be a genetic defect in- volving the eyes and cave factors of the types described. The differential in survival under the condi- tions of the experiment may mean but little in terms of the Mexican caves but since the experi- mental temperatures drew the blind fish out of the cave nearly daily it is tempting to imagine that, had it been possible to reverse conditions at will, a much larger number of blind individuals would have survived. Summary. 1. Light sensitivity in Anoptichthys is lodged in the apparently functionless remnant eye cap- sule, which may be demonstrated by its removal. 2. Reactions to temperature overrides the influence of light in both the eyed and blind forms and rheotaxis interferes to a lesser extent. 3. Eyed individuals apparently do not attempt to destroy blind ones except under conditions of close confinement. 4. Normal river fish, which never before had an opportunity, entered a simulated cave on temperature differentials of less than 1.5° F. as well as on a slight flow emerging from the darkness. 5. Survival in an outdoor pool divided about equally between light and dark was in favor of the eyed forms slightly better than 4 to 1 . 6. Three factors have been experimentally demonstrated as parts of the normal behavior pattern of both eyed and eyeless fishes, which would tend to make them enter caves similar to the one in which found, namely: negative photo- taxis, positive rheotaxis and positive thermotaxis. Bibliography. Breder, C. M., Jr. 1939. Variations in the nesting habits of Ameiurus nebulosus (LeSueur). Zoologica 24 (25): 367-378. 1941] Breder & Gresser: Light Sensitivity of Mexican Blind Characin 295 1940. The nesting behavior of Eupomotis gib- bosus (Linnaeus) in a small pool. Zoo- logica 25 (23) : 353-360. Breder, C. M., Jr. & Gresser, E. B. 1941. Correlations between structural eye de- fects and behavior in the Mexican blind Characin. Zoologica 26 (16): 123-131. Bridges, W. 1940. The blind fish of La Cueva Chica. Bull. N. Y. Zool. Soc. 43 (3) : 74-97. Doudoroff, P. 1939. Reactions of marine fishes to temperature gradients. Biol. Bull. 75 (3): 494-509. Gresser, E. B. and Breder, C. M. Jr. « 1939. The histology of the eye of the cave characin, Anoptichthys. Zoologica 25 (10): 113-116. Shlaifer, A. & Breder, C. M., Jr. 1940. Social and respiratory behavior of small tarpon. Zoologica 25 (30): 493-512. 296 Zoologica: New York Zoological Society [XXVI: 28 EXPLANATION OF THE PLATE. Fig. 1. The pool and “cave” as it appeared with the construction work finished but before the placement of soil and planting. Fig. 2. The pool and “cave” as it appeared shortly after planting. Later the vegetation became considerably more lush, hiding all of the woodwork. 8REDER & GRESSER. PLATE I. FIG. 2. FURTHER STUDIES ON THE LIGHT SENSITIVITY AND BEHAVIOR OF THE MEXICAN BUND CHARACIN. 1941] Crane: Crabs of the Genus Ocypode 297 29. Eastern Pacific Expeditions of the New York Zoological Society. XXIX. On the Growth and Ecology of Brachyuran Crabs of the Genus Ocypode.1 Jocelyn Crane Technical Associate, Department of Tropical Research, New Yoik Zoological Society. (Plates I— II. Text-figures 1-7.) [This is the twenty-ninth of a series of papers dealing with the collections of the Eastern Pacific Expeditions of the New York Zoological Society made under the direction of William Beebe. The present paper is concerned with specimens taken on the Eastern Pacific Zaca Expedition (1937-1938). For data on localities and dates of this expedition, refer to Zoologica, Vol. XXIII, No. 14, pp. 287-298. Also included in the present paper are results of observations made by the author on the Pacific coast of Panama during January, 1941.] I. Introduction 297 II. Summary of Important Points 297 III. Key to Immature Ocypode from the Eastern Pacific 298 IV. Crabs of the Genus Ocypode Taken by the Eastern Pacific Zaca Expedition 299 Ocypode gaudichaudii Milne Edwards & Lucas. 299 Ocypode occidentalis Stimpson 308 V. Bibliography 309 I. Introduction. The present paper is the third of a series dealing with the brachyuran crabs of the Eastern Pacific Zaca Expedition. My sincere thanks go to Mr. Templeton Crocker for the opportunity of making these collections and of studying the crabs in the field, to Dr. William Beebe for his unfailing helpfulness and for leave of absence making possible a special trip to Panama, and to Miss Janet Wilson for her excellent drawings. ■Contribution No. 632, Department of Tropical Re- search, New York Zoological Society. The references given under each of the two species discussed include the type description, Rathbun’s monograph (1917), and occurrence records which have appeared since the latter date. The collection is deposited in the labora- tory of the Department of Tropical Research, New York Zoological Society. II. Summary of Important Points. 1. Young crab stages of Ocypode gaudichaudii and 0. occidentalis differ from adults chiefly in the flatter, more narrow carapace, longer legs, larger eyes, more oblique orbits, smaller chelipeds and different coloration. In gaudichaudii the ocular stiles and truncate chelae begin to develop when the crab is less than half grown, the cara- pace being around 7 mm. in length when the first trace of stiles appears, and 10 mm. when the chelae tips begin to show change in form. In adult males the stile is only slightly longer than in females; this difference is not apparent in young crabs. 2. Adult gaudichaudii apparently alone of all the genus feeds habitually on microscopic organic matter in the sand, just as do crabs of the related genus Uca. In gaudichaudii, however, the tech- nique differs: first, a raking motion of the truncate chelae is substituted for the pinching of sand by the pointed, spooned chelae in Uca; second, the specialized endites of the first maxil- lipeds form sand pellets which drop of their own weight, instead of being wiped off as in Uca. Young gaudichaudii habitually, and adults rarely, prey on small crustaceans, insects and worms, and sometines add wood and seaweed to 298 Zoologica: New York Zoological Society [XXVI: 29 SAN LUCAS B. V' ARENA BANK . GORDA BANKS > MAZATLAN BANDERAS BA CHAMELA b\ CLARION ISL. TENACATITA B. MANZANILLO SIHUATANEJO ACAPULCO DULCE R.J* PORT ANGELES PORT GUATULCO SANTA CRUZ B.' , TANGOLA-TANGOLA B. ^ GULF OF FONSECA^" GULF OF (.LCAT..Z SAN JUAN DEL SUR-r?^ s PORT PARKER GULF OF DU MU RC I EL AGO Br V^'“/K (..im.) POTRERO GRANDE B.-} PORT CULEBRA- BRAXILITO B. - vy, \ t PIEDRA BLANCA • ToA PEyRL0 S' NICOYA BAHIA HONDA EASTERN PACI FIC E X P E D T 1 0 N S NEW Y 0 R X ZOOLOG ICAL SOCIETY S H 0 R E C OLLECTI NG STATIONS COIBA ISL. (.»*«»■•**• •*«») GORGONA GA L APAGOS IS. Af Text-figure 1. Shore collecting stations of the Eastern Pacific Expeditions of the New York Zoological Society. their diet. 0. occidentalis is a predator and scavenger. 3. 0. gaudichaudii is diurnal, occidentalis noc- turnal except when very young. 4. In contrast to the racing dodging, burrowing adults, the young of gaudichaudii run only a short distance when pursued, then flatten them- selves in the sand, where protective coloration aids them in escaping observation. 5. Also included in the present paper are descriptions of color in life, stridulation, habitat, burrows and daily schedule, with special emphasis on gaudichaudii. Abdominal appendages of the three American species, gaudichaudii, occidentalis and albicans, are illustrated. III. Key to Immature Ocypode from the Eastern Pacific. 0. occidentalis and 0. gaudichaudii, although so distinct in the adult, are superficially indistin- guishable up to a length of about 7 mm., when the ocular stiles and truncate chelae of the latter species begin to develop. However, they may be distinguished when alive by color differences, and when preserved by the proportions of the buccal frame, as shown from the following key. The smallest specimens studied were about 4 mm. in length. The length of the ischium of the third maxil- liped is measured along its external margin in a straight line, as far as but not including the 1941] 299 Crane: Crabs of the Genus Ocypode basal process; the breadth of the ischium is measured across the distal margin. The accom- panying diagram illustrates these points. la. Living crab with a spot of brilliant scarlet on middle of posterior gastric region, and others on one or more of the meri of the last three pairs of ambulatories. Breadth of ischium of third maxilliped 58 to 69 per cent of its length occidentalis, p. 308 lb. Living crab without scarlet spots. Breadth of ischium of third maxilliped 77 to 85 per cent of its length gaudichaudii, p. 209 IV. Crabs of the Genus Ocypode Taken by the Eastern Pacific Zaca Expedition. Ocypode gaudichaudii Milne Edwards & Lucas. Text-figs. 2; 4 A, B, C, D; 5 B, D, F, H; 6 B, D; 7E, F; PI. I, Fig. 1; PI. II, Figs. 3, 4. References: Ocypoda gaudichaudii Milne Edwards & Lucas, 1843, p. 26; 1847, pi. 11, figs. 4-4b. Ocypode gaudichaudii, Rathbun, 1917, p. 373; pi. 129, fig. 1; pi. 130, flg. 1; Boone, 1927, p. 267, fig. 96A; 1929, pp. 562, 580; Pesta, 1931, p. 174; Sivertsen, 1934, p. 19; Crane, 1940, p. 67; text-figs. 1-8. Range: Gulf of Fonseca, El Salvador, to Chile; Gal&pagos Islands. Local Distribution: Found on more or less pro- tected beaches and on the shores of lagoons. In one locality (Bahia Honda, Panama), there were more than 3,000 crabs on a beach measuring about 175 X 400 feet. Very rarely found on the same beach with 0. occidentalis, and then usually one or the other species occurs in the young stages only, so that its presence is doubtless accidental. Size: The 55 specimes in the present collection measure between 4.8 and 29 mm. in length of carapace. One of Sivertsen’s (1934) males from Galdpagos was 36 mm. long, and a female 34 mm. Sexual Characters: Except for the probably larger maximum size and slightly longer ocular stiles of the male, there are no apparent secondary sexual characters. The form of the abdominal appendage in this genus, as in Uca, varies among the species, so that it should prove a valuable taxonomic character in the determination of dubious forms. Plate II, Figs. 4, 5 and 6 illus- trate the appendages of the three American species. Color: This species ranges in general color from brilliant coral red to dark brown, marbled or mottled with brown, yellow and yellow-green. The exceedingly variable color is affected by a number of factors, including the color of the sand upon which they live and the sex and size of the individual. No precise observations have yet been made on the influence of light and tempera- ture on the one hand, or of physical and emo- tional states on the other, but these will un- doubtedly prove to be of great importance, and the crab found to be capable of considerable individual variation. The only exact notes taken on the latter subject concerned an oviger- ous female: the sides of the carapace were gray- violet when the crab was captured on a dark, stormy day after a long chase over light sand; they changed to moderately bright coral red after two days in a box kept partly in sunshine, its bottom being covered with the same light sand. There was no color change at night. Compare especially the observations of Cowles (1908) on color change in O. albicans, Abramo- witz & Abramowitz (1940) and Brown (1940) on the effect of eyestalk removal in Uca, and Crane (1941) on color change in courting Uca. Similar experiments and observations made upon O. gaudichaudii should yield interesting results. In regard to the effect of the environment, it was obvious that crabs on light sand were brightest, with a great deal of orange, carnelian and coral pink, while those on dark, volcanic sand were darkest, with these brilliant colors replaced by browns, rusty oranges and pale pinks. In any group the adult males were brightest, the females paler, the young palest. In specimens less than about 15 mm. long, the parts that were yellow in adults were white, those that were coral or carnelian red were lavender or violet; especially noticeable in the young was a band of violet across the posterior part of the carapace. Colors of adult males have been noted in detail as follows : ground color of carapace ranging from bright coral or carnelian red in colonies living on the palest yellow sand, through coral pink and rusty orange, to dark brown in colonies on dark slate-colored sand; carapace with mottlings or marblings ranging from pale lemon yellow or chartreuse with dark brown to ochre or char- treuse with cream and white; sides of carapace brilliant clear orange to scarlet orange, sometimes mottled with pink. Eye sockets brilliant orange to dark brown; stalks coral red to pearl gray; eyes pearl gray; stiles scarlet orange to yellow- brown. Merus and carpus of chelipeds orange or orange-brown above with brown mottlings, buffy yellow below and on inner surface; upper third of manus bright coral pink or orange to dark brown; lower two-thirds pinkish-white to lemon yellow; chelae coral pink, orange or lemon yellow. Posterior (dorsal) surface of ambula- tories usually like carapace, but dactyls are buff when crab lives on light sand, olive green when on dark sand; anterior (ventral) surface of am- bulatories usually pale. Sternum and abdomen bright coral pink, sometimes mottled carnelian red, yellow and white, sometimes fading into white posteriorly. Early Crab Stages: Several excellent studies on the growth of Ocypode have already been pub- 300 Zoologica: New York Zoological Society [XXVI: 29 Ocypode gaudichaudii. Growth stages. Carapace lengths: A, 4.8 mm.; B, 7.2 mm.; C, 10 mm.; D, 29 mm. lished, notably those of Cott (1929), Huxley (1931) and Sandon (1937). It seems worthwhile, however, to summarize and illustrate the changes occurring during growth in gaudichaudii, es- pecially since this species is more specialized than others studied ( ceratophtlialma and aegyptica). The changes in gaudichaudii were studied from a series of 64 specimens ranging in length from 4.8 to 29 mm., the carapace being measured in the median line. The outstanding growth char- acters of the young are as follows, all of the ob- servations being of course relative to the size of the crab. 1. The carapace is broader than in the adult, 1941] Crane: Crabs of the Genus Ocypode 301 D Text-figure 3. Ocypode occidentalis. Growth stages. Carapace lengths: A, 4.2 mm.; B, 6.9 mm.; C, 10 mm.; D, 25 mm. breadth being greatest when the crab is about 7 mm. long. 2. The carapace is less deep. 3. The orbits are more oblique. 4. The eyes are enormous, being broader and thicker as well as longer. 5. The eyestalks are longer and thicker, but there is no trace of stiles until the crab reaches a length of about 7 mm. In their early stages, the stiles are equipped distally with a few hairs, as in adults of certain other Ocypode. The stiles will be further discussed below. 302 Zoologica: New York Zoological Society [XXVI: 29 Text-figure 4. Major chelae of Ocypode. A to D, inch, gaudichaudii: A, female, carapace length 4.8 mm.; B, female, carapace length 7.2 mm.; C, male, carapace length 10 mm.; D, male, carapace length 29 mm. E, F, occidentalis: E, female, carapace length 4.2 mm.; F, male, carapace length 25 mm. 6. All the joints of the chelipeds are more slender, the width of the manus being only three-fourths that of the grown crab in a speci- men 4.8 mm. long. 7. The chelae show the first signs of truncation when the crab reaches a length of about 10 mm. 8. A stridulating ridge is present even in the smallest (4.8 mm.) crabs, but is composed only of minute, homogeneous granules, instead of having parallel, elevated lines in its upper half. In these smallest examples the inside of the palm is slightly hairy. 9. The ambulatories are relatively longest when the crab is about 7 mm. in length. 10. The colors of the young have already been noted (p. 299). 11. The crabs appear, anatomically, to reach maturity at a length of around 22 to 24 mm. ; this point cannot, however, be settled until more is known of their breeding habits. The only ovigerous female taken measured 27.5 mm. in length. Stiles: A number of writers (for example, Ortmann, 1894, p. 768, Lanchester, 1900, p. 759, and Gordon, 1934, p. 9) have mentioned the fact that in other species of Ocypode furnished with ocular stiles the latter are absent in the young and reach their maximum development in large 1941] 303 Crane: Crabs of the Genus Ocypode males. In. 0. ceratophthalma, stiles do not begin to develop much before a length of 20 mm. is reached (Lanchester, ibid.), in contrast to gaudichaudii , in which, as has been said, stiles are first visible at a length of about 7 mm. In gaudichaudii, too, the sexual difference in stile length is relatively slight, and is not apparent at all until the crab measures about 20 mm. In- deed, in smaller crabs, the stiles on the average are even slightly shorter in males than in females. In both sexes, the stile grows more rapidly after the crab has reached a length of about 20 mm. The following table is based on a series of 31 males and 23 females. Length of carapace in mm. 4 . 8 to 6.6 7 to 14 15 to 19 20 to 24 25 to 29 Length of stile in mm. M ales Females Absent 0 . 2 to 1 . 2 1.1 to 1.9 2.4 to 4.3 4.6 to 7.5 Absent 0.1 to 1.0 1.3 to 2.4 2 . 4 to 3 . 5 3.2 to 5.5 Largest male, 29 mm. long, stile length 7.5 mm. Largest female, 27.5 mm. long, stile length 5.5 mm. Dembowski’s detailed description (1913) of the glandular organs of the ocular stile and stalk of ceratophthalma has apparently not been super- ceded, while Parenzan (1931) ascribed to the stiles of the same species a tactile and protective function in addition to their glandiferous capacity. It will be interesting to study in detail the functions of both stiles and glands, and to determine whether the latter control chromato- phore expansion, as do the sinus glands in the eye-stalks of Uca and other crustaceans (see for example Abramowitz & Abramowitz, 1940; Brown, 1939; and Kleinholz & Bourquin, 1941). Food, Feeding and Mouthparts: Unlike other members of the genus, which are confirmed predators and scavengers in the usual sense of the word, gaudichaudii, except when very young, feeds almost entirely on microscopic organic matter washed onto the sand by the tide, exactly as does Uca. As in the latter genus, the minute particles of animal and vegetable matter are gathered by sifting pinches of sand through the mouthparts and dropping the remaining sand in small pellets from the posterior part of the buccal region. In feeding, alternate claws are used, at the rate of four scoops to the second; pellets emerge at two to the second. Rejected pellets of large crabs measure on the average one- quarter to three-eighths of an inch in length and are broadly oval. The details of the method by which the food is gathered, although analogous to the procedure in Uca, are in 0. gaudichaudii unique. In Uca the minor chelae are spooned, and literally pinch up clawsful of sand, the dactyl being widely extended before every pinch. In 0. gaudichaudii, the sand is not picked up between the chelae, but instead the latter are held almost closed, and used simply as broad-toothed rakes, scraping backward, inward and upward. The inner surfaces of both dactyl and pollex are slightly concave, and this characteristic, com- bined with their truncate tips, must make them almost or quite as efficient sand-gatherers as the entirely different chelae of Uca. Each scoop of sand, as in Uca, is swiftly lifted to the anterior end of the buccal cavity, where it is swept from the chelae by the palps of the third maxillipeds. No way has yet been found of watching exactly how the organic matter is separated inside the mouth, although as in similar observations on Uca the parts can be seen to be in rapid motion. Whatever the details of manipulation in gaudichaudii, it is apparent that the sand about to be rejected is formed into a pellet in the middle of the lowest (most posterior) part of the buccal cavity, between the enlarged endites of the first maxillipeds. The pellet is not wiped off by the chelae, as in Uca, nor does the sand simply ac- cumulate outside haphazardly and drop by means of gravity; instead, it appears to be carefully moulded into a pellet, by the up-and-down mo- tion of the endites, and dropped. Since damp sand lacks much of the stickiness of mud, such a procedure is practicable in this crab, whereas it would be impossible in most Uca, which are largely mud-feeders (cf. Crane, 1941). The technique of separation is not perfect, since small amounts of sand are usually found mixed with the organic matter in the alimentary canal. A comparison of the mouthparts of gaudi- chaudii with those of occidentalis is enlightening. The latter in feeding habits is a typical Ocypode, the west coast analogue of albicans, and feeds on the usual assortment of seaweed, carrion, am- phipods, insects, etc. I have never seen it put sand into its mouth, and have rarely found grains in its stomach. In gaudichaudii, the entire mouth field is much larger than in occidentalis, ranging from 25 to 30 per cent, more in both length and width, and with the parts correspondingly larger, a condition which is apparently useful in manipulating large quantities of sand. In gaudichaudii, too, the hairs on the mouthparts are much longer and thicker, although not as long as in mud-living Uca' s, where even more straining work is neces- sary than in sand-living Ocypode. Specifically, the hairs on the inner margins of the third maxil- lipeds are up to more than three times as long, and those on the second maxilliped one and a half times to twice as long. The maxillae are also larger and tougher and the mandibles slightly longer. There are no pectinated spoon- tipped hairs, so characteristic of most fiddler crabs, on the second maxilliped of either species of Ocypode; in both, however, there is a row of short, thickened hairs on the inner (dorsal) side of the merus and a cluster of similarly thickened and slightly concave hairs among the normal ones on the tip of the palp; these stout hairs are slightly longer and more numerous in gaudichau- dii than in occidentalis. The anterior margin of the sternum is broader in gaudichaudii, squared off and slightly concave, perhaps for better re- tention of the sand during manipulation, while 304 Zoologica: New York Zoological Society [XXVI: 29 Text-figure 5. Mouthparts of Ocypode. A, mandible, occidentalis; B, same, gaudichaudii; C, first maxilla, occidentalis; D, same, gaudichaudii; E, second maxilla, occidentalis; F, same, gaudichaudii; G, first maxilliped, occidentalis; H, same, gaudichaudii. All X 4.1, from cf’s 25 mm. long. 1941] Crane: Crabs of the Genus Ocypode 305 Text-figure 6. Mouthparts of Ocypode. A, second maxilliped, occidenlalis ; B, same, gaudichaudii; C, third maxilliped, occidentalis; D, same, gaudichaudii. All X 4.1, from cf’s 25 mm. long. in occidentalis this area is narrower, sloping and convex. The most interesting of the differences, how- ever, lies in the development of the proximal endites (basipodites) of the first maxillipeds. In Ocypode as in Uca there is considerable space between the anterior tip of the sternum, which extends upward between the maxillipeds, and the mandibles. In gaudichaudii, however, there is twice as much as in occidentalis, in which the extra space is negligible. In gaudichaudii this extra space is occupied chiefly by the proximal endites of the first maxillipeds— which, instead of lying almost useless on each side of the sternal projection, with their fiat sides turned backward against the sternum (as in grapsids, portunids, etc.), are in gaudichaudii and Uca enlarged, thickened, and their flat surfaces closely opposed to each other, exactly like a pair of butter-ball paddles. The distal endites (coxopodites) and the maxillae, in normal position, thus support and strengthen them laterally, instead of being anterior to them. In occidentalis the general form and position of the proximal endites are almost the same, but they are so small and weak, and there is so little space for them th^t they obviously could not function as paddles. An examination of a series of alimentary tracts combined with observation in the field, showed first, that feeding habits changed with growth and, second, that large crabs occasionally varied their planktonic diet with other items. Twenty-one stomachs and intestines were ex- amined, from crabs measuring between 5 and 29 mm. in length, taken in eight different localities. This series divided itself naturally into two parts, 306 Zoologica: New York Zoological Society one including crabs less than 9 mm. in length, and the other of those longer. The seven ali- mentary canals comprising the first group con- tained no trace of sand, and the contents con- sisted entirely of amphipods and tiny insects, including beetles, in an easily recognizable state. Obviously at this stage, in which the chelae have not become truncate, the crabs are chiefly predators — just as they were in the megalopal stage — and, doubtless, scavengers as well. Similarly, their food is identical with that of other Ocypode from the young of which, physi- cally, they are scarcely to be distinguished. After a length of 9 or 10 mm. has been attained, however, the crabs commence feeding almost entirely by means of sifting plankton from the sand. Correlated with this is a concomitant development of the truncate chelae. As was to be expected, the stomach contents of this second group, numbering 14 crabs, consisted of particles of organic detritus, including diatoms and minute algae. Ten stomachs included more or less sand in addition. In only four was non- planktonic matter found, and in these sand was present in the intestine, showing that the un- usual food was not habitual : the stomachs of two adults from the same locality held isopods and bits of wood; another had worm spicules and pieces of seaweed ; and a fourth held a small crab, probably Sesarma. These must be regarded as exceptional; several times I have seen large colonies of gaudichaudii scooping up the sand, while a dead fish or bird lay untouched in their midst, although hermit crabs busily feeding on the carcass attested its worth as food to true scavengers. I have never seen a member of this species behave as reported to Miss Rathbun (1917, p. 374) by Tristan, although I have baited them with a number of objects, edible and inedible, including both animal matter and seaweed: “If anything falls near them,” wrote Tristan, “they jump at it with extraordinary rapidity, as a spider in its web, and try to secure it with their claws so as to carry it off quickly to their holes. ” Similarly, I have never found material of any kind in their holes. The only mention I find of other Ocypode feeding in the manner of gaudichaudii is that of Takahasi (1935), who reports that the young of ceratophthalma and cordimana feed on plankton and make “sand-balls,” along with Uca, Sco- pimera, Ilyoplax and Mictyris. It is interesting that the young of gaudichaudii are, in contrast, predators. Burrows: The burrows are of three main types, and I have not been able to discover that each type is the work of a certain size or sex, or dug in a given sort of locality. The most usual form made by adults seems to be a burrow which goes 6 to 12 inches straight down, then turns at a right angle or more and continues for an almost equal distance. Other individuals, however, dig straight down for six inches, and then continue in a slow spiral. Still others dig a simple, oblique [XXVI: 29 burrow about nine inches long, ending about six inches underground. To discover which, if any, is the normal type, plaster casts must be taken in a number of localities. The method of digging agrees in general with the method used by other Ocypode, the sand brought up being carried by the first and second ambulatories on the side of the minor cheliped (the side which usually enters the hole first and leaves last). The minor cheliped is laid flat against the load of sand, bearing no weight, but holding it in place. Small loads may be taken two or three feet from the hole and flicked six inches still farther, being flung sideways and for- ward through the legs and under the cheliped of the other side. Large loads, however, which in large crabs are the usual kind, are simply dumped close to the hole. In Panama City I noticed a habit which had not been apparent in crabs farther north. When the load has been released near the hole, the crabs deliberately stamp the sand flat with the outer sides of the palm of the chelipeds rapping alternately and quickly against the ground. Simultaneously similar motions are made by the dactyls of the first two pairs of ambulatories, which, instead of bearing the weight of the crab on their tips in the usual fashion, are turned in- ward, so that the weight rests on the dactyls’ entire length, and more pounding surface is given. By just a few of the patting motions the heap of sand becomes almost indistinguishable from its surroundings; on the way back to the hole after an absence the crab usually continues the process, apparently to take care of any stray slippings. Also, they frequently stamp in the immediate vicinity of the hole even when they have not brought up a load. The procedure was exactly similar to that ob- served in male U ca in preparing a smooth display ground, free from pellets and digging lumps alike, in front of its burrow. All the instances were seen on rather dry sand in January. No evidence of courting activity was seen, although it may be that this action is connected with it. It was not seen in equally detailed observation of the species at Bahia Honda, in northern Panama, in March. It may, of course, be connected only with the type of sand, but it is worth noting that it may prove to be a definite part of courtship. Among these stamping crabs a number of rudi- mentary fights were seen; chelae were never locked as in Uca, but were used only in very brief sparring matches. In one instance, a large crab, the sex of which was not determined, stamped down a fresh load carefully, then darted eighteen inches away to the newly started hole of a smaller crab. The latter had brought up three or four loads and dumped them in as many different places, with- out stamping them down at all. The first crab stamped down one of the second crab’s loads, then chased the animal several feet, watched it keep on running, then returned to the newly started hole and deliberately kicked sand into it 1941] 307 Crane: Crabs of the Genus Ocypode until it was almost indistinguishable from the surrounding beach. The aggressor then returned to its own hole and resumed repair work. Defense: Crabs of this species employ four distinct types of defense mechanism, according to their size. The smallest noticed, measuring less than 8 mm. in length, are perfectly matched to the sand in color, being finely and evenly marbled with buff and black on carapace and legs. They take every advantage of the simi- larity in color, since they run only a few feet, or dodge a few inches, when pursued, then flatten themselves and remain motionless on the open sand. These smallest crabs were not seen to escape down holes, even when they were at hand, and apparently had none of their own; the smallest crabs found digging burrows measured between 7 and 8 mm. in length. If a moderate-sized crab — that is, one too small to have any orange developed dorsally — is dug up and, its hole having been destroyed, allowed to escape, it will, if pursued with moder- ate force, try to take refuge in a footprint or any other kind of depression in the sand, where it flattens itself and blends in color exceedingly well. However, if still pursued, it will descend any available hole. Large crabs when pursued never depend on flattening in a depression to escape notice, which would, of course, be impossible thanks to their usually striking color, Instead, they run swiftly, in typical Ocypode fashion, to their own hole if possible, or, if not, to that of any neighbor; if this is impossible they dodge and double on the open sand, or seek the shelter of nearby rocks. None of this species, of any size, was seen to take to the water except for a single ovigerous female, which was apparently aerating her eggs. Large crabs at bay employ the fourth means of defense, which, of course, is the assumption of a threatening attitude, with the nippers upraised and open, as the crab makes every attempt to pinch the attacker. Stridulation: Alcock’s (1892, p. 336) and Anderson’s (1894, p. 138) observations on the use of stridulation apply equally well to this species. Whatever its other uses, if any, it cer- tainly is a means of warning trespassers that a burrow is occupied. More than a dozen crabs of both sexes were dropped into the holes of as many others. In each case, when the ear was bent over the hole, distinct twittering squeaks were heard, especially at first. However, if I remained in position, apparently giving evidence of my presence through the shading of the hole, the two crabs always seemed to come to some sort of compromise in regard to sanctuary, the squeaking ceased, and the trespasser remained down the strange hole indefinitely. Naturally, strange large males caused the most commotion. The usual result, as found by subsequent dig- ging, is for the newcomer hastily to dig itself a little niche in the side of the tunnel, the displaced sand forming a loose plug in the mouth. The scrape of the bodies against each other and the digging sounds can be heard, quite distinct from the stridulation. Although the latter may be imitated artificially with a crab held in the hand, it is higher pitched when performed naturally in a burrow. Presumably only the rightful owner stridulates, but this point has not been checked, as it could perhaps be in a glass-sided terrarium. If the observer retires promptly to a distance of about twelve feet, the trespasser will emerge within two minutes, and pause at the mouth for several minutes more until certain that the danger is past. Miss Rathbun (1917, p. 374) quoted a note from Tristan regarding this species as follows: “ ... at low tide the crabs begin to come out sideways, remaining on the edge of the holes and hiding very quickly at the slightest noise.” Although they are extremely sensitive to move- ment, my own observation and experiments in- dicate that noises (excluding, of course, their own stridulation) have absolutely no effect on them. Daily Schedule: The following notes are based chiefly upon observations at Bahia Honda, Panama, on a large colony living on light sand. The beach is close to the mouth of the bay, but protected from the full strength of the surf by an island close to shore. The holes as usual were all in the upper third of the beach, and covered at high tide. These crabs are decidedly diurnal, and their daily routine is similar whatever their habitat. An individual, upon emergence from his hole near high tide line as the water recedes, is ap- parently very tired, and sits beside its hole, obviously resting, leaning first on one side, then on the other. At the slightest disturbance, even the close approach of a neighboring crab, it will reenter its hole. However, hermit crabs some- times come up and touch a recently emerged Ocypode without the latter’s paying the slightest attention. After some minutes’ rest, during which some of the clinging sand has dried and fallen off, the crab cleans itself thoroughly, be- ginning with the polishing of the eyes with the palp of the third maxilliped. Not until almost an hour has passed do the largest crabs start toward the edge of the tide to feed. At first they walk slowly, then more rapidly, and at last gain their usual racing gait. Feeding is carried on just in advance of the tide for an hour or more. Then the crabs return to the upper beach. (It is not yet certain that they return to the same holes they occupied during the preceding high water.) In any case each crab finds itself a burrow, which it repairs or enlarges according to its needs, or, rarely, it digs a new one. This work is punctuated with periods of feeding close to the hole. Sometimes there is still time after burrow repairs to feed at a distance again. Then the crabs gradually re- tire to their burrows, usually pulling in a plug of sand after them, until, fifty minutes to an hour before high tide, not a crab is left on the beach. An interesting relationship was noted at Honda between the crabs and a nocturnal hermit 308 [XXVI: 29 Zoologica: New York Zoological Society crab, not yet identified. This crustacean often was seen to come down from the jungle and enter the recently vacated hole of a beach crab, pull in a plug of sand, and spend the rest of the day, including, apparently, the succeeding high tide, if darkness did not fall before. Ocypode returning from the lower beach to find holes were never seen to pay the least attention to these newly occupied burrows, and no crabs appeared to be lost, or searching, in their vicinity. It may be that Ocypode, on broad beaches such as this, where they feed far from their holes, do not have a sense of property in regard to their burrows, once they have abandoned them after high tide, whereas crabs on narrow beaches, never straying far, may inhabit the same holes day after day. 0. gaudichaudii, as has been said, is active only during the day. In Panama City, on two nights, one brilliantly moonlit and the other dark, visits were made to the quite deserted beach where these crabs were numerous during the day. Although the holes had been opened by the crabs since the high tide (which fell after dark in each case), still not a crab had been feeding, as shown by the complete absence of pellets. The crabs themselves were usually found to be near the top of their sloping burrows, apparently asleep, since they were dug out with- out difficulty, whereas in daylight they always had to be dug from the very bottom of the long burrow. The following is a typical time table of their daily activities, worked out from several days’ observations on the broad beach at Bahia Honda. Sporadic observations made elsewhere agreed with these. Hours After High Tide V* 1 2 2-6 Hours Before High Tide 3-1 1 First young crabs emerge. (Since holes are not deep, tide leaves them first.) First large crab emerges. First large crab goes down to water’s edge. Migration of crabs to water’s edge; the largest last, since they have deepest holes. Most crabs redigging holes and feeding intermittently around holes and at water’s edge. Last crab in hole. Largest are last to descend. Obviously, this program can be carried out without postponement, interruption or curtail- ment due to darkness only when high tide is in the early morning or late afternoon. At Panama City, where there are exceptionally high tides, some members of a colony inhabited burrows which were not covered by neap tides. These individuals apparently spent most of their time quiescent in their holes during these periodic “droughts.” More study is needed on this subject. Material: In addition to the zoeae and mega- lopa recorded in a previous paper (Crane, 1940, 2 pp. 67, 70), a total of 55 specimens of 0. gaudi- chaudii was taken by the Eastern Pacific Zaca Expedition in the following localities: Mean- guera I., Gulf of Fonseca, El Salvador (Cat. No. 37,677); Corinto (37,736) and San Juan del Sur (3838), Nicaragua; Port Parker (3857), Port Culebra (38,141), Piedra Blanca Bay (38,198), Uvita Bay (38,442), Golfito (38,529) and Parida Island (38,651), Costa Rica; Bahia Honda (38,682), Panama; Gorgona Island (38,838), Colombia. In addition, 1 1 specimens were taken in Panama City, Panama, in February, 1941 (Cat. No. 4150). Ocypode occidentalis Stimpson. Text-figs. 3; 4 E, F; 5 A, C, E, G; 6 A, C; 7 A, B; PI. II, Fig. 5. References: Ocypoda occidentalis Stimpson, 1862, p. 229. Ocypode occidentalis, Rathbun, 1917, p. 372, pi. 129, figs. 2, 3; Rathbun, 1923, p. 632; Boone, 1929, pp. 562, 580, fig. 16; Crane, 1940, p. 70, text-figs. 3-8. Range: Turtle Bay, west coast of Lower California, to Peru. Local Distribution: Almost always found on unprotected beaches which are beaten by heavy surf. Occurs rarely on outer beaches somewhat protected by mud flats. Size: The 35 specimens in the present collection measure between 3.9 and 24.5 mm. in length of carapace. The male co-type measured 43.2 mm. long. Color: This species is much less highly colored than gaudichaudii, being always very pale, and always found on light sand. The carapace and legs of adults show marblings of gray and white on olive; the joint between merus and carpus of chelipeds is darkest, usually brownish; manus of chelipeds, all dactyls and underparts white. The young are much more boldly patterned, with dark brownish-gray or even black marblings and splotches on olive buff. Legs banded ir- regularly with dark brown and gray. In the smallest there is a spot of scarlet in the middle of the posterior gastric region, and another on the merus of the last three pairs of ambulatories. In specimens of around 10 mm. in length, the red spots remain only on the gastric region and the most posterior ambulatory. In large crabs, measuring around 20 mm. or more, no trace of scarlet is found. Early Crab Stages: Excluding the remarks con- cerning the specialized ocular stiles and chelae, the account of the young stages of gaudichaudii (p. 299 ff.) applies equally well to the present species, except that the stridulating ridge is not discernible in occidentalis until a length of about 6 mm. is reached. As shown in the key (p. 299), the young of the two forms may be distinguished 2 In this paper reference was inadvertently omitted to Kemp’s (1915) quotation of Annandale’s notes concerning the behavior of the megalopa of O. macrocera. Dr. Annandale found that they made rudimentary burrows under the shelter of beached catamarans, became pale in strong light, and that they were preyed upon by adult crabs and, apparently, by ants. (Kemp, “Fauna of the Chelka Lake. Crustacea Decapoda. ” Mem. Ind. Mus. Calcutta 5. 1915, pp. 219-220, text-fig.) 309 1941] Crane: Crabs of the Genus Ocypode D Text-figure 7. Major chelae of adult males in Ocypode. A, occidentalis, outer side; B, same, inner side; C, albicans, outer side; D, same, inner side; E, guadichaudii, outer side; F, same, inner side. All X 1.6. by their coloration when alive and, when pre- served, by the proportions of the ischium of the third maxilliped. From a phylogenetic point of view, it is interesting that in young occidentalis the orbital angles are even less advanced than in adults of the Atlantic species, albicans, although in adult occidentalis the reverse is true. Speci- mens measuring around 20 mm. in length appear to be mature. None of the few large specimens in the present collection seems to be near breeding condition. General Habits: This species is clearly the analogue of the Atlantic albicans, and their habits turn out to be very similar, while utterly dissimilar to those of gaudichaudii. Like the Atlantic form, the present species is wholly nocturnal when adult, although the young are also active during the day. They are typical predators and scavengers at all ages. The alimentary canals of twelve specimens were examined, from crabs measuring between 6 and 25 mm. in length, taken in eight different localities. The contents were distributed as follows: Amphipods were present in four speci- mens, beetles in four, seaweed in three, worms (setae and jaws) in two, Thysanura in two, a minute sea urchin in one, and one was empty. In four a few sand grains were mixed with the organic matter. Material: In addition to the megalopa recorded in a previous paper (Crane, 1940, p. 70), a total of 35 specimens of 0. occidentalis was taken by the Eastern Pacific Zaca Expedition in the fol- lowing localities: Banderas Bay (Cat. No. 37,132), Chamela Bay (37,165), Tenacatita Bay (37,199), Acapulco (37,278), and Port Guatulco (37,445), Mexico; Gulf of Fonseca near Potosi River (37,702) and Corinto, Nicaragua; Port Culebra (38,125), Costa Rica; Bahia Honda (38,698), Panama. V. Bibliography. Abramowitz, R. K., & A. A. Abramowitz. 1940. Moulting, growth and survival after eye- stalk removal in Uca pugilator. Biol. Bull., Vol. 78 (2): 179-188. 1 fig. Alcock, A. 1892. On the stridulating apparatus of the red Ocypode crab. Ann. Mag. Nat. Hist. (6), Vol. X, p. 336. Anderson, A. R. 1894. Note on the sound produced by the ocypod crab ( Ocypoda ceratophthalma) . Jour. Asiat. Soc. Bengal, Vol. LXVIII, pp. 138-9. Beebe, W. 1938. Eastern Pacific Expeditions of the New York Zoological Society, XIV. Introduc- tion, Itinerary, list of stations, nets and dredges of the Eastern Pacific Zaca Expe- dition, 1937-1938. Zoologica, Vol. XXIII, pp. 287-298. Boone, L. 1927. The littoral crustacean fauna of the Galapagos Islands. Part I. Brachyura. Zoologica 8 (4): 127-288. 1 pi., 69 fig. 1929. A collection of Brachyuran Crustacea from the Bay of Panama and the fresh waters of the Canal Zone. Bull. Amer. Mus. Nat. Hist. 58 (11): 561-583. 18 fig. Brown, F. A., Jr. 1939. The source of chromatophorotropic hor- mones in crustacean eyestalks. Biol. Bull., Vol. 77 (2), p. 329. 1940. The crustacean sinus gland and chromato- phore activation. Physiol. Zool., Vol. XIII, pp. 343-355. 1 fig. Cott, H. B. 1929. The Zoological Society’s expedition to the Zambesi, 1927: No. 3. Observations on the natural history of the racing crab Ocypoda ceratophthalma from Beira. Proc. Zool. Soc. London 1929 (4), pp. 755-765. 310 [XXVI: 29 Zoologica: New York Zoological Society Cowles, R. P. 1908. Habits, reactions and associations in Ocypoda arenaria. Papers from the Tortugas Laboratory of the Carnegie Inst. Wash. II, pp. 3-39, 4 pis., 10 text-fig. Crane, J. 1939. The Painted Ghosts of Honda Bay. Bull. N. Y. Zool. Soc., Vol. XLII, pp. 19-25. 1940. Eastern Pacific Expeditions of the New York Zoological Society, XVIII. On the post-embryonic development of braehyu- ran crabs of the genus Ocypode. Zoologica, Vol. XXV, pp. 65-82; text-figs. 1-8. 1941. Eastern Pacific Expeditions of the New York Zoological Society, XXVI. Brachy- uran crabs of the genus Uca from the west coast of America. Zoologica, Vol. XXVI, pp. 145-208; text-figs. 1-8; pis. I-IX. Dembowski, J. 1913. Uber den bau der augen von Ocypoda ceratophthalma Fabr. Zool. Jahrb. Jena Abt. fur Anat. Vol. XXXVI, pp. 513- 524; 1 pi. Gordon, I. 1934. Crustacea Brachyura. Res. Sci. du Voyage aux Indes Oiientales Neerlandaises de 11. AA. RR. le Prince et Princesse Leopold de Belgique. Vol. Ill, Fasc. 15, pp. 1-78. 37 figs. Huxley, J. S. 1931. Notes on differential growth. Amer. Naturalist 65 (699); pp. 289-315. 7 fig. Kleinholz, L. H., & Bourquin. 1941. Effects of eye-stalk removal on decapod crustaceans. Proc. Nat. Acad. Sci. Vol. XXVII, pp. 145-149. Lanchester, W. F. 1900. On a collection of crustaceans made at Singapore and Malacca. Part I, Crusta- cea Brachyura. Proc. Zool. Soc. London. 1900, pp. 719-770, pis. xliv-xlvii. Milne Edwards, H., & Lucas, H. 1843. Crustacea. D’Orbigny’s Voyage dans l’ Amer. Meiid., Vol. VI. 1847. Atlas. D’Orbiqny’s Voyage dans l’ Amer. Merid., Vol. IX. Ortmann, A. E. 1894. Die Decapoden-Krebse der Strassburger Museums. Zool. Jahrb., Jena, 2 Abt. fur Syst. VII, pp. 683-772; pi. XXIII. Parenzan, P. 1931. Osservazioni biologische sull’ Ocypoda ceratophthalma Fabr. Atti R. 1st. Veneto Sci. Lett, ed Arti 90 (9/10); pp. 1001-1008. 3 pi. 1930 (1931). Pesta, O. 1931. Ergebnisse der Osterreichischen Biolo- gischen Costa-Rica-Expedition 1930. I. Teil. Crustacea Decapoda aus Costa- Rica. Ann. Naturhist. Mus. Wien 45, pp. 173-181. 2 pi., 1 fig. 1931. Rathbun, M. J. 1917. The grapsoid crabs of America. Bull. U. S. Nat. Mus., No. 97. 1923. The brachyuran crabs collected by the U. S. Fisheries Steamer Albatross in 1911; chiefly on the west coast of Mexico. Bull. Amer. Mus. Nat. Hist., N. Y., Vol. 48, pp. 619-637; 9 pis. Sandon, H. 1937. Differential growth on the crab Ocypoda. Proc. Zool. Soc. London, 1937, pp. 397- 414; 4 figs. Sivertsen, E. 1934. Littoral Crustacea decapoda from the Galapagos Islands (1933). (Norwegian Zool. Exped. to the Galapagos Is. 1925. VII.) Nyt. Mag. Naturvidensk. 74: 1-23. 4 pi., 1 fig. 1934. Stimpson, W. 1862. Notes on North American Crustacea in the Museum of the Smithsonian Institu- tion. No. 2. Ann. Lyc. Nat. Hist. N. F., Vol. VII, pp. 176-246; 2 pis.; 1860 (1862) Takahasi, S. 1935. Ecological notes on the ocypodian crabs (Ocypodidae) in Formosa, Japan. Annot. Zool. Jap., Vol. XV, pp. 78-87; 1 pi.; 1 text-fig. EXPLANATION OF THE PLATES. Plate I. Fig. 1. O. gaudichaudii. Buccal frame of adult female, carapace length 25 mm. E marks enlarged endite of first maxilliped. Bahia Honda, Panama. X 3.5. Fig. 2. O. occidentalis. Buccal frame of adult fe- male, carapace length 24.5 mm. Port Guatulco, Mexico. X 3.5. Plate II. Fig. 3. 0. gaudichaudii beside burrow, with heap of excavated sand, feeding pellets and scraping marks of truncate chelae. Bahia Honda, Panama. Fig. 4. O. gaudichaudii. Right abdominal append- age, right lateral view. Carapace length 25 mm. Panama City, Panama. X 6.6. Fig. 5. 0. occidentalis. Right abdominal appendage, right lateral view. Carapace length 25 mm. Chamela Bay, Mexico. X 9.2. Fig. 6. 0. albicans. Right abdominal appendage, right lateral view. Carapace length 24.5 mm. Bermuda. X 6.4. CRANE. PLATE I. FIG 1. FIG. 2. ON THE GROWTH AND ECOLOGY OF BRACHYURAN CRABS OF THE GENUS OCYPODE CRANE PLATE II. ON THE GROWTH AND ECOLOGY OF BRACHYURAN CRABS OF THE GENUS OCYPODE. 1941] Edgerton & Breder: High Speed Photographs of Flying Fishes in Flight 311 30. High Speed Photographs of Flying Fishes in Flight. H. E. Edgerton Massachusetts Institute of Technology & C. M. Breder, Jr. New York Aquarium (Plates I-VIII). Introduction. The study of the mechanism of flight of the Exocoetidae has long been handicapped by a lack of good photographs of the performance. Such knowledge as we have of the factors involved has been based on simple observation and the inter- pretation of studies on the morphology of the Exocoetidae. That this was still in a not alto- gether satisfactory state led Breder (1937 and 1938) to remark rather sharply about the general attitude on the problem. Further no one can deny that there are still a host of details, a knowledge of which would be both of practical aerodynamic and academic value. The first satisfactory photographs of flying fishes in various stages of flight are presented herewith with such interpretations as may be made from them. The photographs were taken on the east coast of Catalina Island, California, at night, during July, 1940, by one of the authors, Edgerton. They all represent the species Cypselurus californicus (Cooper) and are all results of the high-speed electrical flash photog- raphy method developed at the Massachusetts Institute of Technology. The value of this device for biological pursuits concerning loco- motor matters involving high speeds is obvious. See Edgerton & Killian (1939) for a compre- hensive bibliography on technique. If it had been practicable, high-speed motion pictures would have been even more valuable and it is to be hoped that such may be made in the future. As it is the stills presented herewith show a host of items unsuspected and illuminate a variety of others which for long have resided in the limbo of half-knowledge. The time of exposure in all cases is about 1/10,000 of a second, which is sufficiently short to “stop” all motion. Acknowledgment is made of the splendid co- operation of the Catalina Island Company which furnished the motor boat Blanche W. This ship is equipped with a 110 volt D. C. generator and search light for observing the flying fishes at night. A small converter was used to produce alternating current to operate the electrical flash equipment. Two flash lamps were used in parallel. One of these was an experimental lamp about four feet in length (in two sections) in a cylindrical reflector. The other was a spiral lamp in a spherical reflector that was loaned for the occasion by the Los Angeles Herald-Express.1 The straight tube was mounted on a pipe-work frame as far out over the water as possible, while the other was hand-held and directed at the fish that was photographed. A miniature Speed Graphic camera was used. A photograph is shown of the entire arrangement on the foredeck of the Blanche W in Plate I, Fig. 1. Analysis op Plates. The details which these photographs show are analyzed in the following section while the re- sulting interpretations are given under the heading “Discussion,” both the remarks and interpretations being those of Breder. Plate II, Fig. 2. The fish in full flight. Since the right pectoral is evidently arched upward near its center of pressure it would appear that the fish is rolling to that side with the left pectoral high. At the same time the tail is being swung to the right as is evidenced by the weaker central rays bending to the left. This would then represent normal resistance to extrinsic turning with banking to the right. Since the effects are 1 This second lamp is the same as the Eastman Koda- tron Speedlamp. 312 Zoologica: New York Zoological Society both slight, as is evident, it should represent an incipient turn of large radius which is being corrected. It may be noted that the right pelvic also appears to be pressed upwards with the left normal. The dorsal fin so far as can be seen in this picture seems unaffected. This shows well the extent of expansion of the wing surfaces under actual flight. Dead and preserved material give the impression of much less wing area due to shrinkage of the delicate membrane between the supporting ribs. Plate III, Fig. 3. Just before emergence the paired fins may be seen closely appressed to the sides, while the upper caudal lobe already raises a wake, throwing spray high in the air before the head of the fish breaks the surface. Plate III, Fig. 4. Similar to Fig. 3 but with even greater spray throwing and the fish de- scribing a curve. It is doubtful if this attempt actually lead to a flight. Plate IV, Fig. 5. Two fish just at emergence. The pectorals are lifted but the pelvics seem to be only about half extended. At the extreme left are to be seen the first few flicks of spray raised by the upper caudal lobe. The emergence of the snout and body is represented by the heavy mass of spray. Note the extent of travel by the fish before the earliest spray has fallen. Plate IV, Fig. 6. Full lateral view of a fish just before taking off. Both pectorals show a large amount of arching due to the pressure on them just before flying speed is obtained. The pelvics are only partly, if at all, unfolded. When this occurs the tail lifts and the greater amount of supporting surface relieves the load on the pectorals. Plate V, Fig. 7. A diagonal view of a fish about to take off. The one visible pelvic appears to be about one-half expanded. Note the zigzag track which marks the “taxi” period and gives some idea of the value of each tail thrust in terms of the fish’s length. Plate VI, Fig. 8. A fish coming head on while in the “taxi” period. Apparently the pelvics are still being held close to the body for, if other- wise, in this photograph they should be con- spicuous. As usual during this time the fish appears to be rolling badly, at the moment bearing down on the right pectoral and lifting the left. As some measure of the vigor of this rolling movement the flexible posterior margin of this fin is clearly bending down as the fin presses upward. It is this rolling that has given rise to the oft-repeated claim of wing flapping flight in these fishes. Plate VII, Fig. 9. A fish just having cleared the water and steadying off to a smooth glide. The pelvics are still at a sharp angle pressing the tail upward. A moment later they would be ap- proximately parallel to the pectorals. Plate VII, Fig. 10. A second view of essentially the same position as that of Fig. 2 but not as marked a turning. The left pectoral of this fish has been damaged on its posterior border. Another specimen in the background is just [XXVI: 30 leaving the water with the long lower caudal lobe still immersed. Plate VIII, Fig. 11. A fish not quite main- taining flying speed. The body is somewhat arched in a vertical plane anticipatory to the tail’s touching the surface of the water, when the “taxi” stage will be resumed and flight continued without complete submergence. Discussion. The photographic evidence here recorded is in good agreement with recent descriptive inter- pretation, e. g. Breder (1930), Hubbs (1933, 1935 and 1937), Carter & Mander (1935), Forbes (1936), and Loeb (1936). Reference to these papers shows that nearly all of the descriptive details given by them are evident from these photographs. In addition there are a number of items that could not be made out by field ob- servation or interpretations based on anatomical study. For example, in Figs. 6, 7 and 8 it may be seen that the mouth is held open. In all the others this feature cannot be distinguished because of position or some other reason. If this can be taken to mean that flying fishes usually or normally hold their mouths open during flight it is to say the least surprising. Even that they ever do was not expected. Two reasons for holding the mouth closed would be to preserve intact the streamline form of the head and to prevent the desiccating effect of a current of air passing over the delicate gill membranes. How- ever this may be it is clear that these fishes do not always preserve form and conserve moisture by this simple expedient. The two pictures of Plate III showing the fishes completely submerged and swimming nearly parallel with and very close to the surface indicate another unexpected feature; that of throwing spray with their tails before breaking the surface. All this happens so quickly that under mere ocular observation it is normally lost in the general flurry of the fish’s plunging out into the air. Checking back, however, the proportions of these fish are such that the upper caudal lobe would project through the surface if they were swimming close to and parallel to it. Actual measurements show that the tail tip of a Cypselurus californicus 12 inches in standard length is a little over % inch above the level of the flat back while the dorsal fin when erect is not quite x/i inch above the back. In other words a fish 12 inches long would be just breaking the surface film with its tail tip when swimming with z/± inch of water over its back. This would seem to indicate that these fishes travel in this fashion just prior to emerging for some little distance, for otherwise it would be very unlikely that the photographer could get such photo- graphs at all. Their sometimes apparent rela- tively steep angle of emergence is then probably generally due to the rapid tip up of the snout as the unfolding wings encounter air resistance. An angle of emergence of more than about 4° would prevent the tail and dorsal from breaking 1941] 313 Edgerton & Breder: High Speed Photographs of Flying Fishes in Flight the surface before the snout came out. Inci- dentally these photographs show very good agreement with the sketches of Hubbs (1933). In considering the various flexures of the fin rays that are shown in these photographs it must be borne in mind that these are long, thin, more or less flexible rods that are attached only at their bases and there operated by relatively small muscle bundles. Since the fish can only wave these rods about in various ways, mostly backward and forward and to a lesser degree up and down and since they have no voluntary control over the curvature of them it follows that the contortions shown are the result of wind pressure. This being the case the interpretation of the meaning of the curves seen becomes rela- tively simple. Even, as in these cases, where we have no direct knowledge of what the fish was doing before or after the instant of exposure, these curves, as indicated under “Analysis of Plates,” give clues as to what the fish was undertaking. In Fig. 11, for example, it is clear that the fish is descending and is all set for a renewal of the “taxi” period. Note that the wings are relatively straight and show a large lateral dihedral, for stability, while in Fig. 6, where the fish is driving ahead under the impulse of power- ful tail thrusts, the wings held at a lower angle (see the basal part) are actually blown back and up and cupped by wind pressure. In other words they are loaded relatively more heavily than, as in Fig. 11, where the fish is merely falling freely. Many other points already fairly well under- stood are reinforced by the details in these photographs. The next logical step looking toward a further clarification of exocoetid flight would be the application of high-speed cinema- tography whereby actual series of steps in this performance could be studied. Carter & Mander (1935) used motion picture technique to check the speed of flight but presumably they obtained a very small image not of value in studying details of manipulation of the fishes’ structures. Recently Woodcock (1940a and 1940b) has discussed the instability of air over the ocean showing that bands of updraughts occur, by an ingenious noting of the differential behavior of soaring gulls. These bands are responsible for the lines of Sargassum commonly found in the Atlantic according to the studies of Langmuir (1938). Since exocoetids may fly in a straight line or in various curving flights it may well be that these too are taking advantage of such atmospheric characteristics instead of merely being blown off their course as has been generally assumed. Carter & Mander (1935) indicated that they found their fish flying greater distances over rough water than over smooth and inferred that advantage was taken of the greater air turbulence in the former condition. Hubbs (1933 and 1936) could not find a difference in duration of flight to be correlated with travel over smooth or rough water. In any case the situation calls for a study of the flight of these fish in the light of the work of Woodcock on gulls and in reference to the general recent advances of micrometerology. Bibliography. Breder, C. M., Jr. 1930. On the structural Specialization of Flying Fishes from the Standpoint of Aerody- namics. Copeia, No. 4: 114-121. 1937. The Perennial Flying Fish Controversy. Science 86 (2236) : 420-422. 1938. A Contribution to the Life Histories of Atlantic Ocean Flyingfishes. Bull. Bing- ham Oceanographic Coll. 6 (5) : 1-126. Carter, G. S. & Mander, J. A. H. 1935. The Flight of the Flying-Fish, Exocoetus. Rep. Brit. Assn. Adv. Sci., 105: 383-384. Edgerton, H. E. & Killian, J. R. 1939. Flash. Hale, Cushman & Flint. Boston. Forbes, A. 1936. Flying Fish. Science, N. S., 83: 261-262. Hubbs, C. L. 1933. Observations on the Flight of Fishes, with a Statistical Study of the Flight of the Cypselurinae and Remarks on the Evolution of the Flight of Fishes. Pap. Mich. Acad. Sci., Arts and Letters, 17: 575-611. 1935. Nature’s Own Seaplanes. Smiths. Rep., 1933:333-348. 1937. Further Observations and Statistics on the Flight of Fishes. Pap. Mich. Acad. Sci., Arts and Letters, 22: 641-660. Langmuir, I. 1938. Surface Motion of Water Induced by Wind. Science, 87: 119-123. Loeb, L. B. 1936. The “Flight” of Flying Fish. Science, N. S., 83: 260-261. Woodcock, A. H. 1940a. Observations on Herring Gull Soaring. Auk. 57: 219-224. 1940b. Convection and Soaring Over the Open Sea. Jour. Mar. Res. 3: 248-253. 314 Zoologica: New York Zoological Society [XXVI: 30 EXPLANATION OF THE PLATES. Photographs by Dr. H. E. Edgerton. Taken at night by means of a Kodatron type of electrical flash lamp. Photographs all unretouched. All specimens are Cypselurus californicus (Cooper). Plate I. Fig. 1. The photographic equipment and its ar- rangement as used for taking pictures of flying fish in flight. Plate II. Fig. 2. A fish in full flight. The object in the upper left corner is part of the special illuminating device. Plate III. Fig. 3. Just before emergence, showing that the water is splashed before the fish breaks the surface. Fig. 4. A more advanced stage in the water splashing period. Plate IV. Fig. 5. Two fish just about to leave the water near the end of the “taxi” stage. Fig. 6. Full lateral view of a fish in the “taxi” stage. Plate V. Fig. 7. A diagonal view of an advanced “taxi.” Note the trail left by the oscillating tail. Plate VI. Fig. 8. Head on view of a fish in an advanced “taxi.” Note the evidences of roll in the differentially warped wings. Plate VII. Fig. 9. Just as the tail raises after the “taxi” stage. Fig. 10. One fish in full flight in the foreground. Note the torn left wing. Another in the background with only the long lower caudal lobe immersed. Plate VIII. Fig. 11. Losing flying speed and about to dip the tail in the water for a resumed “taxi. ” EDGERTON & BREDER PLATE HIGH SPEED PHOTOGRAPHS OF FLYING FISHES IN FLIGHT EDGERTON & BREDER PLATE II. FIG. 2. HIGH SPEED PHOTOGRAPHS OF FLYING FISHES IN FLIGHT. FIG. EDGERTON & BREDER t;pFFn PHnTnr^RtPHF of fi yimh fifhfq iim fi inHT EDGERTON 8c BREDER PLATE HIGH SPEED PHOTOGRAPHS OF FLYING FISHES IN FLIGHT. EDGERTON & BREDER. PLATE VI. FIG. 8. HIGH SPEED PHOTOGRAPHS OF FLYING FISHES IN FLIGHT. EDGERTON & BREDER PLATE VII. HIGH SPEED PHOTOGRAPHS OF FLYING FISHES IN FLIGHT. HIGH SPEED PHOTOGRAPHS OF FLYING FISHES IN FLIGHT. 1941] Index to Parts 1-4 315 INDEX. Names in bold face indicate new genera, species or varieties; numbers in bold face indicate illustra- tions; numbers in parentheses indicate the paper containing the Plate numbers immediately following. A Acanthias vulgaris, 121 Acantbicolepis longicirrata, 26, 27 Achirus lineatus, 243, (25) Plate I ASgithus burmeisteri, 283 punctatissimus, 282 separandus, 283 Aetobatus californieus, 271, 271 peruvianus, 271, 271 Alaudidae, 241 Allassostoma parvum, 74 Alopecias barrae, 97 longimana, 97 Alopias vulpinus, 97 Ammotrypane bermudiensis, 23, 29 Amphistoma lunatum, 65, 85 Anas epecutlri, 65 melanotus, 65 moschata, 80 platyrhynchos, 65 Anchoviella, 96 Anoptichthys, 289 (28) Plate I jordani, 123, (16) Plates I-IV Aphrodita, 21 Aprionodon fronto, 105, 105 Arabella pacifica, 19, 23 Argyreiosus vomer, 31, 37, 37, 38, opp. 46 Astyanax mexicanus, 123 Autolytus bidens, 27, 28 B Basiliscus galeritus, 160 Betta splendens, 55 Blennius montagui, 235 sphynx, 235 Bos taurus, 65, 79 Brachioptilon hamiltoni, 276 Brachycoelium, 83 Brachysphoenus ardens, 284, (27) Plate I duplicatus, 285 moniliferus, 284 Branchiostoma californiense, 89, 89 caribaeum, 90 elongatum, 90 florldae, 90 virginiae, 90 Busycon perversum, 230 C Callorhinchus callorhynchus, 279 smythii, 279 tritoris. 279 Cambarincola macrodonta, 4 Cambarus blandingii, 1 b. acutus, 1, 3 b. cuevachicae, 1, 3 Carangidae, 31, opp. 46 Caranx bartholomaei, 42, opp 46 caninus, 107, 221 crysos, 43, opp. 46 guara, 31 latus, 43, 44, opp. 46 ruber, 42, 43, opp. 46 Carcharhinus lamiella, 111 Carcliarias aethiops, 104 fronto, 106, 110 gracilis, 104 platyrhyncbus. 111 pugae, 104 Carcharinus natator, 108 Carcharodon carcharias, 98 Caretta caretta, 13 Centroscyllium nigrum, 120. 120 ruscosum, 121 Cephaloscyllium uter, 99, 99 Cephalurus cephalus, 100, 100 Cercaria poconensis, 65 Cervus dichotomus, 65 Cestracion pantherinum, 119 Cetorhinus maximun, 98, 98 Chaetodon humeralis, 221 Chelonia mydas, 13, (5) Plates I-IV Chlamydoselachus anguineus, 96 Chloeia euglochis, 18 Chloroscombrus chrysurus, 32, 39. 40-42, opp 46 Clinus argentatus, 234 Clonorchis sinensis, 14 Corydoras rabauti, 5 Cotylophoron cotylophorum, 65 Crossoptilon, 143 Ctenosaura similis, 160 Cypselurus californieus, 311, (30) Plates I- VIII D Dasyatis brevis, 260, 261, (26) Plate III hastatus, 49, 50, 51, 52, 53 longus, 261, 262 pacificus, 262. 262, (26) Plate II sabinus, 49, 50, 51, 52, 53 schmardae, 263 Dasychonopsis conspersa, 29 Decapterus macarellus, 31, 32, 32, opp. 46 punctatus, 31, 33. 33, 34, opp. 46 Delacourigallus, 143 Diardigallus, 143 Dicrocoelium dentriticum, 14 Diodon holacanthus, 113 hystrix. 111 Diopatra ornata, 22 Diplodiscus subclavatus, 74 tempera tus, 70 Discopyge ommata, 249, 249, (26) Plate I tschudi, 250, 250 Distomum constrictum, 13 E Entocythere cambaria, 4 Epimachus meyeri meyeri, 47, (9) Plate I Erotylidae, 281, (27) Plate I Erotylus variegatus, 283 Eulalia magnapupula, 19, 21 megalops, 26 Eulamia aethalorus, 106, 107, (15) Plate II azureus, 106, 109, 109 cerdale, 109, 109 commersonii, 111 galapagensis, 111, 112, 112 lamiella, 110, 110 316 Zoologica: New York Zoological Society limbatus. 108 milberti, 106, 110 platyrhynchus, 111, 111 velox, 108, 108, (15) Plate II Eupholoe nuda, 19, 20 Eupolymnia magnifica, 29 Eupomotls gibbosus, 55 Eurythoe complanata, 18 dubia, 18 oculata, 18, 19 Exocoetidae, 311 F Farranula gibbula, 213, 224 rostrata, 219, 224 Fasciola hepatica, 14, 69 Fundulus heteroclitus, 237, (23) Plates I-III G Galeocerdo arcticus, 113, 113 Galeorhinus chilensis, 114 galeus, 114 molinae, 114 sp., 114 zyopterus, 114 Garrulax chinensis, 241 c. lochmius, 241 Gastrodiscus aegyptiacus, 74 Gelasimus vocatur, 178 Gennaeus edwardsi, 143 imperialis, 143 Ginglymostoma cirratum, 9, 9-12, 96, 96, (4) Plates I, n Grapsus grapsus. 151 Gymnosarda alletterata, 113 Gymnura marmorata, 263, 263 H Haemonchus contortus, 80 Haliseetus leucocephalus leucocephalus, 7, (3) Plates I- IV Halosydna brevisetosa, 20 Hapalotrema constrictum, 13 „ mistroides, 13 Haplosyllis gula, 29 Harriotta curtiss- jamesi, 278, 278 Helisoma antrosum, 65 trivolvis, 65 Hemipodus mexicanus, 23 Heptranchias, 96 Hermodice carunculata, 17, 25 Heronimus chelydrae, 69 Heterodontus francisci, 117, 117 galeatus, 120 peruanus, 119 quoyi, 113, 117, 118, 120 Heterostichus rostratus, 235 Hexanchus griseus, 96 Hierophasis swinhoii, 143 Himantopus wilsonii, 65 Histiopborus immaculatus, 209 pulchellus, 209 Houpifer, 143 Hyalinoecia juvenalis, 22 Hypoprion brevipinnis, 106 Hypselonotus gibbosus, 283 sp. A; sp. B; sp. C, 283 (27) Plate I I Iguana iguana, 113 Ilyoplax, 306 Isistius brasiliensis, 121, 121 Istiophorus, 209, 210 americanus, 209, 219, 219, 222, 226, (20) Plate V [XXVI greyi. 209, 210, 213, 214, 216, 217, 219, 221. 221, 222 (20) Plates I-IV Isurus glaucus, 98 nasus, 98 L Lamna hudobroii, 98 philippi, 98 Learedius learedius, 13 Lebistes reticulatus, 55. 243, (25) Plate I Leodice culebra, 28 longisetis, 22 mutilata, 28 paloloides, 22 stigmatura, 28 Lepidasthenia elegans, 19, 19 picta, 18 Lepidocephalus guntea, 58 Lepidonotus pilosus, 26 Lepomis auritis, 55 Lobiophasis, 143 Lopadorhynchus uncinatus, 27 Lophura. 143 Lybas calidus, 285 dorsalis, 285 guianacus, 285, (27) Plate I triangularis, 285, (27) Plate I M Manta birostris, 276 hamiltoni, 274, 274, (26) Plate IV pinchoti, 276 Megalodiscus, 83 Megischyrus catenatus, 287 Mictyris, 306 Mirafra assamica, 241 a. subsessor, 241 Mobula lucasana, 273, 273, (26) Plate IV Moniezia, 78 Monocirrhus polyacanthus, 235 Morone labrax, 55 Mus norvegicus, 65 Mustelus californicus, 102, 102 dorsalis, 103, 103 lunulatus, 103, 103 Mycotretus coccineus. 286 dorsonotatus, 286 durius, 287 maculatus, 287 pygmaeus, 286 sanguinosus, 286 Mycteroperca jordani, 113 Mystides gracilis, 26, 27 Myxine circifrons. 90. 90 N Narcine brasiliensis, 248 entemedor, 247, 247 ommata, 248 vermiculatus, 248, 248, (26) Plate I Nereis agassizi, 27, 28 bairdii, 28 dumerilii, 28 kobiensis, 28 mirabilis, 28 Notomastus sp., 23 Notopygos ornata, 18 O Ocypode aegyptica, 300 albicans, 298, 309, (29) Plate II ceratophthalma, 300, 303 cordimana, 306 gaudichaudii. 176, 297, 299, 299, 300, 302, 304, 305, 309, (29) Plates I, IE occidentalis. 297, 299, 301, 302, 304, 305, 308, 309, (29) Plate II 1941] Index to Parts 1-4 317 Ogcocephalus sp., 113 Omoiotelus, 281 pallidus, 282 Opisthorchis felineus, 14 Opsanus beta, 229, (21) Plates I, II pardus, 231 tau. 229 Otario jubata, 113 Otus scops, 133 s. botelensis, 137 s. coecus, 140 s. cycladum, 135 s. cyprius, 136 s. distans, 138, 138, 139 s. elegans, 137 s. feae, 140 s. flammeolus, 141, 141 s. graueri, 140 s. hendersoni, 138, 141, 141 s. intermedius, 141 s. interpositus, 137 s. japonicus, 137, 137 s. latipennis, 141 s. leggei, 139 s. malayanus, 137, 138 s. modestus, 138 s. pamelae, 140 s. pulchellus, 135 s. pusillus, 141 s. pygmeus, 140 s. rarus, 142 s. rufipennis, 139 s. scops, 133, 135, 135 s. Senegal ensis, 140 s. socotranus, 140 s. stictonotus, 136, 136 s. sunia, 138, 139, 139 s. turanicus, 135 s. ugande, 140 Ovalipes punctatus, 108 Ovis aries, 65 P Pachygrapsus, 205 transversus, 151 Pandarus satyra, 113 Paraclinus marmoratus, 229, 233, 234, (22) Plates I-IH Paramphistomum cervi, 70 Pectinaria gouldii, 23, 29 Penella filosa, 221 Phricobacis beebei, 286, (27) Plate I Phyllodoce groenlandica, 21 oculata, 21 Plagitura, 83 Polyodontes californicus, 19, 20 Pomatorhinus erythrogenys, 241 e. celatus, 241 Pontinus strigatus. 111 Pontobdella muricata, 96, 109 Postharmostomum laruei, 83 Prepopharus notatus, 283 obscurior, 283, (27) Plate I undatus, 283 Prionace glauca, 104, 104 Pristiurus xaniurus, 99, 100 Pritis 7 pectinatus, 253 perotteti, 253 zephyreus, 253, 253 Psammobatis brevicaudatus, 259, 259 lima, 259, 259 scobina, 258, 258 spinosissimus, 259, 260, (26) Plate II Pselaphacus giganteus, 287 signatus, 287 Pteromylaeus asperrimus, 272, 272 Puffinus auricularis, 113 R Raja aguja, 254, 254 badia, 254. 255 cliilensis, 256 ecuadoriensis, 255, 255 equatorialis, 256 inornata, 256, 256 sp. A, 257, 257 sp. B, 257. 258 steindachneri, 255 stellulata, 258 Rhineodon typus, 97, 97, (15) Plate I Rhinobatus glaucostigma, 251, 251 leucorhynchus, 252, 252 planiceps, 251, 251 productus, 251, 251 Rhinoptera steindachneri, 273, 273 Rhytidodoides intestinalis, 14 similis, 14, 15, (5) Plates I-IV Rivulus, 243 Rocinela aries, 107, 110, 113 Rupiscartes atlanticus, 96 S Sabella melanostigma, 23 Sardinia caerulea, 107 Scaphidomorphus quinquepunctatus, 283 Scoliodon longurio, 112, 112 Scopimera, 306 Sphyrna corona, 115, 116 media, 116, 116 tiburo, 116 tudes, 115, 115, 116 vespertina, 116, 116 Spirobranchus tricornis, 29 Squalus fernandinus, 121 lebruni, 121 sp., 121 suckieyi, 121 Squatina armata, 122 californica, 121, 121 californicus. 122 philippi, 122 squatina, 122 Stoasodon narinari, 272, 272 Stunkardia dilymphosa, 85 Stylarioides sp., 23 T Tarpon atlanticus, 55 Telake epipolasis, 29 Terebella gorgonae, 23 Terebellides stroemi, 23 Timaliidae, 241 Tomopteris longisetis, 28 opaca, 22 Trachinotus goodei, 35, 36, opp. 46 palometa, 34. 35, opp. 46 Trachurops crumenophtkalma, 38, 39, opp 46 Travisiopsis atlantica, 22, 28 Triaenodon obesus, 101, 101, (15) Plate II smithii, 102 Triakis maculata, 101 semifasciatum, 100, 101 Trimeresurus atrox, 62 barbouri, 61 bicolor, 61 dunni, 62 godmani, 61 lansbergii, 61 lateralis. 61 melanurus, 63 nasutus, 62 nigroviridis aurifer, 62 nummifer nummifer, 62 ophryomegas, 62 schlegelii. 61 318 [XXVI Zoologica: New York Zoological Society undulatus. 63 yucatanicus, 62 U Uca, 145. 166, 297, 306 annulipes, 152 argillicola, 149, 164, 183, (19) Plates I, II batuenta, 149, 164, 187, 188, (19) Plate VI beebei, 149, 164, 192, (19) Plates IV-VI brevifrons, 149, 164, 177, (19) Plate VII var. delicata, 178 coloradensis, 166 crenulata, 149, 164, 198 deichmanni, 149, 164, 199, (19) Plates IV-VI ecuadoriensis, 166 festae, 166 galapagensis, 145, 149, 164, 176 guayaquilensis, 167 helleri, 145, 149, 164, 198 heteropleura, 149, 164, 171 inaequalis, 149, 164, 185, (19) Plates n, III insignis, 149, 173 latimanus, 149, 162, 163, 164, 201, (19) Plates VI- vm leptodactyla, 150 limicola, 149, 164, 198, (19) Plates IV-VI macrodactyla, 149, 164, 178 minax, 157 monilifera, 165 mordax, 149, 162-164, 176 musica, 166, (19) Plate VI oerstedi, 149, 162-164, 184 panamensis, 149, 164, 204 princeps, 149, 162, 163, 170 pugilator, 150 pygmaea, 149, 164, 174, (19) Plates I, II rectilatus, 167 saltitanta, 149, 164, 189, (19) Plates II, m, VI signata, 152 stenodactyla, 149, 164, 195, (19) Plates IV-VI, IX stylifera, 149, 171 tangeri, 152 tenuipedis, 149, 164, 186, (19) Plates n, m terpsichores, 149, 164, 202, (19) Plates IV-VII thayeri, 183 tomentosa, 149, 164, 179, 180 umbratila, 149, 159, 164, 181, 182, (19) Plate VII zacae, 149, 164, 175, (19) Plates I, II Urobatis concentrlcus, 268, 269, (26) Plate III halleri, 113, 269, 269 maculatus, 270, 270 Urolophus asterias, 265 rnundus, 264 Urotrygon aspidurus, 264 asterias, 265, 265 binghami, 266, 266 chilensis, 267, 267 goodei, 267 mundus, 268, 268 rogersi. 265 V Vanadis fuscapunctata, 27 Verongia flstularis, 230, 233 X Xiphias gladius, 223 Z Zapteryx exasperata, 252, 253 Zonarius xanthomelas, 284 Zygocotyle ceratosa, 65 lunata, 65, (13) Plates I-IV NEW YORK ZOOLOGICAL SOCIETY General Office: 630 Fifth Avenue, New York City OFFICERS President, Fairfield Osborn First Vice-President, Alfred Ely Chairman, Executive Committee & Second Vice-President, Laurance S. Rockefeller Treasurer, Cornelius R. Agnew General Director, Zoological Park and Aquarium, Allyn R. Jennings Assistant General Director, Harry Sweeny, Jr. Zoological Park Raymond L. Ditmars, Curator of Reptiles and Insects Lee S. Crandall, Curator of Birds Claude W. Leister, Curator of Mammals and Educational Activities Leonard J. Goss, Curator of Health William Bridges, Editor and Curator of Publications Aquarium in the Zoological Park Charles M. Breder, Jr., Director Christopher W. Coates, Aquarist Ross F. Nigrelli, Pathologist G. M. Smith, Research Associate in Pathology Homer W. Smith, Research Associate in Physiology Myron Gordon, Research Associate in Genetics Department of Tropical Research William Beebe, Director John Tee-Van, General Associate Gloria Hollister, Research Associate Jocelyn Crane, Technical Associate Consultant, Jean Delacour SCIENTIFIC STAFF Editorial Committee Fairfield Osborn, Chairman Allyn R Jennings William Beebe Charles M. Breder, Jr. Harry Sweeny, Jr. Jean Delacour William Bridges S'* - * ^SlSONIAN INSTITUTION LIBRARIES 3 9088 01 405 8994